[opensuse-factory] Road-map for openSuse 13.3?
Dear openSuse, is there a road-map for the next openSuse release 13.3? I could not find one at the usual location https://en.opensuse.org/openSUSE:Roadmap According to the 8 months plan and the last release in November, I could expect early June for the next release. Obviously, this is not the plan. What's the plan instead? If you make changes, please tell us users. Bye Christoph -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/28/2015 10:09 AM, Christoph Grüninger wrote:
Dear openSuse,
is there a road-map for the next openSuse release 13.3? I could not find one at the usual location https://en.opensuse.org/openSUSE:Roadmap
According to the 8 months plan and the last release in November, I could expect early June for the next release. Obviously, this is not the plan. What's the plan instead? If you make changes, please tell us users.
Bye Christoph
I *think* they decided to go with 12 months instead? Someone else will have to confirm but that's what I remember reading on here. - -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVP6QuAAoJEM66EOTZRH6+6a8P+gKPtL7/k8BA4SA5/RWW+pHc qJ38QD6FapntNYIjnEdA9rjo+jJpO7ioXpHDNQ14lDlgF9V704XemVeciUFZbSfH D41+PAXsTmpDENkqqjuK+SwxyhAkNqhWO19Ll8E+ot5Dv1F4kRS5mwMpnuXxYnzV bYOViasW6DVVcZyu4NVlWPkVMPfCFqZPy0oU+qaAzUU2axtShTUZLVsLcK5QMTqA DLm9UtTYsmAszt5ArexZmeX1aRTM4pBlhv6jtBISZJD0gt6fjhBrnGti+WEMhD5c DsutrOSVT7KgoaPd/ynpetRce/vVLd09LhpDQqrAPvGmzbBZxswrH91NOM4eBXqH pZ2FnhzQtbm2n2KVb35t+RGSualNC8SYyVyFk8wVdr5MqNbkfHjwOcw/gZ+m/NnF YohrFHbYnkDGytX2e2Gbwcpulf+PRMhph5lNKNiewLIa5uHewYvT1OeiFf0xKFxn s2cNQ740HfbhNvC51P8kjpAoGZdOfw9AaBfg2yvNRgPveSfjEpdUnYjqR7VmA85O H3ZGWEEDJkccN0630lHkphvV+h1wkZfp7RTnZKWb1TpUy0wdQY2EcQ25kzic/oIJ fqwGS/XUYP6opMpekeAcSk+qvIqiLC7tqZm9gaDtHdF1iEWdhsOba4BSVrKzbwSV bUrQP2CjSJfDgZXgfibd =n+wM -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Tue, Apr 28, 2015 at 11:09 AM, Christoph Grüninger <foss@grueninger.de> wrote:
Dear openSuse,
is there a road-map for the next openSuse release 13.3? I could not find one at the usual location https://en.opensuse.org/openSUSE:Roadmap
According to the 8 months plan and the last release in November, I could expect early June for the next release. Obviously, this is not the plan. What's the plan instead? If you make changes, please tell us users.
Unless I am missing something, the release schedule has not been discussed yet on this list. My personal opinion is to encourage people to switch to tumbleweed snapshots who satisfy a yet to be discussed release criteria, so let's say next release would be openSUSE 2015.$quarter .. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez <crrodriguez@opensuse.org> wrote:
On Tue, Apr 28, 2015 at 11:09 AM, Christoph Grüninger <foss@grueninger.de> wrote:
Dear openSuse,
is there a road-map for the next openSuse release 13.3? I could not find one at the usual location https://en.opensuse.org/openSUSE:Roadmap
According to the 8 months plan and the last release in November, I could expect early June for the next release. Obviously, this is not the plan. What's the plan instead? If you make changes, please tell us users.
Unless I am missing something, the release schedule has not been discussed yet on this list.
No definitive answer, but definitely discussion: http://markmail.org/thread/cnzilu3erwkjekvi There was even a community poll with 1278 votes currently shown. https://plus.google.com/+openSUSE/posts/g5FdGJ9iSqB Most popular by far was a 1 year release schedule. Coolo said he'd rather not do a release at all and just use Tumbleweed. http://markmail.org/message/iiheijcgclfvb2cy I'd say at this point the answer is "openSUSE 13.3 will not be released before Nov 2015". Greg -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-( - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVAIk4ACgkQja8UbcUWM1xnKQD/eD2Vzyd2K5q/EgOmitr9vmp1 ljS13fxO/rdROvf2pzQA/ivAOeW3WEveuUB/gkMNVdjUu+zyBJ95O+aL6Qe1KQhT =2g7J -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/28/2015 08:14 PM, Carlos E. R. wrote:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
Heh, very true. I have tried it, really great actually but I think I prefer 13.2 as my stable environment :) - -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVQCwQAAoJEM66EOTZRH6+y8MQALVQAHxZ5NNNIAeTMtCpqQgT 3ozALeK7lq1+mAP121hzvBl5QvuLE8ctJgJJDqphjJIW9VqqmXjOtqm6nWSvo74e v4aDLI21bzWOpb/821IiseRMBivNNpEFzk+57Oz47imFSgt6rPmCGWP/Vfa26mFA sXiGdSGqd5ipn9/gksWA01DJW7AfMSpGYwNizitBNDzangm2TDldKiHNAJXMmfoS oSG0QKQ8OVHjhiE6R5AIjFORMqAG83ORY7XnEkoEVCsNZY+6SwHMlMnECnxg4eFR YVr0TZYPnvQFMl7oEtNX4COps8tPUu5q8SD3UVWUWFnnzOJDXE2Wu30MCQlco8qd 9cuMKZLa/iqxZWuGpk8rIQeEk80bYKJfCm5MOyKosF5NlKRIaOvNZgt3XFrxZBAb YkQObBtLbBTlwdgwv64vCYKgAZMiN4X4yRciUSMxnAS9VLSfOsXlzjVzuIXY9BjC MpP0uZg/KzZjeGcGYItJ2mKZz0aTniGbJpz0nHQl6qKPZM/IAFtNWa6R+O73JpCY JQY+gK2829Dbza+c/+EBTGSEXH125XPY8ewZAApL3aIHZU0l0ROv2me/mnqqVlpK GE/9jPig04uW87VXdTLrfEn8JOMPJHLFL5I0kswaXkSJsiapwKJzzYsgAn1lz8u5 J4w68GsRhmJO0u+8dMB0 =Gvrj -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 29.04.2015 02:14, Carlos E. R. wrote:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
Carlos E. R.
(from 13.1 x86_64 "Bottle" (Minas Tirith))
13.2 obviously isn't either - so what's your point? Greetings, Stephan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlVAiqkACgkQwFSBhlBjoJaXfwCdEpFgyV7+kP2RQ6zB6nDw4yTJ mN4An1izhczFQCdTypQQnVzh6LWVhxeT =n4OY -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 29/04/15 01:55, Uzair Shamim wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 04/28/2015 08:14 PM, Carlos E. R. wrote:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian RodrÃguez Coolo said he'd rather not do a release at all and just use Tumbleweed. TW is not suitable for everybody :-(
Heh, very true. I have tried it, really great actually but I think I prefer 13.2 as my stable environment :)
- -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2
iQIcBAEBAgAGBQJVQCwQAAoJEM66EOTZRH6+y8MQALVQAHxZ5NNNIAeTMtCpqQgT 3ozALeK7lq1+mAP121hzvBl5QvuLE8ctJgJJDqphjJIW9VqqmXjOtqm6nWSvo74e v4aDLI21bzWOpb/821IiseRMBivNNpEFzk+57Oz47imFSgt6rPmCGWP/Vfa26mFA sXiGdSGqd5ipn9/gksWA01DJW7AfMSpGYwNizitBNDzangm2TDldKiHNAJXMmfoS oSG0QKQ8OVHjhiE6R5AIjFORMqAG83ORY7XnEkoEVCsNZY+6SwHMlMnECnxg4eFR YVr0TZYPnvQFMl7oEtNX4COps8tPUu5q8SD3UVWUWFnnzOJDXE2Wu30MCQlco8qd 9cuMKZLa/iqxZWuGpk8rIQeEk80bYKJfCm5MOyKosF5NlKRIaOvNZgt3XFrxZBAb YkQObBtLbBTlwdgwv64vCYKgAZMiN4X4yRciUSMxnAS9VLSfOsXlzjVzuIXY9BjC MpP0uZg/KzZjeGcGYItJ2mKZz0aTniGbJpz0nHQl6qKPZM/IAFtNWa6R+O73JpCY JQY+gK2829Dbza+c/+EBTGSEXH125XPY8ewZAApL3aIHZU0l0ROv2me/mnqqVlpK GE/9jPig04uW87VXdTLrfEn8JOMPJHLFL5I0kswaXkSJsiapwKJzzYsgAn1lz8u5 J4w68GsRhmJO0u+8dMB0 =Gvrj -----END PGP SIGNATURE----- In dealing with well over a dozen OS's in my time, stable seems relative and at times an oxymoron.
Rip and replace of large bits puts it in the unstable category. Fixes keep rolling in with no cast iron guarantee that they won't seriously impact some odd box out there. Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Senior Staff Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi!
TW is not suitable for everybody :-(
13.2 obviously isn't either - so what's your point?
The point is to have both. A release every 12 months would be ok. But not having a stable release at all, would be a bumper for me. Anyhow, I would expect to share such thoughts to the wider audience. I was not aware of the current spirit against stable releases. Whenever you want to change openSuse to a rolling release, please announce it or start an according discussion. But letting creepingly die the releases, is kind of betraying the users. What does the openSuse Board thinks about the recent development? Can they either rule or moderate a discussion what openSuse should aim for? Best Christoph -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 29 of April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
TW is not suitable for everybody :-(
Carlos E. R.
(from 13.1 x86_64 "Bottle" (Minas Tirith))
13.2 obviously isn't either - so what's your point?
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi!
TW is not suitable for everybody :-(
13.2 obviously isn't either - so what's your point?
The point is to have both. A release every 12 months would be ok. But not having a stable release at all, would be a bumper for me. Anyhow, I would expect to share such thoughts to the wider audience. I was not aware of the current spirit against stable releases. Whenever you want to change openSuse to a rolling release, please announce it or start an according discussion. But letting creepingly die the releases, is kind of betraying the users. What does the openSuse Board thinks about the recent development? Can they either rule or moderate a discussion what openSuse should aim for? Best Christoph -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 29.04.2015 10:58, Michal Kubecek wrote:
On Wednesday 29 of April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
TW is not suitable for everybody :-(
Carlos E. R.
(from 13.1 x86_64 "Bottle" (Minas Tirith))
13.2 obviously isn't either - so what's your point?
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades.
I'm not saying anything else - I just don't like Carlos' one sentense replies. Expressing his problems with Tumbleweed (or better the possible lack of 13.3) is one thing, but throwing statements onto mailing lists doesn't help anyone. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, 29 Apr 2015 11:12, Stephan Kulow wrote:
On 29.04.2015 10:58, Michal Kubecek wrote:
On Wednesday 29 of April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
TW is not suitable for everybody :-(
Carlos E. R.
(from 13.1 x86_64 "Bottle" (Minas Tirith))
13.2 obviously isn't either - so what's your point?
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades.
I'm not saying anything else - I just don't like Carlos' one sentense replies. Expressing his problems with Tumbleweed (or better the possible lack of 13.3) is one thing, but throwing statements onto mailing lists doesn't help anyone.
Greetings, Stephan
What is the situation? for a fixed release (version number as you like), the past has shown that six month is what it takes at minimum to get a (mostly) stable set of packages that can be called "release" without being ashamed afterwards. We've had the "great" releases, (10.2, 11.4, and even 12.3) and the "not-so-great" releases (the last was 13.2), what was the difference? For 13.2 it was the whole network-config issue that took down the overall impression. Wicked was just NOT ready yet, and this was felt like a supersonic-knell through the whole system. Every network-depended service hat its own issues with that, some could cope, some through up. For me a openSUSE release is not about a "killer-feature", but about a complete system done right. If I really want just a specific "killer-feature", I take a distro that features just that. DistroWatch and Linux magazines are great for that. The real question is: Is TW in a shape that with 6 month work it can be worthy of the title "release", or do we get another set of last-minute untested software get pressed on? Keypoints for me: * Btrfs on / is not in a shape that acceptable for the masses. - fillup: either the snapper defaults get corrected, or other config screws get turned in the right directions. - stability: Ready to run for the next 3-4 Years after install? Sadly not at the moment. - Missing repair tools, and step-by-step manuals for crashes. * Snapper docu is still not complete in the packages, config missing. * Network: - where IS the full wicked docu? Not in the packages! - up-to-date HowTo's for the different scenarios: + prefer cable, but use wifi if no cable, + vlan, + ipsec, + ipv6 (state of the art, please), + dnssec, authenificating forwarders, + firewall and virtual-machines (hello, docker!) To me it looks like that we really have to get started if we want a "release" in November. - Yamaban. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi All, I've just made an announcement on the opensuse-project mailinglist which may influence this discussion somewhat http://lists.opensuse.org/opensuse-project/2015-04/msg00104.html After my presentation on Friday, where I'll be able to go into more details of what SUSE are doing and how we may (or may not, this is all proposals at this point) wish to adopt that inside openSUSE, I think it might make more sense to revisit this topic then with that new information. Hope this helps, Richard On 29 April 2015 at 10:55, Christoph Grüninger <pr@grueninger.de> wrote:
Hi!
TW is not suitable for everybody :-(
13.2 obviously isn't either - so what's your point?
The point is to have both. A release every 12 months would be ok. But not having a stable release at all, would be a bumper for me.
Anyhow, I would expect to share such thoughts to the wider audience. I was not aware of the current spirit against stable releases. Whenever you want to change openSuse to a rolling release, please announce it or start an according discussion. But letting creepingly die the releases, is kind of betraying the users.
What does the openSuse Board thinks about the recent development? Can they either rule or moderate a discussion what openSuse should aim for?
Best Christoph -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 29 April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
13.2 obviously isn't either - so what's your point?
Why not? 13.2 has been great for me so far. As end user, who requires stable desktop, openSUSE with 6-8-12 month cycle gives the best possible solution ever. Rolling distros, like Tumbleweed or Fedora, are extremely inconvenient unless you are a developer *of* that distro. And on the other side, CentOS/RHEL is just too outdated, I do want to have latest Qt5+QtCreator, newer git/mercurial/giflib, and so on. openSUSE has this very attractive niche, being "stable but not ancient", and there is no competition in this niche as far as I know. Taking the best from both worlds, I would suggest the following: 1. Create repos for every part of the system separately. Already done (other distros can only envy). 2. Create 2 versions of those repos, stable and tumbleweed. Snapshotting from tumbleweed into stable once in a while, stabilizing that version. That is also done for most repos. That, in my opinion, gives amazing opportunity for developers to work on their piece of the system without encountering issues from other pieces. For example the development of phonon may become crazy hard if every two days gstreamer or alsa or pulseaudio gets broken or changes behaviour. 3. Create the installation media that will be configured to use 128 smaller repos instead of 2 "have-it-all" repos. That will allow changing only specific parts of the system to the tumbleweed by replacing "13.3" in the repo path to "tumbleweed". Speaking very egoistically, for me that would be the best possible option, as I will just replace the KF5 and Qt5 repos from stable to tumbleweed and have latest shiniest plasma, while not having to fix periodical kernel/systemd/Xorg/etc issues that may creep into their tumbleweeds. Of course that was very rough idea and some additional steps may be required, like some "stabilizing" repos where tumbleweed snapshots will live for some time before going to stable. I don't know all the nuances of the distro development, but I really hope that those options will be more convenient for both opensuse developers and target audience. -- Regards, Stas -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 29 April 2015 13:29:00 Richard Brown wrote:
Hi All,
I've just made an announcement on the opensuse-project mailinglist which may influence this discussion somewhat
http://lists.opensuse.org/opensuse-project/2015-04/msg00104.html
After my presentation on Friday, where I'll be able to go into more details of what SUSE are doing and how we may (or may not, this is all proposals at this point) wish to adopt that inside openSUSE, I think it might make more sense to revisit this topic then with that new information.
Hope this helps,
Richard
Hi Richard, Please make a recording of that presentation and put it online somewhere, for people who will not be able to attend. -- Regards, Stas -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 29 April 2015 at 13:52, Stanislav Baiduzhyi <baiduzhyi.devel@gmail.com> wrote:
Hi Richard,
Please make a recording of that presentation and put it online somewhere, for people who will not be able to attend.
-- Regards, Stas
We will :) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-29 10:58, Michal Kubecek wrote:
On Wednesday 29 of April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
TW is not suitable for everybody :-(
13.2 obviously isn't either - so what's your point?
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades.
Absolutely. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVA5/IACgkQja8UbcUWM1zuywEAkI04QCwOdL/9I85o5BcXrxTK pQj232NfIMABRtAvpn8A/jmOD43bwDQhsLNEjTHu5kAqF3h8/3GMIlyNC3/ApnTL =cIPd -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-29 11:12, Stephan Kulow wrote:
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades.
I'm not saying anything else - I just don't like Carlos' one sentense replies. Expressing his problems with Tumbleweed (or better the possible lack of 13.3) is one thing, but throwing statements onto mailing lists doesn't help anyone.
I don't understand your point, either. What do you want me to say? Do you want me to explain why I said that tumbleweed is not for everybody? Writing a long text about that maybe would be tiring to read. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVA6QgACgkQja8UbcUWM1yLjgEAnUrYZCdyPtVHNMG8UPmmxq8y WlpDW0bLCrCLb3/gW/QA/0R07O9PcYf7n50+Y+wtQrdMGT9oDo0k0lbXRrWc3/tE =0E/c -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 04/29/2015 07:17 AM, Carlos E. R. wrote:
On 2015-04-29 10:58, Michal Kubecek wrote:
On Wednesday 29 of April 2015 09:39:25 Stephan Kulow wrote:
On 29.04.2015 02:14, Carlos E. R. wrote:
> > TW is not suitable for everybody :-( 13.2 obviously isn't either - so what's your point?
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades. Absolutely.
That person should be using an Evergreen release, at this time. - -- James Mason Technical Architect, Public Cloud openSUSE Member SUSE jmason@suse.com - ------------------------------------------------------------------------ SUSECon 2015: Register at susecon.com -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBCAAGBQJVQOkhAAoJEBs5UYhsRJAj1eoIAIxiayUaWs5VgKDwcFCRDPK1 xp8yesBQRf32oW+cuQseg2czfTqh+8HYPkDhzhTAVV9qOte5LHdk3z1ZjkBL7sAG Fj3cYDVs2Vl/MDy7cPmkKMIJGLY94HGSlaATYueGlcmD7BdnPnDmJ9S4Bc+sqntu WnPGfLuMRIaj15KvrudM+uZ/IUSFmUS+tpRJ824KJjdeEmCFRkd2fC/1WGSf38N7 PE7iSDH1IoXjt8HCc3ac1UXlsY1/GwdAEk8IrzkWBHlad0kDpI2XDKKBp4Y0XJ+a XbHrEIlIDy7vEawndLPFKMbzz9dr31hako3VRo2RJ07hrJN7ERmjGNirEZKnr/M= =3zC6 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-29 16:22, James Mason wrote:
On 04/29/2015 07:17 AM, Carlos E. R. wrote:
I don't quite understand yours. To me, it makes pretty good sense that someone who doesn't upgrade to every distribution release and decides to skip one is even less likely to be happy with rolling upgrades. Absolutely.
That person should be using an Evergreen release, at this time.
Not necessarily. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVA6doACgkQja8UbcUWM1yiJAD+KqsO9eEI5EIP9v8Y7LRles5a gZb7JN+6X88ZQJBISkwA/R4Ybgkd6iRd9ZTTW6UCvYxpIESVgoW+dCksXnBv3NxB =MLFC -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/29/2015 07:53 AM, Richard Brown wrote:
On 29 April 2015 at 13:52, Stanislav Baiduzhyi <baiduzhyi.devel@gmail.com> wrote:
Hi Richard,
Please make a recording of that presentation and put it online somewhere, for people who will not be able to attend.
-- Regards, Stas
We will :)
Thank you! I look forward to watching it :) - -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVQPWnAAoJEM66EOTZRH6+1uMP/RF/L0H68bwx+xKgpNxdPSma 1Evy/S+I0iQ2P8jf3d9kuMpCe6Kcjbv2AdzctU9tZhRsNRNUbVG/0MMN/Xw3hSx5 LPPTFgDgkwl52tOuvSJ0jvLCX+f/KLLHCWasa86XaNab2R77LSnXrHNnSsOkSYAr 3CKrXQWC/64+Ryc58Q6RxP/TgBYiPaaKw6gQASOAhSDKO9YQNtAqDMRsUjqoaZLI zICFciDPrsRuqxDFBnN3CwUJu23lK4uUXGccyV5iHkjE/+xI8LQuIxUNYTRouMHw vLe1Gg5ULOFjsu7KOCzffI6LlAyZ3AjAbOtRlrqh4rKD9v+cFpWwvPvXwwGh4jhi L+Bm5cP9MfxHufcHgNjHqQwUzdcar4gV1+MjV3FcBvWggOwCelC1b+reBOQC5py3 XA1NBFJZC2bpqljmh8xUl/rm8ew1b7Q7Xue1mNVIngI16hr5d9EFi8qrQzo8Wdw4 e43RmbTe84g50qrx3HY7yJaDdgWS442Szaacr9zCbd+sthnMkXOrVzwA3u9hDdLS PvYpqpk7gFts/j71lunO7NXL4gh45pJ4xP7v5Viml8iV+R7diiW5AYJSxuTDg7Qm tp6J2LVjpyubznbY8TIgkllclxKVEYIKmRVY1UWN7FIRA3mYNf6vS4h0kEn8MmYL JWOHvmETdYeymqa1HAEn =cfiW -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Onsdag den 29. april 2015 02:14:06 skrev Carlos E. R.:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
But the co-existence of stable releases and TW, will/would mean that openSUSE as a whole would be for everybody, just about. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Tue, Apr 28, 2015 at 9:14 PM, Carlos E. R. <carlos.e.r@opensuse.org> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases.
For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me. That's why I'm not using Tumbleweed on any of my machines and I do not intend to. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases.
For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
That's why I'm not using Tumbleweed on any of my machines and I do not intend to.
Yep, that is the issue with a rolling release. +1 -- Per Jessen, Zürich (10.7°C) http://www.dns24.ch/ - free dynamic DNS, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases. For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
That's why I'm not using Tumbleweed on any of my machines and I do not intend to. Yep, that is the issue with a rolling release. +1
I think so, too. Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, Apr 30, 2015 at 10:18:14AM +0200, Torsten Gruner wrote:
Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases. For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
That's why I'm not using Tumbleweed on any of my machines and I do not intend to. Yep, that is the issue with a rolling release. +1
I think so, too.
Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only.
What do you mean with affected packages? We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already. My Laptop is running Tumbleweed and I have so far not have had a single troublesome status. Ciao, marcus -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Marcus Meissner wrote:
On Thu, Apr 30, 2015 at 10:18:14AM +0200, Torsten Gruner wrote:
Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases. For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
That's why I'm not using Tumbleweed on any of my machines and I do not intend to. Yep, that is the issue with a rolling release. +1
I think so, too.
Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only.
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
Unless you have a really unusual laptop it is probably not very representative of the majority of use cases :-) At some point last year I had an issue with icmp redirects not working after an upgrade, it took three months to fix. -- Per Jessen, Zürich (12.2°C) http://www.dns24.ch/ - free dynamic DNS, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 29/04/15 19:41, Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases.
For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time. It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc. It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
That's why I'm not using Tumbleweed on any of my machines and I do not intend to.
Michal Kubeček
I considered Tumbleweed, but decided against it based on the advice here: https://en.opensuse.org/Portal:Tumbleweed I use the NVidia proprietary graphics drivers to control my twin screen display. When I read the following... "When you should keep staying at stable release At this point there is no guarantee to have all additional modules available in the stable release like for Vmware or Virtualbox. And while the Packman Tumbleweed Essential repository attempts to deliver them there is no guarantee they will always succeed due to the incompatibilities with the quickly advancing Linux Kernel. The problems with proprietary Graphics drivers are similar and there is no guarantee they will work tomorrow, even if they do today. If you don't know how to compile your own additional kernel modules and you don't wish to learn or keep a very close eye on what is being updated, please don't use Tumbleweed." and "Special Concerns Third Party Drivers If you have 3rd party kernel modules it is STRONGLY suggested that you not use Tumbleweed. Seriously, it's not worth the pain and extra work, unless you _really_ want to do it. And if you do do it, then again, you are on your own, sorry." ...I accept the advice and stick with the stable releases. And even that is not without the occasional problem. :-( Bob - -- Bob Williams System: Linux 3.16.7-7-desktop Distro: openSUSE 13.2 (x86_64) with KDE Development Platform: 4.14.3 Uptime: 06:00am up 7:55, 3 users, load average: 0.16, 0.05, 0.06 -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlVB898ACgkQ0Sr7eZJrmU5dkACfaRXeebTAAauQiC3PxEf4DgAN xLoAnjP80g6gOeMbPkw7LdG9eRmMomfQ =MHte -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 30 of April 2015 10:30:23 Marcus Meissner wrote:
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
The risk of regression is much higher on version upgrades than on usual updates. And as I wrote before, there are also intentional incompatibilities where the new version behaves differently, has some functions dropped, requires different config file syntax or interprets certain config values differently. That's why I consider the "rolling" word misleading: it suggests a mental image of a continuous process; but there are changes that can't be "rolled" continuously over and will always be discontinuous. With the traditional periodic releases, these things happen on distribution upgrades and I expect them to happen then. With Tumbleweed, I would have to expect they might potentially happen any time I run "zypper update". What Torsten suggests would be marking such risky upgrades (what is risky does vary by package, of course) and allowing users to postpone them until they are ready to take the risk and handle the potential fallout. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 30.04.2015 11:20, Bob Williams wrote: > I use the NVidia proprietary graphics drivers to control my twin
screen display. When I read the following...
Quick question about this: I have a Macbook with an intel and nvidia graphics card. I used the nouveau driver and nvidia card for some time. But my laptop got really hot. So recently I blacklisted the noveau driver. And only use intel card. But then I realized: multiple monitor doesn't work with it! So now I consider switching back to nvidia, is the proprietary driver better? Also I read this: https://en.opensuse.org/SDB:NVIDIA_Bumblebee but it seems that won't help me cooling down my laptop, since I will use multiple monitors all the time I guess all the time it will also use the nvidia card, is this correct? I understand bumblebee like this: If I need not much good graphics, it switched to lower card (intel), if I need good graphic, it enabled the high card (nvidia). But when I will need multi monitor, it will use nvidia all the time, right? PS: Sorry for hijaking thread. Just was a good opportunity to ask, and get someone who might now! -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, 30 Apr 2015 10:30:23 +0200 Marcus Meissner <meissner@suse.de> wrote:
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
So the current touch-pad configuration problem hasn't affected your laptop? Lucky. -- Graham Davis [Retired Fortran programmer - now a mere computer user] openSUSE Tumbleweed (64-bit); KDE 4.14.6; Kernel: 3.19.4; Processor: AMD Phenom II X2 550; Video: nVidia GeForce 210 (using nouveau driver); Sound: ATI SBx00 Azalia (Intel HDA) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 30.04.2015 um 11:27 schrieb Michal Kubecek:
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status. The risk of regression is much higher on version upgrades than on usual updates. And as I wrote before, there are also intentional incompatibilities where the new version behaves differently, has some functions dropped, requires different config file syntax or interprets certain config values differently. That's why I consider the "rolling" word misleading: it suggests a mental image of a continuous process; but
On Thursday 30 of April 2015 10:30:23 Marcus Meissner wrote: there are changes that can't be "rolled" continuously over and will always be discontinuous.
With the traditional periodic releases, these things happen on distribution upgrades and I expect them to happen then. With Tumbleweed, I would have to expect they might potentially happen any time I run "zypper update".
What Torsten suggests would be marking such risky upgrades (what is risky does vary by package, of course) and allowing users to postpone them until they are ready to take the risk and handle the potential fallout.
Michal Kubeček
Yes, those are exactly my thoughtsyou wrote. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Tue, Apr 28, 2015 at 9:14 PM, Carlos E. R. <carlos.e.r@opensuse.org> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez Coolo said he'd rather not do a release at all and just use Tumbleweed. TW is not suitable for everybody :-( Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases. My main problem with having a rolling release is my PC at work, It's important that its secure so I do security updates often, but its also important that it doesn't break when I’m in the middle of important
On 04/30/2015 03:24 AM, Cristian Rodríguez wrote: projects and up against deadlines. As a developer its also important that it has relatively up to date software on it because that's what I need to run some tools. The 8-12 month cycle is ideal for this scenario it means I can wait till a few weeks after the release for any niggles to go away, then sometime in the next couple of months when I have some downtime between projects I can do the upgrade normally starting in some sort of VM because I need to deal with weird old stuff like NIS check that everything will probably still work then upgrade my machine and if I loose half a day to a day fixing stuff and doing it there’s not really any big deal. On a rolling distro I can't get security updates without also getting new versions of packages that may break stuff when I can't set aside time to fix it. I also run 13.2 on my laptop at home because for the select few pieces of software I like the latest version of I just add a development repo, when I used to run the old tumbleweed a year or so back I got sick of constantly having large downloads of stuff like libre office and also kernel updates were painful when running a older NVIDIA card 6150m that needs a legacy version of the proprietary drivers. So for me the end of releases would probably result in me being a lot less involved in openSUSE just because its not practical to run rolling releases on my systems and I don't want the hassle. Cheers Simon Lees -- Enlightenment Maintainer openSUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/30/2015 05:27 AM, blue hut wrote:
On 30.04.2015 11:20, Bob Williams wrote: > I use the NVidia proprietary graphics drivers to control my twin
screen display. When I read the following...
Quick question about this: I have a Macbook with an intel and nvidia graphics card. I used the nouveau driver and nvidia card for some time. But my laptop got really hot. So recently I blacklisted the noveau driver. And only use intel card.
But then I realized: multiple monitor doesn't work with it!
Cant say for the macbook but nouveau definitely supports multiple monitors, you have to configure it using the DE tools (eg. System Settings in KDE). - -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVQkMjAAoJEM66EOTZRH6+yLYP/iQODR+sEh6i5abkPPP9pWze t2mYp6thXxhf58IsXPSgQNm45XPyaI4P9agOfWNuydKG3poImRLiph1WiGAaxHKj jf2A1h48n3ltVroOd1TarwJeteBrL+2oerNsACD697MkLLdYmR8/XPNO7dAOyMDb BX9tWyGJHIGJ/EleOkj6l82tsnvkrj5Xwchr1Zj1zYbyhIkfk6PwuVBleNRadWTz VCqCe6HXAOTWtoG6gbVajr3A1Rgofyh7mZrVhB7yRPtU6v+GoUyGB6XOSYTCAm1L L3IfIDvIIlsdwCy8pFUuzwU6VQybnamhfJZcZFHcb3C8Qbno48RPIgpQTBWU49Xk vBlVkViX+DBkDg+0lYuR3oxh1CiXG70O938SSBQ7rDtpbuZPIbcXTZHIyDR+3YmI QWAyNp6uPlp5SnmMzyZOPYMQf+qZTPeYdlc6R8oGc7NiJDbdJgm0oqVgskMmFAe/ Nkw+PNObs807olfBxaNmzMAaHY9A1VDWC8oF1+GhvERFC+8SQQ4odG3rMkaraUsx 20yHHiWrb7kWdYGqJFNdXFWmUgAEdjDXFa4SX/kWSim1hoGIPav7y4ll2Jrv9lub 9ikGr+0qDoj5QogeMwVDc3U/CLQ4VJ6Fgbd1ifqWfR4y/NscEbF57p8aIOT21OXy zzFDnIeNpzb5KyLcGWgz =JsAL -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 29 April 2015 19.51:05 Martin Schlander wrote:
Onsdag den 29. april 2015 02:14:06 skrev Carlos E. R.:
On 2015-04-28 21:59, Greg Freemyer wrote:
On Tue, Apr 28, 2015 at 2:41 PM, Cristian Rodríguez
Coolo said he'd rather not do a release at all and just use Tumbleweed.
TW is not suitable for everybody :-(
But the co-existence of stable releases and TW, will/would mean that openSUSE as a whole would be for everybody, just about.
Hi Martin, No no no everybody that's a forbidden target, don't say for everybody, it's bad they told me :-) Still I share this vision. Still the unanswered question, by who? -- Bruno Friedmann Ioda-Net Sàrl www.ioda-net.ch openSUSE Member & Board, fsfe fellowship GPG KEY : D5C9B751C4653227 irc: tigerfoot -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/30/2015 10:58 AM, Uzair Shamim wrote:
On 04/30/2015 05:27 AM, blue hut wrote:
On 30.04.2015 11:20, Bob Williams wrote: > I use the NVidia proprietary graphics drivers to control my twin
screen display. When I read the following...
Quick question about this: I have a Macbook with an intel and nvidia graphics card. I used the nouveau driver and nvidia card for some time. But my laptop got really hot. So recently I blacklisted the noveau driver. And only use intel card.
But then I realized: multiple monitor doesn't work with it!
Cant say for the macbook but nouveau definitely supports multiple monitors, you have to configure it using the DE tools (eg. System Settings in KDE).
And... I should learn to read... :S - -- Regards, Uzair Shamim -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVQmMxAAoJEM66EOTZRH6+cMMQAI0nYRdAHRzROk7zMNrsV0Xh iXSrhPoubbrC2Xa9ZTexaOpIsR22AyGt564eMETI7vB0tEHVZjTWDIgPKAbNLw8c AxS6DXe8o2yuz1JhSQ4GAXv3cSEkXJEIw6zXEkw6G3z+C8cqpyWGM99Ok7GpFH94 S26HIlnj8FcF5VmyCeMWPYJfthSNUcSmc0mDixhSb2knojkZvNEDY/YGPOxegQjx gG9saLR6ezlpPU5OzxSlKp9mBpRvOx2CgBzZtqHDC3fe4hvFwuiTSQOv+wl+CIUV mDFka/C62j2JTr+f1uYTq6TOBWqNmD731+sLaSJ6N+Oi7u3CqOrHskwy7b+hJRrs 23QRSc1cnj3C88yL1W3q6e35KV5ywHft5nLHUWURYbKIbdfnZNjlNrwShKo9MAIn JuWo/fjBIDnHPLo1l1mPUFYQ18m0WB0Nk/ElFITn3wW2RmqywW3B8Ks5yMZCXngH Xv7t5pr2586gpleI4F2VYq/wER2Tov63PBUpkucZu0Xhdcf48IfvYae1f6Om2WLb HFpJYpanMwZwwecPIYfWJo1ykXRFZEkgEpBcKKC/I2GNXCfnQtC7sSRm8MRgkNO1 61+M8uT4q2qa72yOUXGiK+T9MV4byY1ryymugH1tKosR+08dMNZZrqH91NkaoP79 3WHd67uK4i5BfcOQMKn5 =CqBG -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 30 April 2015 11:27:28 blue hut wrote:
On 30.04.2015 11:20, Bob Williams wrote: > I use the NVidia proprietary graphics drivers to control my twin
screen display. When I read the following...
Quick question about this: I have a Macbook with an intel and nvidia graphics card. I used the nouveau driver and nvidia card for some time. But my laptop got really hot. So recently I blacklisted the noveau driver. And only use intel card.
But then I realized: multiple monitor doesn't work with it!
That is the strange part for me, I know for sure that intel driver supports up to 3 monitors (laptop + 2 external using docking station). Maybe you should raise that as the first problem, and solving this issue will mean that you do not need nvidia any more? I have bumblebee on my older asus zenbook, but I'm using it only to keep nvidia disabled at all times (bbswitch). If I understand it correctly, linux support for optimus is weaker than on windows (thanx, nvidia!), so you actually have to launch stuff using bumblebee as wrapper, mostly intended for games. I may be wrong though, never used it myself, only heard others complaining. -- Regards, Stas -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 30.04.2015 um 10:30 schrieb Marcus Meissner:
On Thu, Apr 30, 2015 at 10:18:14AM +0200, Torsten Gruner wrote:
I think so, too.
Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only.
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore. My confidence in TW as a working horse is really quite low while I had actually some hopes in it. Wolfgang -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi, On Thu, Apr 30, 2015 at 10:30:23AM +0200, Marcus Meissner wrote:
On Thu, Apr 30, 2015 at 10:18:14AM +0200, Torsten Gruner wrote: [ 8< ]
Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only.
What do you mean with affected packages?
The kernel for example. At some point we had kernel-default-3.19.4-1.1.x86_64 and that got replaced by kernel-default-3.19.4-1.2.x86_64 Both make use of /boot/vmlinuz-3.19.4-1-default resulting in a file conflict.
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
On one of my real systems this worked quite well - cause I power it on less often. But with my daily updated workstation I had trouble several times. I still have two rdsosreport.txt files stored. As it works 3 of 4 times and I don't see a pattern - sometimes booting an older kernel and calling dracut for the failing one help but sometimes I had to make use of a rescue system - and have not reported it to bugzilla yet.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
In the current state Tumbleweed isn't able to pass the relatives test yet. Even getting them using openSUSE 13.2 with a KDM which sets a non KDE session type with one of the recent KDE 13.2 updates is a no go for such users. All they intend more or less is to start a web browser and a text processor. And all this is for me a good reason to have regular openSUSE releases. Once every 12 months if fine. But hey, maybe the openSUSE conference will even bring a much better news to us. :) Cheers, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team + SUSE Labs SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all. As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers. Greetings, Stephan -- Ma muaß weiterkämpfen, kämpfen bis zum Umfalln, a wenn die ganze Welt an Arsch offen hat, oder grad deswegn. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 01.05.2015 um 13:14 schrieb Stephan Kulow:
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all.
As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers.
Xorg is immediately dying with a segfault with current TW version using vboxvideo for me. And yes, it might very well be an issue in the VBox version but the end result is: unusable (unfortunately) Wolfgang -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 04/30/2015 06:48 PM, Wolfgang Rosenauer wrote:
Am 30.04.2015 um 10:30 schrieb Marcus Meissner:
On Thu, Apr 30, 2015 at 10:18:14AM +0200, Torsten Gruner wrote:
I think so, too.
Another solution might be to mark the affected packageswith this kind of changes. And each packet has to be assessed whether a jump to the next version without consequence for the function. Can several versions are skipped? And zypper has to handle this. This is a nice feature but a dream only.
What do you mean with affected packages?
We try to push out only "stable" and working tumbleweed snapshots and that seems to be working quite well already.
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore. My confidence in TW as a working horse is really quite low while I had actually some hopes in it.
Well, but TW is not really intended to run on a "production" server. TW is targeted at developers that want the latest of X and do not mind getting the latest of Y in the process. On a personal note I run TW on my laptop as it gets frequently rebooted anyway, but do not rnu TW on my desktop as I hate rebboting that machine. Simply it takes me 1/2 hour to get all my workspaces and terminals and and and back to the state before the reboot. Later, Robert -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/01/2015 07:31 AM, Wolfgang Rosenauer wrote:
Am 01.05.2015 um 13:14 schrieb Stephan Kulow:
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all.
As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers.
Xorg is immediately dying with a segfault with current TW version using vboxvideo for me. And yes, it might very well be an issue in the VBox version but the end result is: unusable (unfortunately)
Wolfgang
I have installed Tumbleweed on VmWare and X works. VirtualBox will be releasing V5.0 Beta 3 in a few days. I'm downloading Beta 2 to see if X works. Cheers! Roman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/01/2015 03:10 PM, Roman Bysh wrote:
On 05/01/2015 07:31 AM, Wolfgang Rosenauer wrote:
Am 01.05.2015 um 13:14 schrieb Stephan Kulow:
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all.
As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers.
Xorg is immediately dying with a segfault with current TW version using vboxvideo for me. And yes, it might very well be an issue in the VBox version but the end result is: unusable (unfortunately)
Wolfgang
I have installed Tumbleweed on VmWare and X works. VirtualBox will be releasing V5.0 Beta 3 in a few days.
I'm downloading Beta 2 to see if X works.
Cheers!
Roman
Follow Up Just tried Tumbleweed snapshot 20150430 and the problem still exists with X. Yast ncurses starts up. No gui. -- Cheers! Roman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Roman Bysh <rbtc1@rogers.com> [01-01-70 12:34]:
On 05/01/2015 03:10 PM, Roman Bysh wrote:
I have installed Tumbleweed on VmWare and X works. VirtualBox will be releasing V5.0 Beta 3 in a few days.
I'm downloading Beta 2 to see if X works.
Cheers!
Roman
Follow Up
Just tried Tumbleweed snapshot 20150430 and the problem still exists with X. Yast ncurses starts up. No gui.
Just so you know it does work for some of us: Tumbleweed 20150430 NVIDIA-Linux-x86_64-349.12 GeForce GTS 450 http://wahoo.no-ip.org/~pat/TWsnapshot1.jpg for the record, I have been running tw since Greg KH started it and have always been able to run yas2 (gui) althought I have had short periods of inability to use the nvidia prop drivers which usually were quickly solved with the help of Larry Finger, thanks for the support. I have also but not currently using vb, been able to run versions of openSUSE and other distros in a vb but only for experience, not productive use (my choice). -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am Freitag, 1. Mai 2015, 13:31:07 schrieb Wolfgang Rosenauer:
Am 01.05.2015 um 13:14 schrieb Stephan Kulow:
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all.
As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers.
Xorg is immediately dying with a segfault with current TW version using vboxvideo for me. And yes, it might very well be an issue in the VBox version but the end result is: unusable (unfortunately)
I'm running tumbleweed on virtualbox for the same reason for a while. X runs quite accurate here. Greetings Willi -- openSUSE 20150430 "Tumbleweed" (x86_64) GNU/Linux 4.0.0-1-desktop on VirtualBox 4.3.26 on host OpenSUSE 13.1 "Bottle" -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/01/2015 06:37 PM, Wilhelm Boltz wrote:
Am Freitag, 1. Mai 2015, 13:31:07 schrieb Wolfgang Rosenauer:
Am 01.05.2015 um 13:14 schrieb Stephan Kulow:
Am 01.05.2015 um 00:48 schrieb Wolfgang Rosenauer:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore.
I wonder how much of that is related to Virtualbox though. I already added virtualbox to the staging projects so we make sure it always builds, but we don't test it at all.
As my google mojo fails me to find a way to run VB inside qemu VMs, an option is to revive the VB backend of openQA - or run VB tests on IMPI servers.
Xorg is immediately dying with a segfault with current TW version using vboxvideo for me. And yes, it might very well be an issue in the VBox version but the end result is: unusable (unfortunately)
I'm running tumbleweed on virtualbox for the same reason for a while. X runs quite accurate here.
Greetings Willi
Perhaps you misunderstood. In VB, during the initial installation the Yast installer does not stay with the familiar gui. It switches to Yast ncurses. It looks like X server segfaults as Yast is about to start. I can see the Yast notifying me that it is starting up. Under normal circumstances I should be seeing the regular gui as in VmWare. Once the installation finishes and restarts X is working. My card is a GeForce 8600 GTS. When installing on a real partition everything is normal just like in openSUSE 13.2. I remember when the earlier isos were released. You know the ones that used to have the factory repos. Some time between March and April, X started behaving strangely during the initial set up, selection of packages and Grub2 of Tumbleweed in VirtualBox. I believe they are already looking into this. In the meantime, I'll stick with installing Tumbleweed on to real hardware. Cheers! Roman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-30 10:30, Marcus Meissner wrote:
My Laptop is running Tumbleweed and I have so far not have had a single troublesome status.
You should subscribe to the Tumbleweed forum ;-) - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVEKxoACgkQja8UbcUWM1xEpAD/SU+cHWoMFmqP/FzCxYbBXgDX LGkhty1rb0s4fmfVx20A/3NA/pEI3uogi0xb66LxANmAZyW3D0ZJJpbhoY64VR0q =Nx7n -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-30 11:27, Michal Kubecek wrote:
What Torsten suggests would be marking such risky upgrades (what is risky does vary by package, of course) and allowing users to postpone them until they are ready to take the risk and handle the potential fallout.
Would not work. Other packages would expect the new features, too, and would in turn fail in unexpected ways. Too complex to handle. Tumbleweed most be at consistent status with itself. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVEK88ACgkQja8UbcUWM1xKvgD/eRPis5gAIt4oVibeKz57u0pH T5Fdf3SJiCvYqj5/ObcA/256wOu6QchTYTvoLil80vIas51MyrSNKJGQW/dxDFT8 =7XU0 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-30 11:20, Bob Williams wrote:
I considered Tumbleweed, but decided against it based on the advice here: https://en.opensuse.org/Portal:Tumbleweed
...
...I accept the advice and stick with the stable releases. And even that is not without the occasional problem. :-(
Absolutely. It is documented! LOL. :-) - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVELnAACgkQja8UbcUWM1xNwQD/SK1envHsqme2lB6pkC5DfDtG kpSeNVpknjsqwceRP94A/RsrkQ8EAfGP16gtEH7P9oSedwggfPKaoOc+0DGiev1j =r9Hj -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-30 15:56, Simon Lees wrote:
My main problem with having a rolling release is my PC at work, It's important that its secure so I do security updates often, but its also important that it doesn't break when I’m in the middle of important projects and up against deadlines.
Yep.
As a developer its also important that it has relatively up to date software on it because that's what I need to run some tools.
Aha. True.
The 8-12 month cycle is ideal for this scenario it means I can wait till a few weeks after the release for any niggles to go away, then sometime in the next couple of months when I have some downtime between projects I can do the upgrade normally starting in some sort of VM because I need to deal with weird old stuff like NIS check that everything will probably still work then upgrade my machine and if I loose half a day to a day fixing stuff and doing it there’s not really any big deal. On a rolling distro I can't get security updates without also getting new versions of packages that may break stuff when I can't set aside time to fix it.
Same here.
I also run 13.2 on my laptop at home because for the select few pieces of software I like the latest version of I just add a development repo, when I used to run the old tumbleweed a year or so back I got sick of constantly having large downloads of stuff like libre office and also kernel updates were painful when running a older NVIDIA card 6150m that needs a legacy version of the proprietary drivers.
Yep.
So for me the end of releases would probably result in me being a lot less involved in openSUSE just because its not practical to run rolling releases on my systems and I don't want the hassle.
Yes, too... - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVEL1EACgkQja8UbcUWM1wA1gD/UbVzB+wLFZZ9Cu2gHR0g3/LH azXrpihUf+e3Gp7/abAA/jJiJnbYxoRxGoQxOmwFv3hpkjSsNIBO9zybbI950ZMS =UR/k -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-04-29 20:41, Michal Kubecek wrote:
On Wed, Apr 29, 2015 at 02:54:30PM -0300, Cristian Rodríguez wrote:
Neither is every single software product out there.. ;-) Please make your case on what's the exact problem with having only rolling releases.
For me, the main reason is the fact that upgrades necessarily cause various kinds of problems from time to time.
Yes.
It's not only bugs and regressions but also intentional changes in program behaviour, config file syntax or semantics etc.
Yes. Or having to replace a package with another one, very different. ifup vs wicked, Initrd vs iforgotwhat....
It's acceptable for me to do a distribution upgrade once a year or two (less often on a server) when I reserve enough time to resolve these issues and I expect things to be broken temporarily. Having such issues continuously hitting me at random times in a "rolling upgrades" model is not acceptable for me.
Yes, absolutely!
That's why I'm not using Tumbleweed on any of my machines and I do not intend to.
I agree completely. Tumbleweed is great. Just not for me. If I have time, I might install it on a spare machine, for testing; but I don't have it. I may install it under vmware, though, for testing things. But then some devs don't like my reports. There are more issues with Tumbleweed. For instance, updates are bigger (no patches), and they are more, so you need a better Internet connection; and not everybody has it. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVEL3MACgkQja8UbcUWM1z6wAEAnZkVOTZs7RJrUDUQ5VHISxJx CJ43gDgRNAAwDamZxkcA/RB8ERy3sDdvN+78o/vlEvks3Bhp2gFFxR/IX3KBXYQR =lmFJ -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Carlos E. R. <carlos.e.r@opensuse.org> [05-01-15 22:30]: [...]
Or having to replace a package with another one, very different. ifup vs wicked, Initrd vs iforgotwhat....
Ah, rpm -qf `which ifup` wicked-service-0.6.18-1.1.x86_64 perhaps, NetworkMissManager as I have four remote boxes which fail 80% of the time, but not with wicked. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Patrick Shanahan composed on 2015-05-01 22:43 (UTC-0400):
perhaps, NetworkMissManager as I have four remote boxes which fail 80% of the time, but not with wicked.
I have one test box without NM and with Wicked, with default target multi-user, with nearly 200 noauto nfs mounts in fstab (same exact entries as all my PCs running Linux, cut and pasted from template), that if TW booted with 5 on cmdline produces: May 02 01:53:19 gx620 rpc.gssd[1307]: ERROR: gssd_refresh_krb5_machine_credential: no usable keytab entry found in keytab /etc/krb5.keytab for connection with host <FQHN> May 02 01:53:19 gx620 rpc.gssd[1307]: ERROR: No credentials found for connection to server <FQHN> instead of any NFS mounts when I try mounting them, simply booting with 5, but not without. If booted normally without 5 or with 3, those mounts mount normally. My LAN has no more than one NIC per PC, all on fixed IP, and none have wireless. A difference between whether NFS mounting works or not according to whether X is running seems inexplicable, and never seen here before there was Wicked. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Felix Miata <mrmazda@earthlink.net> [05-02-15 02:39]:
Patrick Shanahan composed on 2015-05-01 22:43 (UTC-0400):
perhaps, NetworkMissManager as I have four remote boxes which fail 80% of the time, but not with wicked.
I have one test box without NM and with Wicked, with default target multi-user, with nearly 200 noauto nfs mounts in fstab (same exact entries as all my PCs running Linux, cut and pasted from template), that if TW booted with 5 on cmdline produces:
May 02 01:53:19 gx620 rpc.gssd[1307]: ERROR: gssd_refresh_krb5_machine_credential: no usable keytab entry found in keytab /etc/krb5.keytab for connection with host <FQHN> May 02 01:53:19 gx620 rpc.gssd[1307]: ERROR: No credentials found for connection to server <FQHN>
instead of any NFS mounts when I try mounting them, simply booting with 5, but not without. If booted normally without 5 or with 3, those mounts mount normally. My LAN has no more than one NIC per PC, all on fixed IP, and none have wireless. A difference between whether NFS mounting works or not according to whether X is running seems inexplicable, and never seen here before there was Wicked.
And using NM booting to runlevel 3/multi-user has *always* failed to provide network communication on any of my machines, wireless or wired, besides only providing networking "when it feels like it" when booting to graphical mode. But I must say that on one machine wicked may also fail "when it feels like it" when booting to either multi-user or graphical. <soap-box> Currently the suit linux was always strong on, networking, is no longer so. </soap-box> -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Saturday 2015-05-02 13:53, Patrick Shanahan wrote:
* Felix Miata <mrmazda@earthlink.net> [05-02-15 02:39]:
Patrick Shanahan composed on 2015-05-01 22:43 (UTC-0400):
perhaps, NetworkMissManager as I have four remote boxes which fail 80% of the time, but not with wicked.
I have one test box without NM and with Wicked, with default target multi-user, with nearly 200 noauto nfs mounts in fstab (same exact entries as all my PCs running Linux, cut and pasted from template), that if TW booted with 5 on cmdline produces:
instead of any NFS mounts when I try mounting them, simply booting with 5, but not without. If booted normally without 5 or with 3, those mounts mount normally. My LAN has no more than one NIC per PC, all on fixed IP, and none have wireless. A difference between whether NFS mounting works or not according to whether X is running seems inexplicable, and never seen here before there was Wicked.
And using NM booting to runlevel 3/multi-user has *always* failed to provide network communication on any of my machines, wireless or wired, [...] <soap-box> Currently the suit linux was always strong on, networking, is no longer so. </soap-box>
Since networkd is in TW now, new possibilities open up :) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Jan Engelhardt <jengelh@inai.de> [05-02-15 10:29]:
On Saturday 2015-05-02 13:53, Patrick Shanahan wrote:
* Felix Miata <mrmazda@earthlink.net> [05-02-15 02:39]:
Patrick Shanahan composed on 2015-05-01 22:43 (UTC-0400):
perhaps, NetworkMissManager as I have four remote boxes which fail 80% of the time, but not with wicked.
I have one test box without NM and with Wicked, with default target multi-user, with nearly 200 noauto nfs mounts in fstab (same exact entries as all my PCs running Linux, cut and pasted from template), that if TW booted with 5 on cmdline produces:
instead of any NFS mounts when I try mounting them, simply booting with 5, but not without. If booted normally without 5 or with 3, those mounts mount normally. My LAN has no more than one NIC per PC, all on fixed IP, and none have wireless. A difference between whether NFS mounting works or not according to whether X is running seems inexplicable, and never seen here before there was Wicked.
And using NM booting to runlevel 3/multi-user has *always* failed to provide network communication on any of my machines, wireless or wired, [...] <soap-box> Currently the suit linux was always strong on, networking, is no longer so. </soap-box>
Since networkd is in TW now, new possibilities open up :)
as in /usr/lib/systemd/systemd-networkd, systemd-219-5.1.x86_64, as I fail to find the package "networkd". What does it accomplish re wicked/NetworkManager? tks, -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi, Am 01.05.2015 um 16:51 schrieb Robert Schweikert:
On 04/30/2015 06:48 PM, Wolfgang Rosenauer wrote:
Since I'm afraid of putting TW on a production machine I decided to try it in VirtualBox for quite some time (I reported issues earlier) and I can only say that still every other TW update will leave my VirtualBox TW installation unusable. Usually no X coming up anymore. My confidence in TW as a working horse is really quite low while I had actually some hopes in it.
Well, but TW is not really intended to run on a "production" server. TW is targeted at developers that want the latest of X and do not mind getting the latest of Y in the process.
I wasn't talking about servers but TW as my every day desktop. It could very well be that my desktop wouldn't be affected since I would run natively there but there is not even a chance to check that out if it does not run under my preferred virtualization.
On a personal note I run TW on my laptop as it gets frequently rebooted anyway, but do not rnu TW on my desktop as I hate rebboting that machine. Simply it takes me 1/2 hour to get all my workspaces and terminals and and and back to the state before the reboot.
So now this is an unexpected statement since many many openSUSE community members claim that all you need is TW and nothing else and this was also what this part of the thread was about. Good to hear that this is apparently just an opinion of some very vocal part of the community. Wolfgang -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 03.05.2015 um 09:09 schrieb Wolfgang Rosenauer:
So now this is an unexpected statement since many many openSUSE community members claim that all you need is TW and nothing else and this was also what this part of the thread was about.
I for one never claimed that TW is the only thing needed. All I'm saying is that the general contributor is a lot less interested in contributing into a stable release. 13.2 beta had very few testers, that's just no base to continue with the way we did openSUSE releases. Greetings, Stephan -- Ma muaß weiterkämpfen, kämpfen bis zum Umfalln, a wenn die ganze Welt an Arsch offen hat, oder grad deswegn. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Sunday 03 May 2015 14:23:03 Stephan Kulow wrote:
[...] All I'm saying is that the general contributor is a lot less interested in contributing into a stable release. 13.2 beta had very few testers, that's just no base to continue with the way we did openSUSE releases.
I am not sure this has changed to a better with the new TW. I never did run Factory, I did test some of the betas when I was still a student and had time, but until recently I just used 13.2 with addon repos (XOrg, KDE, ...). With the new TW, this has changed. One of my computers is running it (maybe more in the future), but for me it would be fine to switch it to some Tumbleweed-on-Ice for a month or two and help stabilize it. I need something I can be sure it works for the other computers, and although TW is quite good (or I am lucky) is not there yet, and I doubt it ever will be. TW has it made possible to run a testing version of openSUSE regularly - that is, every day. Regards, Stefan -- Stefan Brüns / Bergstraße 21 / 52062 Aachen phone: +49 241 53809034 mobile: +49 151 50412019 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Dear Christoph, Em ter 28 abr 2015, às 16:09:57, Christoph Grüninger escreveu:
Dear openSuse,
is there a road-map for the next openSuse release 13.3? I could not find one at the usual location https://en.opensuse.org/openSUSE:Roadmap
According to the 8 months plan and the last release in November, I could expect early June for the next release. Obviously, this is not the plan. What's the plan instead? If you make changes, please tell us users.
Bye Christoph
Perhaps you want to see the presentation of Richard Brown "The Future is Unwritten" in OSC 2015 that addresses this specific term. He proposed a very interesting "merge" of openSUSE and SLE, since the source has been made available recently. Here is the link in YouTube: https://www.youtube.com/watch?v=BH99TSrfvq0 Best Regards, Ronan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Sat, 2 May 2015 07:53:14 -0400 Patrick Shanahan <ptilopteri@gmail.com> wrote:
But I must say that on one machine wicked may also fail "when it feels like it" when booting to either multi-user or graphical.
Please collect logs as per: https://en.opensuse.org/openSUSE:Bugreport_wicked and open a bug report. Then we can fix it (if anything on wicked's side) to behave as expected. But first of all please upgrade to wicked 0.6.18. There were lots of fixes and improvements delivered. Should be soon available for 13.2. Currently available at OBS's network:wicked:maste.r and network:wicked:factory (AFAIR it has been also accepted to TW). -- Best Regards, Pawel Wieczorkiewicz <pwieczorkiewicz@suse.de>, Linux System Developer SUSE LINUX GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 21284 (AG Nürnberg) Maxfeldstraße 5 / 90409 Nürnberg / Germany / Phone: +49-911-740 53 - 613 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Pawel Wieczorkiewicz <pwieczorkiewicz@suse.de> [05-05-15 03:43]:
On Sat, 2 May 2015 07:53:14 -0400 Patrick Shanahan <ptilopteri@gmail.com> wrote:
But I must say that on one machine wicked may also fail "when it feels like it" when booting to either multi-user or graphical.
Please collect logs as per: https://en.opensuse.org/openSUSE:Bugreport_wicked and open a bug report. Then we can fix it (if anything on wicked's side) to behave as expected. But first of all please upgrade to wicked 0.6.18. There were lots of fixes and improvements delivered.
Should be soon available for 13.2. Currently available at OBS's network:wicked:maste.r and network:wicked:factory (AFAIR it has been also accepted to TW).
Running Tw and at 20150430 which has: wicked-service-0.6.18-1.1.x86_64 Fri May 1 09:27:10 2015 wicked-0.6.18-1.1.x86_64 Fri May 1 09:27:08 2015 Have not rebooted sub box since then. Will observe and if wicked shows another failure, I will file a bug report against wicked. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://linuxcounter.net -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Tue, 5 May 2015 07:53:58 -0400 Patrick Shanahan <ptilopteri@gmail.com> wrote:
Running Tw and at 20150430 which has: wicked-service-0.6.18-1.1.x86_64 Fri May 1 09:27:10 2015 wicked-0.6.18-1.1.x86_64 Fri May 1 09:27:08 2015
Have not rebooted sub box since then. Will observe and if wicked shows another failure, I will file a bug report against wicked.
Perfect! Thank you. You may also consider switching to nanny (providing nanny=1 to linuxrc upon installation or changing to <use-nanny>true</use-nanny> in /etc/wicked/common.xml). This usually helps with hotplug stuff. BTW this is going to be the default from wicked 0.6.19 onwards. Another frequent source of problems is default requirement to wait for device's link to be present. This can be avoided by adding extra LINK_REQUIRED=no into the ifcfg file of a given interface (more can be found in man ifcfg(5)). Not sure what is the problem in your case, but we shall see with the logs. -- Best Regards, Pawel Wieczorkiewicz <pwieczorkiewicz@suse.de>, Linux System Developer SUSE LINUX GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 21284 (AG Nürnberg) Maxfeldstraße 5 / 90409 Nürnberg / Germany / Phone: +49-911-740 53 - 613 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Søndag den 3. maj 2015 14:23:03 skrev Stephan Kulow:
Am 03.05.2015 um 09:09 schrieb Wolfgang Rosenauer:
So now this is an unexpected statement since many many openSUSE community members claim that all you need is TW and nothing else and this was also what this part of the thread was about.
I for one never claimed that TW is the only thing needed. All I'm saying is that the general contributor is a lot less interested in contributing into a stable release. 13.2 beta had very few testers, that's just no base to continue with the way we did openSUSE releases.
It was also the first release based on the "new" factory/tumbleweed, with a very short testing phase (two devel releases with 2-3 weeks between) and iirc the roadmap itself was published very late. Secondly, openSUSE releases shouldn't need as many testers/much testing as before, with Tumbleweed already being fairly well tested by a decent number of people running it full-time. At least I thought that was the idea. The whole process of Tumbleweed based stable releases just needs a few small adjustments, to work well imo. (A roadmap, a slightly longer window to test things, a fixed 12 month release cycle, 26 months of lifetime (2 releases+2 months) would be a very good start). -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 5 May 2015 at 21:40, Martin Schlander <martin.schlander@gmail.com> wrote:
It was also the first release based on the "new" factory/tumbleweed, with a very short testing phase (two devel releases with 2-3 weeks between) and iirc the roadmap itself was published very late.
Secondly, openSUSE releases shouldn't need as many testers/much testing as before, with Tumbleweed already being fairly well tested by a decent number of people running it full-time. At least I thought that was the idea.
The whole process of Tumbleweed based stable releases just needs a few small adjustments, to work well imo. (A roadmap, a slightly longer window to test things, a fixed 12 month release cycle, 26 months of lifetime (2 releases+2 months) would be a very good start).
From an Engineering perspective it was 'just' a Tumbleweed snapshot, but from a Marketing, Branding, 'actually getting a release out of the door' perspective, it required as much work as any other software release. For 13.2 this fell on the shoulders of a tiny handful of people. The reality is that Tumbleweed has captured the excitement of a huge
I understand where you're coming from, but I think there are a few things I'd like you to consider The 13.2 release required a large amount of work from a small number of people to make it a success. portion of our current contributing community, and grown it greatly, and that's great, but it does have the side effect of leaving less enthusiasm and less people to work on the Regular Release. As was discussed at the Project Meeting at oSC 15, the 'Tumbleweed-Snapshot-based-Release' model also has an interesting side effect when it comes to this part of getting a release published and marketed. Just as they share development models, both Tumbleweed and snapshot-based-releases have very similar core messages - "The latest of everything, each release". The major differences are the gap between releases and the method of maintenance after release. On each Release Day for the Regular Release, this ends up leaving us with relatively little exciting to talk about (Everything is already in Tumbleweed) and at the same time, way too much to talk about (the Changelog of 13.1 to 13.2 is insanely long and hard to filter down into exciting things to include in Release announcements) If snapshot-based-releases were to continue, we'd need to find a way to tackle that, and get people excited, talking about, and helping put together the release and it's marketing, something which is basically just the same as what we do every week, but maintained differently. But this is where my proposal for a Regular Release based on the SLE Sources comes into play. ( Video available at https://www.youtube.com/watch?v=BH99TSrfvq0 ) Our Regular Release has always been pulled in two directions, with some wanting it faster and more innovative, some wanting it slower and more stable. Tumbleweed has addressed the 'go faster' crowd - I think we have the *best* Rolling Distribution out there, and it's getting better Now we have the sources for SUSE Linux Enterprise we have an opportunity to build the *best* Stable Distribution out there. Something that really addresses the needs of those users who want a stable Linux distribution for 'real work', with Enterprise stability at it's core, with the community turning it into something exceptional. The best of both worlds, which is really something different from what everyone else has (Ubuntu LTS, CentOS) - and it's steps in that direction which will help us get new people involved to help work on it and make it even better going forward. I realise this proposal would mean the change of expectations for each openSUSE regular release - maybe we wouldn't upgrade every version of [insert software here] each release. I realise this is a huge change from the promise of 'The latest of everything, each release'. The individual teams of our project will have the opportunity to choose what goes into the new Regular Release and how often they want to upgrade it. With the clear division in direction between Tumbleweed and this new Regular Release, hopefully decisions about which versions make sense for inclusion when become easier for those teams. They'll know the kind of expectations our users have for either distribution, and be able to make informed judgement calls as to when to include whatever they want from the various upstream projects they maintain in openSUSE. In the case of the software provided in the shared SLE Sources, I also think we also have a lot of collective learning to do, as I think many of the fears regarding the 'age' of that software and their capabilities don't hold up under scrutiny (eg. the SLE Kernel gets a TON of backported features and patches) So while, of course snapshot-based-releases is a valid option, I'd rather see what we can make of this opportunity to build something better for openSUSE Project and our users. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-05-05 23:11, Richard Brown wrote:
For 13.2 this fell on the shoulders of a tiny handful of people. The reality is that Tumbleweed has captured the excitement of a huge portion of our current contributing community, and grown it greatly, and that's great, but it does have the side effect of leaving less enthusiasm and less people to work on the Regular Release.
Yes.
On each Release Day for the Regular Release, this ends up leaving us with relatively little exciting to talk about (Everything is already in Tumbleweed) and at the same time, way too much to talk about (the Changelog of 13.1 to 13.2 is insanely long and hard to filter down into exciting things to include in Release announcements) If snapshot-based-releases were to continue, we'd need to find a way to tackle that, and get people excited, talking about, and helping put together the release and it's marketing, something which is basically just the same as what we do every week, but maintained differently.
As a stable release user, I would be very excited by every new feature that I know is available in Tumbleweed that makes into Stable. And I would be very disappointed to find out that they will not be and that I have to wait another year for them.
Our Regular Release has always been pulled in two directions, with some wanting it faster and more innovative, some wanting it slower and more stable.
Yes, but that we want it slower doesn't have the meaning of using versions that already a year old compared to tumbleweed at release time. What we want (I have my user hat on) is that when it is released we can keep using it for about two or three years if we want to. Slower means that we do not want to be *forced* to reinstall or upgrade about every eight months, but that we *can*, at a reasonable interval (which can be 8 or 12 months, whatever is feasible). And we do want the new shiny improvements when we do install. It is not the same meaning of slower. You also have to consider that we are a mix of server sysadmins and desktop users, both professional and home. Basing on SLES is wonderful for the sysadmin crowd, but not that much for the rest. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVJRwQACgkQja8UbcUWM1xglQD/TnMPp2oVjpW/y9SDVq42+rsg vhan4FlSqpAH+WuHkbEA/2guaoblffSc53N4rwj2UZ1JtzdIq1Z/XWTc07yEhfQ9 =4hI3 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 06 of May 2015 00:41:09 Carlos E. R. wrote:
As a stable release user, I would be very excited by every new feature that I know is available in Tumbleweed that makes into Stable. And I would be very disappointed to find out that they will not be and that I have to wait another year for them.
One thing we should keep in mind is that openSUSE kernel maintenance doesn't work very well. You could read previous discussion on this topic for details. The short story is that unless we are lucky to pick a longterm kernel version (and we haven't been for quite long), upstream stable branch is closed even before openSUSE reaches its release and after that, only security bugs and small amount of functional fixes is backported into it. We do not have resources to do openSUSE kernel maintenance comparable to at least upstream stable branches. Piggybacking on SLE kernel maintenance allows to achieve much better maintenance with much less effort than anything else. That's why Evergreen 11.4 uses 3.0 kernel (SLE11-SP2 based) and Evergreen 13.1 is planned to use 3.12 kernel (SLE12 based - I consider a move to SLE12-SP1 but that's for a different discussion in other lists). Even if openSUSE provides features and drivers not provided or not supported by SLE (including the whole i586 architecture now), the result is still much better that anything else we can realistically think of. Of course, we can still do what we have been doing so far: take the latest released kernel available at the moment and do the minimum maintenance. But I can't help thinking it would be a pity if we didn't consider the opportunity to utilize the work behind SLE kernel. Perhaps having both might satisfy both types of users. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Michal Kubecek <mkubecek@suse.cz> [2015-05-06 11:24]:
On Wednesday 06 of May 2015 00:41:09 Carlos E. R. wrote:
As a stable release user, I would be very excited by every new feature that I know is available in Tumbleweed that makes into Stable. And I would be very disappointed to find out that they will not be and that I have to wait another year for them.
One thing we should keep in mind is that openSUSE kernel maintenance doesn't work very well. You could read previous discussion on this topic for details. The short story is that unless we are lucky to pick a longterm kernel version (and we haven't been for quite long), upstream stable branch is closed even before openSUSE reaches its release and after that, only security bugs and small amount of functional fixes is backported into it.
We do not have resources to do openSUSE kernel maintenance comparable to at least upstream stable branches. Piggybacking on SLE kernel maintenance allows to achieve much better maintenance with much less effort than anything else. That's why Evergreen 11.4 uses 3.0 kernel (SLE11-SP2 based) and Evergreen 13.1 is planned to use 3.12 kernel (SLE12 based - I consider a move to SLE12-SP1 but that's for a different discussion in other lists). Even if openSUSE provides features and drivers not provided or not supported by SLE (including the whole i586 architecture now), the result is still much better that anything else we can realistically think of.
The main problem is that SLE and openSUSE have very different target audiences, the SLE kernel targets entreprise customers which tend to use enterprise hardware which is substantially different from consumer grade hardware with completely different lifecycles. Providing a kernel which lags years behind means (even with limited backports) that openSUSE releases likely cannot be used by a large share of endusers using more current consumer hardware (this is particularly an issue with consumer-grade laptops for which driver support is already problematic in the current situation). So I don't think this works at all and makes it much worse that what we have today, i.e. every 8 months a release containing the most current kernel at the time of the release even if its maintenance leaves much to be desired.
consider the opportunity to utilize the work behind SLE kernel. Perhaps having both might satisfy both types of users.
Perhaps, do you have the resources to support two kernels for each openSUSE release? -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 6 May 2015 at 12:43, Guido Berhoerster <gber@opensuse.org> wrote:
The main problem is that SLE and openSUSE have very different target audiences, the SLE kernel targets entreprise customers which tend to use enterprise hardware which is substantially different from consumer grade hardware with completely different lifecycles. Providing a kernel which lags years behind means (even with limited backports) that openSUSE releases likely cannot be used by a large share of endusers using more current consumer hardware (this is particularly an issue with consumer-grade laptops for which driver support is already problematic in the current situation). So I don't think this works at all and makes it much worse that what we have today, i.e. every 8 months a release containing the most current kernel at the time of the release even if its maintenance leaves much to be desired.
Are these statements that the "SLE Kernel doesn't have the hardware support required" based on fact, actual experience, or just theory based on assumptions? SUSE do lots of work for hardware enablement, there's already plenty in the current SLE 12 Kernel and more on the way with SLE 12 SP1. As we're not that far away from SLE 12 SP1, the current SLE 12 kernel is probably the 'oldest' it will ever be in regards to hardware enablement, so I'd be very curious to hear of actual examples of people who've downloaded (and patched) the current SLED 12 Kernel from the free trial SUSE offer and found bonefide 'this isn't good enough for my hardware' issues You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
consider the opportunity to utilize the work behind SLE kernel. Perhaps having both might satisfy both types of users.
Perhaps, do you have the resources to support two kernels for each openSUSE release?
As Michal has pointed out, we don't really have the resources to fully/properly support the one kernel we're currently shipping with the current openSUSE Regular Release. So I think it's probably safe to say the answer to your question is 'no' -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 2015-05-06 13:45, Richard Brown wrote:
On 6 May 2015 at 12:43, Guido Berhoerster <gber@opensuse.org> wrote:
consider the opportunity to utilize the work behind SLE kernel. Perhaps having both might satisfy both types of users.
Perhaps, do you have the resources to support two kernels for each openSUSE release?
As Michal has pointed out, we don't really have the resources to fully/properly support the one kernel we're currently shipping with the current openSUSE Regular Release.
So advance the kernel version when one goes out of support upstream. Those who want to stay at one version can do so with the SLE one. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Richard Brown <RBrownCCB@opensuse.org> [2015-05-06 13:45]:
On 6 May 2015 at 12:43, Guido Berhoerster <gber@opensuse.org> wrote:
The main problem is that SLE and openSUSE have very different target audiences, the SLE kernel targets entreprise customers which tend to use enterprise hardware which is substantially different from consumer grade hardware with completely different lifecycles. Providing a kernel which lags years behind means (even with limited backports) that openSUSE releases likely cannot be used by a large share of endusers using more current consumer hardware (this is particularly an issue with consumer-grade laptops for which driver support is already problematic in the current situation). So I don't think this works at all and makes it much worse that what we have today, i.e. every 8 months a release containing the most current kernel at the time of the release even if its maintenance leaves much to be desired.
Are these statements that the "SLE Kernel doesn't have the hardware support required" based on fact, actual experience, or just theory based on assumptions?
SUSE do lots of work for hardware enablement, there's already plenty in the current SLE 12 Kernel and more on the way with SLE 12 SP1.
As we're not that far away from SLE 12 SP1, the current SLE 12 kernel is probably the 'oldest' it will ever be in regards to hardware enablement, so I'd be very curious to hear of actual examples of people who've downloaded (and patched) the current SLED 12 Kernel from the free trial SUSE offer and found bonefide 'this isn't good enough for my hardware' issues
Huh, are you actually serious? Unless SUSE is planning on backporting every single new driver that matters for endusers using cheap laptops then this is inevitably becomes an issue and moreso than today where we get at least _all_ the latest drivers every 8 months.
You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
And I'd prefer if the discussion continued with less insinuations and handwaving of concerns. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On May 6, 2015 5:23:38 AM EDT, Michal Kubecek <mkubecek@suse.cz> wrote:
On Wednesday 06 of May 2015 00:41:09 Carlos E. R. wrote:
As a stable release user, I would be very excited by every new
feature
that I know is available in Tumbleweed that makes into Stable. And I would be very disappointed to find out that they will not be and that I have to wait another year for them.
One thing we should keep in mind is that openSUSE kernel maintenance doesn't work very well. You could read previous discussion on this topic for details. The short story is that unless we are lucky to pick a longterm kernel version (and we haven't been for quite long), upstream stable branch is closed even before openSUSE reaches its release and after that, only security bugs and small amount of functional fixes is backported into it.
We do not have resources to do openSUSE kernel maintenance comparable to at least upstream stable branches. Piggybacking on SLE kernel maintenance allows to achieve much better maintenance with much less effort than anything else. That's why Evergreen 11.4 uses 3.0 kernel (SLE11-SP2 based) and Evergreen 13.1 is planned to use 3.12 kernel (SLE12 based - I consider a move to SLE12-SP1 but that's for a different discussion in other lists). Even if openSUSE provides features and drivers not provided or not supported by SLE (including the whole i586 architecture now), the result is still much better that anything else we can realistically think of.
Of course, we can still do what we have been doing so far: take the latest released kernel available at the moment and do the minimum maintenance. But I can't help thinking it would be a pity if we didn't consider the opportunity to utilize the work behind SLE kernel. Perhaps
having both might satisfy both types of users.
Michal Kubeček
From: http://en.m.wikipedia.org/wiki/SUSE_Linux_distributions There was a 2 1/2 year period without a major SLE kernel release recently Feb 2012-kernel 3.0 (and still used for SLE 11.3) Oct 2014-kernel 3.12 Assume there was a summer 2014 opensuse release. What are you proposing as relates to leveraging the SLE kernel? Release with 3.0 then have an opensuse "service pack" to get the 3.12 kernel out? If not, will SLE even support 3.0 through the lifetime of a summer 2014 opensuse release? I.e.. until fall 2016 I assume. When in 2014 would the SLE 3.12 kernel have been available to opensuse? Maybe the opensuse release could have been delayed to wait for it? Btw: am I right that Evergreen leveraged the SLE 3.0 kernel and plans to leverage the 3.12 kernel? Greg -- Sent from my Android device with K-9 Mail. Please excuse my brevity. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 06 of May 2015 14:20:53 Guido Berhoerster wrote:
Huh, are you actually serious? Unless SUSE is planning on backporting every single new driver that matters for endusers using cheap laptops then this is inevitably becomes an issue and moreso than today where we get at least _all_ the latest drivers every 8 months.
The point is if you want "all the latest drivers", you can use Tumbleweed and have them every month or so. This discussion is rather about a stable, reliable and reasonably maintained distribution to provide for (e.g.) three years as an alternative for those who prefer the "stable" direction over the "latest and greatest" one. IMHO the main misunderstanding is that you base your comments on the assumption that the way openSUSE kernel has been maintained until now worked pretty well so that it would be best to go on with it. I don't believe this is the case and previous discussions (there were e.g. suggestions to move 13.1 to 3.12 kernel even within regular support period) show that I'm not the only one. Michal Kubeček -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 06 of May 2015 08:29:55 greg.freemyer@gmail.com wrote:
Assume there was a summer 2014 opensuse release. What are you proposing as relates to leveraging the SLE kernel?
IMHO we shouldn't automatically assume that if we decide to transform "release openSUSE" to be base it on SLE core packages, we keep the release schedule we've had in the previous model (every 8/12 months). If we go this way, it may make sense to have longer intervals and to adjust the timing to SLE releases. In other words...
When in 2014 would the SLE 3.12 kernel have been available to opensuse? Maybe the opensuse release could have been delayed to wait for it?
Yes, it would IMHO make good sense to align that openSUSE release (if we were already in the mode we are only discussing now) to SLE12 release and use SLE12 based kernel.
Btw: am I right that Evergreen leveraged the SLE 3.0 kernel and plans to leverage the 3.12 kernel?
Yes. Evergreen 11.4 has 3.0 kernel based on SLE11 SP2 (you may find my earlier e-mail(s) explaining in detail what exactly this means) that replaced the original 2.6.37 and there is a SLE12 based kernel for 13.1 (in home:mkubecek:evergreen-13.1 OBS project) that is supposed to become Evergreen 13.1 kernel unless we decide to base it on SLE12-SP1 instead (which becomes quite appealing as 13.1 regular support is likely to end later than originally expected). It wasn't painless but I believe the overall experience was positive. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Michal Kubecek <mkubecek@suse.cz> [2015-05-06 14:46]:
On Wednesday 06 of May 2015 14:20:53 Guido Berhoerster wrote:
Huh, are you actually serious? Unless SUSE is planning on backporting every single new driver that matters for endusers using cheap laptops then this is inevitably becomes an issue and moreso than today where we get at least _all_ the latest drivers every 8 months.
The point is if you want "all the latest drivers", you can use Tumbleweed and have them every month or so. This discussion is rather
Sure, but I don't want to deal with the churn and occasional fallout on the laptop which I need for doing work. Of course *I* can run Kernel:stable (and have done so in the past because of lacking touchpad and graphics driver support) but the question is whether we only want to make a distro for openSUSE/SUSE developers or people who simply want towork on their laptops and cannot or do not want to deal with Tumbleweed.
about a stable, reliable and reasonably maintained distribution to provide for (e.g.) three years as an alternative for those who prefer the "stable" direction over the "latest and greatest" one.
Sure, that's great if you buy/lease a server today and want to run the same distro and kernel until it's written off. But for the desktop it is not so great, I have no strong opinion on a more stable userland middle layer but kernel/Mesa/X should be exempt from that and receive a refresh every 8 or maybe 12 months as they do today.
IMHO the main misunderstanding is that you base your comments on the assumption that the way openSUSE kernel has been maintained until now worked pretty well so that it would be best to go on with it. I don't believe this is the case and previous discussions (there were e.g. suggestions to move 13.1 to 3.12 kernel even within regular support period) show that I'm not the only one.
I do not assume that is has been working out well, but to me it sounds SUSE is scaling back even more in this area. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
IMHO the main misunderstanding is that you base your comments on the assumption that the way openSUSE kernel has been maintained until now worked pretty well so that it would be best to go on with it. I don't believe this is the case and previous discussions (there were e.g. suggestions to move 13.1 to 3.12 kernel even within regular support period) show that I'm not the only one.
I do not assume that is has been working out well, but to me it sounds SUSE is scaling back even more in this area.
I’m curious about your last point there, what makes you think they are scaling back even more in this area? Sincerely, Bob Martens Martin Luther College Webmaster/Technician http://mlc-wels.edu
On May 6, 2015, at 8:16 AM, Guido Berhoerster <gber@opensuse.org> wrote:
* Michal Kubecek <mkubecek@suse.cz> [2015-05-06 14:46]:
On Wednesday 06 of May 2015 14:20:53 Guido Berhoerster wrote:
Huh, are you actually serious? Unless SUSE is planning on backporting every single new driver that matters for endusers using cheap laptops then this is inevitably becomes an issue and moreso than today where we get at least _all_ the latest drivers every 8 months.
The point is if you want "all the latest drivers", you can use Tumbleweed and have them every month or so. This discussion is rather
Sure, but I don't want to deal with the churn and occasional fallout on the laptop which I need for doing work. Of course *I* can run Kernel:stable (and have done so in the past because of lacking touchpad and graphics driver support) but the question is whether we only want to make a distro for openSUSE/SUSE developers or people who simply want towork on their laptops and cannot or do not want to deal with Tumbleweed.
about a stable, reliable and reasonably maintained distribution to provide for (e.g.) three years as an alternative for those who prefer the "stable" direction over the "latest and greatest" one.
Sure, that's great if you buy/lease a server today and want to run the same distro and kernel until it's written off. But for the desktop it is not so great, I have no strong opinion on a more stable userland middle layer but kernel/Mesa/X should be exempt from that and receive a refresh every 8 or maybe 12 months as they do today.
-- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-- This electronic communication, including any attached documents, may contain confidential and/or legally privileged information that is intended only for use by the recipient(s) named above. If you have received this communication in error, please notify the sender immediately and delete the communication and any attachments. Views expressed by the author do not necessarily represent those of Martin Luther College. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed 06 May 2015 01:45:06 PM CDT, Richard Brown wrote: <snip>
You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
<snip> Hi The only thing I've had to compile for the kernel was the it87 module for SLES 12, it is present for SLED 12... On this system HP ProBook 455 G1 since it's an AMD APU is to run the proprietary driver for boost states which means no secure boot. Else no issues with HP 2000, HP ProBook 4330s and H ProBook 4440s with SLED 12 in my tests and using day to day... There have been a few forum posts with respect to newer dual graphics and wireless cards. -- Cheers Malcolm °¿° LFCS, SUSE Knowledge Partner (Linux Counter #276890) SUSE Linux Enterprise Desktop 12 GNOME 3.10.1 Kernel 3.12.39-47-default up 23:34, 3 users, load average: 1.21, 0.42, 0.30 CPU AMD A4-5150M APU @ 3.3GHz | GPU Richland Radeon HD 8350G -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Le mercredi 06 mai 2015 à 08:29 -0500, Malcolm a écrit :
On Wed 06 May 2015 01:45:06 PM CDT, Richard Brown wrote:
<snip>
You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
<snip>
Hi The only thing I've had to compile for the kernel was the it87 module for SLES 12, it is present for SLED 12...
This means it is part of kernel-*-extra package (which is only available on SLED and not on SLES). My proposal would be for an openSUSE kernel based on SLE kernel, we could have a single kernel package which would include the content on kernel + kernel-extra (if kernel team is fine with that)..
On this system HP ProBook 455 G1 since it's an AMD APU is to run the proprietary driver for boost states which means no secure boot.
Else no issues with HP 2000, HP ProBook 4330s and H ProBook 4440s with SLED 12 in my tests and using day to day...
Glad to read that ;) -- Frederic Crozat Enterprise Desktop Release Manager SUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 05/06/2015 06:16 AM, Guido Berhoerster wrote:
have no strong opinion on a more stable userland middle layer but kernel/Mesa/X should be exempt from that and receive a refresh every 8 or maybe 12 months as they do today.
I *strongly* disagree. Churn on kernel/Mesa/X is what gives users the most grief. SLE'S stack here is exceptionally stable, and the ABI-compatibility is part of why there are nvidia binaries for SLE. Leaving aside that I've been bitten by a recent Mesa/nouveau bug that made it through both Freedesktop.org & Tumbleweed's QA, I can say that firming up that part of the stack is going to solve *a lot* of problems for people who want a *stable* system. - -- James Mason Technical Architect, Public Cloud openSUSE Member SUSE jmason@suse.com - ------------------------------------------------------------------------ SUSECon 2015: Register at susecon.com -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBCAAGBQJVSigAAAoJEBs5UYhsRJAjYE8IAJyov80YZuDoef95km/q/aae DSC/jOcmKnoIurfJDvz3G4D8P+9ygD/QazaHtskbOGiR8EtaB0j46z9ponur+2Y2 mKMb/Q3do1yR2tb3nCtKUoUzceHQ3fIUUZsx5m/pH9ZomLhGcXoFK7rnkzqrg62n wPshE4mvMH0UNSwCbL/rgtF86G1BOMxAxtvD07iJGtIDUI2tWkAUckovF98ROFxQ Jxo5AF2MD1gpHo6714AwWZx+tpPxDDgMSKIXUQA8NEzWYxk/b5AibsdtzQ2+8Bf1 pdB/lznqWalymd69LJGHLPS7f7mLmObVG8PnMuUqQbOiAOxEUFm2X/ybS0zJZfE= =jfu9 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Robert Martens <martenrt@mlc-wels.edu> [2015-05-06 15:29]:
IMHO the main misunderstanding is that you base your comments on the assumption that the way openSUSE kernel has been maintained until now worked pretty well so that it would be best to go on with it. I don't believe this is the case and previous discussions (there were e.g. suggestions to move 13.1 to 3.12 kernel even within regular support period) show that I'm not the only one.
I do not assume that is has been working out well, but to me it sounds SUSE is scaling back even more in this area.
I’m curious about your last point there, what makes you think they are scaling back even more in this area?
AFAICT the kernel and X stack in openSUSE releases are currently maintained by SUSE staff. Especially for the kernel the current resources seem barely sufficient. The plan seems to be to cut these resources back by leveraging the same kernel/X stack that is already maintained in SLE. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/05/2015 05:11 PM, Richard Brown wrote:
On 5 May 2015 at 21:40, Martin Schlander <martin.schlander@gmail.com> wrote:
It was also the first release based on the "new" factory/tumbleweed, with a very short testing phase (two devel releases with 2-3 weeks between) and iirc the roadmap itself was published very late.
Secondly, openSUSE releases shouldn't need as many testers/much testing as before, with Tumbleweed already being fairly well tested by a decent number of people running it full-time. At least I thought that was the idea.
The whole process of Tumbleweed based stable releases just needs a few small adjustments, to work well imo. (A roadmap, a slightly longer window to test things, a fixed 12 month release cycle, 26 months of lifetime (2 releases+2 months) would be a very good start).
I understand where you're coming from, but I think there are a few things I'd like you to consider
The 13.2 release required a large amount of work from a small number of people to make it a success. From an Engineering perspective it was 'just' a Tumbleweed snapshot, but from a Marketing, Branding, 'actually getting a release out of the door' perspective, it required as much work as any other software release. For 13.2 this fell on the shoulders of a tiny handful of people. The reality is that Tumbleweed has captured the excitement of a huge portion of our current contributing community, and grown it greatly, and that's great, but it does have the side effect of leaving less enthusiasm and less people to work on the Regular Release. As was discussed at the Project Meeting at oSC 15, the 'Tumbleweed-Snapshot-based-Release' model also has an interesting side effect when it comes to this part of getting a release published and marketed. Just as they share development models, both Tumbleweed and snapshot-based-releases have very similar core messages - "The latest of everything, each release". The major differences are the gap between releases and the method of maintenance after release. On each Release Day for the Regular Release, this ends up leaving us with relatively little exciting to talk about (Everything is already in Tumbleweed) and at the same time, way too much to talk about (the Changelog of 13.1 to 13.2 is insanely long and hard to filter down into exciting things to include in Release announcements) If snapshot-based-releases were to continue, we'd need to find a way to tackle that, and get people excited, talking about, and helping put together the release and it's marketing, something which is basically just the same as what we do every week, but maintained differently.
Correct, we need a better approach to the marketing of the snapshot based releases and we need more people contributing to the snapshot based releases.
But this is where my proposal for a Regular Release based on the SLE Sources comes into play. ( Video available at https://www.youtube.com/watch?v=BH99TSrfvq0 )
Switching the base of the regular release is not a solution to either of our most pressing problems (mentioned above), better marketing for regular releases and more people involvement in the creation of the regular releases. Re-basing the regular release serves a different purpose and thus without addressing the actual issues we face concerning regular releases we will inevitably end up in the same place whether the release is based on SLE sources or on snapshots. The idea of "base it in something new and they will come" is a fallacy. What we really ought to figure out is how to solve the other problems fist. Then consider if re-basing is what we want to do or if we even want a third release with the distro based on SLE sources replacing Evergreen. I don't know. What I do know is that the actual issues surrounding the regular release (a quote from your message): "The 13.2 release required a large amount of work from a small number of people to make it a success." will not be resolved just because we pick a different base. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSlUKAAoJEE4FgL32d2UkRvcH/Rv581fMb+M7pR0uaZzfmiSM E7WYmnFMQb65S67/PAupzr79U/fRQQaWU8ORvSI6tkR06Dvz9z1QpiQC5408Ha9+ C1Lg54VU84SO6l73e8uMpbk7NwcOj0ExCsocgc3oSaAGvp+mXTDBdNP7Z86STby+ 1mSmK8OzC7ZTb2fRDFs2YLppKJgOXOLWtbYQ3ri7pU2pwLnc94NMZ6QpTCkwpcKv Dc/q7Ar4Kv4mkuPiUWW8kHoj4vZYcDebo5uuSUZbRaKh3xpZ7YUnVUallyQJKgfB 98FdqDqMrUjlOmHoSGgAB1LhWE9JLnbVziCz9/cfR6zVjXSbd6V3Md/MHpvgJJA= =h52R -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 05:23 AM, Michal Kubecek wrote:
On Wednesday 06 of May 2015 00:41:09 Carlos E. R. wrote:
As a stable release user, I would be very excited by every new feature that I know is available in Tumbleweed that makes into Stable. And I would be very disappointed to find out that they will not be and that I have to wait another year for them.
One thing we should keep in mind is that openSUSE kernel maintenance doesn't work very well. You could read previous discussion on this topic for details. The short story is that unless we are lucky to pick a longterm kernel version (and we haven't been for quite long), upstream stable branch is closed even before openSUSE reaches its release and after that, only security bugs and small amount of functional fixes is backported into it.
We do not have resources to do openSUSE kernel maintenance comparable to at least upstream stable branches. Piggybacking on SLE kernel maintenance allows to achieve much better maintenance with much less effort than anything else. That's why Evergreen 11.4 uses 3.0 kernel (SLE11-SP2 based) and Evergreen 13.1 is planned to use 3.12 kernel (SLE12 based - I consider a move to SLE12-SP1 but that's for a different discussion in other lists). Even if openSUSE provides features and drivers not provided or not supported by SLE (including the whole i586 architecture now), the result is still much better that anything else we can realistically think of.
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation. Thus for x86 we would end up with a kernel that has a very large number of patches, is based on an older kernel version and will be difficult for others to work with due to the large number of patches if an issue arises that is specific to x86. Plus the chances of getting help from upstream on this kernel are even slimmer than with the approach we take today. I think that for all features that SLE supports we would end up with a better kernel binary than we have in openSUSE today. However for everything else, as stated above I think the argument doesn't apply. Last but not least a backport of a driver for the latest whiz-bang USB plug in gizmo is not going to happen if it is of no interest for SLE. Then we end up in a situation where the upstream kernel supports the feature but openSUSE users cannot access it because it will not be backported because it is of no interest to SLE. Thus, just considering the kernel there are many unanswered questions. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSlgcAAoJEE4FgL32d2UkR1IH/3S3ytnnzUTguV9jIzdehOks +VgNb7H3lKhggTJx+f13/d13twvtQQ/kfv5Ii8XZfscGHFq4Tb2Ly0UMjWMOLG4n vPJmlZxMnD8jdvJBQnDsTXQ5PpfeaxRTufCm4dHRGm8heJ6A7t8J4FXto3vx8GMo qpNSCk4v4ut1dRkMenZOfu4Po/GH9QzO0RV7cIAPcznU04kZK4lobiFD/gvNrvmW 2mjY4MTrMp9ATf1xEzpEu5qDWfgN/J31WLMdeYEItlZVVXizQFNnhftVonv/Now2 QorvGZqStM3M2N+8tl1xDxhmGuelFgjHo0N8eN4LKWFceO+K/MO5F1IkyKHDApw= =O9jw -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 07:45 AM, Richard Brown wrote:
On 6 May 2015 at 12:43, Guido Berhoerster <gber@opensuse.org> wrote:
The main problem is that SLE and openSUSE have very different target audiences, the SLE kernel targets entreprise customers which tend to use enterprise hardware which is substantially different from consumer grade hardware with completely different lifecycles. Providing a kernel which lags years behind means (even with limited backports) that openSUSE releases likely cannot be used by a large share of endusers using more current consumer hardware (this is particularly an issue with consumer-grade laptops for which driver support is already problematic in the current situation). So I don't think this works at all and makes it much worse that what we have today, i.e. every 8 months a release containing the most current kernel at the time of the release even if its maintenance leaves much to be desired.
Are these statements that the "SLE Kernel doesn't have the hardware support required" based on fact, actual experience, or just theory based on assumptions?
Actual experience in my case. Although this effected SLE 11 at the time the problem I ran into was with an Ethernet dongle which worked perfectly fine with openSUSE but the driver had not been backported to SLE 11 kernel, this was passed the kernel upgrade we did in SP3. Given that the kernel development model for SLE hasn't really changed this is a very real issue that we have to deal with. openSUSE users expect a much larger range of hardware to be supported and if the hardware is not on the "interest to SLE" list the backport is not going to happen. Even if it is on the "interest for SLE" list the backport is likely to be slower than getting the feature from a newer upstream kernel in a snapshot based regular release.
SUSE do lots of work for hardware enablement,
Yes, all driven by partners that develop enterprise HW. Almost none driven by requests for consumer grade HW. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSlo7AAoJEE4FgL32d2Uk8tcIAKA7yJ8gXO6W8PXDisr8GaS+ hJIaLBEX1LKyo7uCrWv0cfK3G9AVqltVTsWG2u8ow5zTlTc8G4O15a0w9WLWT9si 2uDxmnyesPgXWbwCr1LLvTTYKEoeKnFj6rvatPDRePu59q6q06sAt0QUPWkKi5ZT 6FECpLRU1tzVt7sX+kdliQXYWUtvy0UwGOIJ+p0Zot/ZD82YdTyqAThiQDOmzLOn VyaY9OvMWhu+xqo3F65xkp04BF97ie6qYkosXwR7uxlkVm0vjqnuj4DmKnlgpaR0 TZ7dhz5Nc+KDcCWNdcTlPcPW/pyMHGZMvhFA0SKMAAsHJNfl2XjXjPHCCkx8QCM= =PfQ9 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 08:46 AM, Michal Kubecek wrote:
On Wednesday 06 of May 2015 14:20:53 Guido Berhoerster wrote:
Huh, are you actually serious? Unless SUSE is planning on backporting every single new driver that matters for endusers using cheap laptops then this is inevitably becomes an issue and moreso than today where we get at least _all_ the latest drivers every 8 months.
The point is if you want "all the latest drivers", you can use Tumbleweed and have them every month or so.
Yes I get it. I either get to reboot my machine many times a week because a new kernel arrives in TW or I get the choice of not getting "new" drivers for years to come. Please consider the [possibility that there is a potentially a sizeable chunk of users that lie in between. As Guido pointed out, for many the "getting all 'new' drivers every 12 months" model appears to be working. What is being proposed is "get some 'new' drivers every 12 month". Lets face it, the chances of consumer product drivers being backported to the SLE base are slim. If the SLE kernel gets re-based every 2 years this implies that instead of "getting all 'new' drivers every 12 months" users end up with "getting some 'new' drivers every 24 months". A doubling of the waiting period and a reduction in features. Not a very compelling story, sorry. Additionally, SUSE is not committed to moving the kernel forward every other service pack (which has no committed release cycle either, just a rough time frame), rather, the "moving the kernel forward" is ad-hoc based on whatever suits the business considerations best.
This discussion is rather about a stable, reliable and reasonably maintained distribution to provide for (e.g.) three years as an alternative for those who prefer the "stable" direction over the "latest and greatest" one.
IMHO the main misunderstanding is that you base your comments on the assumption that the way openSUSE kernel has been maintained until now worked pretty well so that it would be best to go on with it. I don't believe this is the case and previous discussions (there were e.g. suggestions to move 13.1 to 3.12 kernel even within regular support period) show that I'm not the only one.
Well, I personally have encountered only few kernel bugs in our current release model. But then again I try to run on run off the mill hardware. The newer stuff where issues are encountered are the drivers for all the fancy thing one can plug into a USB port these days. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSl3LAAoJEE4FgL32d2UkLnsH/i+Z4QhA80nGwbF7tfHq50mW bYGxgGhnsaCe/hJiomIHidf0tbGXnxVvW552X9VQ1EBNXWTygwiWYCaqQGL2dXxx zqem7vUWkQrFdInbEXD/ibSSwzTi+Kyx4+1OLwASJ5QxPqrn26BSNc+EqWfmqYrJ vfMYibFZrnoWTj1AIz1BblMVaA4a4NIWGFYD0QuvaZy4JE9ZuE5S7i7P6xYy0+7r l749eMYXAuvtS6oqWRPDMRMD2NpbW4xRI4HoWRni3dMaQlGjvf4Qj+2nNbpqAI2V fp+efVPRkX2YVEWa+l4hNWogDlSJvzFU31w/9VuTeglLCwG10YWS1T4r59G3UnE= =OlLN -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard. Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586? How much chance of getting attention to an i586-specific bug would you get? Do you really believe that given the circumstances and popularity of i586 among kernel developers, the state of i586 kernel would be significantly worse if we based openSUSE kernel on SLE (compared to current state)? I don't think so - and even if I did, I don't believe it would outweigh the difference on x86_64. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Guido Berhoerster wrote:
* Michal Kubecek <mkubecek@suse.cz> [2015-05-06 14:46]:
about a stable, reliable and reasonably maintained distribution to provide for (e.g.) three years as an alternative for those who prefer the "stable" direction over the "latest and greatest" one.
Sure, that's great if you buy/lease a server today and want to run the same distro and kernel until it's written off. But for the desktop it is not so great,
You may have different desktop machines, but one of my back-office desktops will (more or less( also run the same distro and kernel until it's written off. -- Per Jessen, Zürich (16.7°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 06, 2015 at 01:53:14PM -0400, Robert Schweikert wrote:
Correct, we need a better approach to the marketing of the snapshot based releases
I don't think marketing is the problem. The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him. It would be even nicer if we could do three distributions: Tumbleweed, something similar to openSUSE <= 13.1 and some "LTS" version (possibly SLE based). But I'm afraid we don't have resources to do all three and there is no easy way to change that fact in short timeframe.
and we need more people contributing to the snapshot based releases. ... "The 13.2 release required a large amount of work from a small number of people to make it a success."
will not be resolved just because we pick a different base.
I disagree. People have free will, they won't come just because we need them. They will if we do something to attract them. I believe offering a solution in the area where we have been failing to provide one for years is more likely to attract users and contributors than offering two versions of a distribution that are not too far from each other. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 02:35 PM, Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586?
Enough that the upgrade from 13.1 to 13.2 on the netbook worked flawlessly. Network, disk and everything I expect works.
How much chance of getting attention to an i586-specific bug would you get? Do you really believe that given the circumstances and popularity of i586 among kernel developers, the state of i586 kernel would be significantly worse if we based openSUSE kernel on SLE (compared to current state)?
I believe that if there were an i586 specific bug and I have a recent enough kernel I can get help upstream. However, with a kernel that is based on an older release and carries a lot of patches that will be more difficult. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSnVQAAoJEE4FgL32d2UkJ3cIAKaQE2XE8RJWHk8x22QFIeOE sHFn1yEwG0y9BCVn39b4hs1Vf3/QBMaPh/NFPbWi+I6G9all3JZQVIkKtLcYBpAv hRvuCUBFTVd0A5QHg56D9z9n5JXIi25csQy/73pt/GTojRIp/ns3aTAVMtk3pDR4 GKN+YnKS03QALX8XF8gu013/nC3K0xxSaglA8vAhAZ/YyQlwjVspmcj34FKIPqMX 3NzFdNVG3vlrdV54apd+khQbYNpVpKrh35p+gwILEbcMaP77t9atayVZi4/ntR81 HsMhBr/2VNLbSrf0lBaloE4WUcIuFIzD3FHyoCGWG2on0coj9OT7AD45UoVGDfs= =1wpS -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 02:53 PM, Michal Kubecek wrote:
On Wed, May 06, 2015 at 01:53:14PM -0400, Robert Schweikert wrote:
Correct, we need a better approach to the marketing of the snapshot based releases
I don't think marketing is the problem. The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him.
It would be even nicer if we could do three distributions: Tumbleweed, something similar to openSUSE <= 13.1 and some "LTS" version (possibly SLE based). But I'm afraid we don't have resources to do all three and there is no easy way to change that fact in short timeframe.
and we need more people contributing to the snapshot based releases. ... "The 13.2 release required a large amount of work from a small number of people to make it a success."
will not be resolved just because we pick a different base.
I disagree. People have free will, they won't come just because we need them. They will if we do something to attract them. I believe offering a solution in the area where we have been failing to provide one for years is more likely to attract users and contributors than offering two versions of a distribution that are not too far from each other.
We shall see. Bottom line is that someone has to do the work. Given that we've had like 5 contributors, if that many to get a regular release out the door it will probably be up to those 5 to decide what hey want to do. Those who do, decide. Once the decision is made and we have something and there is sufficient movement to do something else it will happen, if not it won't . So far I do not think there has been a big rush to actually build packages from the released SLE sources. Maybe this also says something about the overall "excitement" over this event ;) . Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSndFAAoJEE4FgL32d2Uk/qkH/1Mtrxw2j6pJ+5NhVoZMJcqO WK0+Z9nSYQHoPatYmGXa5UEzm6n2zrZky9QZCw1yI482QMgM+cFTJY2tR9GSECqV fiD92gh3U3q0IT7VFstOQ6KCTVcfmUXSbBRc7rcsyJagqU3zQLLBa5cqh+97mfME Lw8HECfeBUvhVb1cDHOUxDehi7Bz+03e31GSq/5E736dR6H9oiYtrxpcLHu/5DGU 6xlct3eGzUFYuxKcwbf/YleyONRHjf8IRfWJUbKOkLymNGX7Cjd99tyF2Vwo1FWK GTsUrUhEKw3JRCQydiobSnDdWre4agU/XTpc3iuaKwRgE+B9amPV3nIRALVafiE= =4wXH -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek composed on 2015-05-06 20:35 (UTC+0200):
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586?
I have many test installations. Most by far are 32 bit regardless of CPU type. At least 3 do not support SSE2.
How much chance of getting attention to an i586-specific bug would you get? Do you really believe that given the circumstances and popularity of i586 among kernel developers, the state of i586 kernel would be significantly worse if we based openSUSE kernel on SLE (compared to current state)? I don't think so - and even if I did, I don't believe it would outweigh the difference on x86_64.
I find 64 bit chronically annoying, and so use it on few machines even capable of using it. Most of mine are not. The one I most use, up 24/7 and I write this with is on i586 even though it's an E7600 with 4G RAM and using RAID. 1-most of my machines have less than 4G RAM, and many don't support more than 1GB, which is ample for most users for email, shopping, banking, youtube, etc. 2-64 bit installations take more disk space, which in turn means updates require more time and bandwidth for the duplicative required -32bit packages 3-last I checked specifically, zypper still offered no option to not clutter search results with i586 packages on 64 bit installations 4-whatever speed advantage 64 bit may have over 32 bit has never become apparent to me 5-need for >4G RAM has yet to manifest here 6-killing off 32 bit means inducing otherwise premature migration of working machines into landfills, econeg. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 06.05.2015 um 22:10 schrieb Robert Schweikert:
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
Robert, I think I heard you say in The Haag that openSUSE needs to stop trying to be everything for everyone and concentrate on a specific user group. We can move i586 support into a ports project - just as we have arm7 and ppc and you can be maintainer. But it really shouldn't be in our focus in 2015. Greetings, Stephan -- Ma muaß weiterkämpfen, kämpfen bis zum Umfalln, a wenn die ganze Welt an Arsch offen hat, oder grad deswegn. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 06, 2015 at 04:10:56PM -0400, Robert Schweikert wrote:
On 05/06/2015 02:35 PM, Michal Kubecek wrote:
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
And you would (unless we are brave enough to drop i586), they would just be less tested than on x86_64. Which wouldn't really be that much different from today's 13.2 or Tumbleweed. Out of curiosity... the low-end netbook I bought in summer 2010 has a 64-bit processor; how old is yours that it doesn't?
Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586?
Enough that the upgrade from 13.1 to 13.2 on the netbook worked flawlessly. Network, disk and everything I expect works.
And so does everything on mine running 13.1 with SLE12 based kernel. BtW, I have this combination on all my machines except two and all run without problems (I just don't consider it a statistically relevant sample). One of the remaining two has Evergreen 11.4 (with SLE11-SP2 based kernel) and runs fine as well. The only machine I have problems with is one I dared to install 13.2 on (but those are not kernel problems). Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 2015-05-06 20:35, Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
One does not even need to try hard. http://www.ebay.de/itm/IBM-eServer-xSeries-x-335-2-x-XEON-2-8GHz-CD-/140672266437?pt=LH_DefaultDomain_77&hash=item20c0b870c5 These kinds of servers are somewhat popular with the gaming community (at least the guys I am in contact with) to host game daemons at gatherings. As you can see, these are quite the vintage systems — previously used systems that other entities have decommissioned. However, they still pack enough punch per watt to be not written off by their current new owners.
Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586?
I can tell by experience of administering aforementioned IBM models: sufficient.
How much chance of getting attention to an i586-specific bug would you get?
i586 is so well-establlished that I do not expect any more issues with the platform.
Do you really believe that given the circumstances and popularity of i586 among kernel developers, the state of i586 kernel would be significantly worse if we based openSUSE kernel on SLE (compared to current state)?
In case of i586, a SLE kernel might do. But since openSUSE produces a $modern kernel _anyway_, snatching of the i586 build result for said modern kernel is cheap. coolo's idea of placing i586 onto the ports backburner sounds acceptable. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 06, 2015 at 04:43:57PM -0400, Felix Miata wrote:
I find 64 bit chronically annoying,
Not sure what to imagine under that.
1-most of my machines have less than 4G RAM,
x86_64 does not require 4 GB to show its advantages. Example 1: these two code fragments show the same simple function compiled for x86_64 and i586: 0x0000000000400560 <+0>: add %rsi,%rdi 0x0000000000400563 <+3>: add %rdi,%rdx 0x0000000000400566 <+6>: lea (%rdx,%rcx,1),%rax 0x000000000040056a <+10>: retq 0x08048450 <+0>: mov 0x4(%esp),%edx 0x08048454 <+4>: mov 0x8(%esp),%eax 0x08048458 <+8>: add %edx,%eax 0x0804845a <+10>: mov 0xc(%esp),%ecx 0x0804845e <+14>: add %ecx,%eax 0x08048460 <+16>: mov 0x10(%esp),%edx 0x08048464 <+20>: add %edx,%eax 0x08048466 <+22>: ret The difference is quite obvious, I would say. Example 2: recently I helped a guy running 32-bit system who hit a problem where adding a netfilter rule failed on memory allocation even if about 2/3 of his 2 GB were still free. I won't go into technical details but his problem was directly caused by the deficiencies of i586 architecture and would never happen if he was running a x86_64 system (with the same amount of memory).
2-64 bit installations take more disk space, which in turn means updates require more time and bandwidth for the duplicative required -32bit packages
None of my installs, desktop or server, needs more than 16 GB for root filesystem, most fit into 8 GB. How much could I possibly save by using i586 system? Not much, really. And most of my installs don't actually need -32bit compatibility packages.
4-whatever speed advantage 64 bit may have over 32 bit has never become apparent to me
Try "openssl speed rsa". Last time I checked, the ratio was somewhere between 3:1 and 4:1, that's apparent enough to me. It might be even better with newer generations of CPU's and newer versions of gcc.
5-need for >4G RAM has yet to manifest here
Same comment as above - plus that I'm really happy that I put 32 GB into my working machine two years ago when the prices were half of today's.
6-killing off 32 bit means inducing otherwise premature migration of working machines into landfills, econeg.
I wouldn't be so sure about that, old machines tend to be less energy efficient. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 07, 2015 at 12:01:09AM +0200, Jan Engelhardt wrote:
On Wednesday 2015-05-06 20:35, Michal Kubecek wrote:
haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
One does not even need to try hard. http://www.ebay.de/itm/IBM-eServer-xSeries-x-335-2-x-XEON-2-8GHz-CD-/140672266437?pt=LH_DefaultDomain_77&hash=item20c0b870c5
OK, I should have said it explicitely: I meant a new one. I guess if I wanted, I could still get a ZX Spectrum, maybe even ZX-81. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hello, Am Mittwoch, 6. Mai 2015 schrieb Robert Schweikert:
We shall see.
Bottom line is that someone has to do the work. Given that we've had like 5 contributors, if that many to get a regular release out the door it will probably be up to those 5 to decide what hey want to do. Those who do, decide.
No objection here, but please keep in mind that replacing the base packages with the SLE packages more or less invalidates the testing done in/with tumbleweed. So if we decide to release 13.3 with the SLE packages as base, we'll need to do quite some additional testing. In exchange, we get the testing done for SLE and easier/shared maintenance.
So far I do not think there has been a big rush to actually build packages from the released SLE sources. Maybe this also says something about the overall "excitement" over this event ;) .
Maybe. It could also be that people just need some time to do that or are still busy with the things that piled up on their desk while they were at oSC ;-) I'd also guess that most people here don't know what exactly is in SLES and SLED (speaking of package list, versions and backported patches) and how they differ from openSUSE. If you or someone else has a list with the major differences between SLE* and openSUSE, it would be more than welcome. I heard at oSC that the SLE* DVD images are available for download somewhere[tm] to test SLE* (does someone have the download link at hand?), but that's also something the average factory contributor might not know. BTW: It would be nice to have the SLE* kernel available in a repo [1] (packages built for tumbleweed and maybe 13.2, not only sources) so that we could change the kernel part of this discussion from "$version looks old" to actual testing ;-) And finally, the SLE* source release is a big event. Not only on the excitement level, but also from the amount of code. I'd guess people just need some time to find their way around before they start using the SLE* sources. This might even be seen positive - people first think about what they want to do with it, instead of using the headless chicken mode ;-) Regards, Christian Boltz [1] no, I didn't check if it exists already ;-) - if it does, please just send a link -- What are you doing?!? The message is over, GO AWAY! -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 6, 2015 at 5:43 PM, Felix Miata <mrmazda@earthlink.net> wrote:
2-64 bit installations take more disk space, which in turn means updates require more time and bandwidth for the duplicative required -32bit packages
The -32bit packages are not required..(maybe except a few small ones) , the problem is that something "recommends" them, either in the patterns or spec files.. (I have no looked where that comes from)
3-last I checked specifically, zypper still offered no option to not clutter search results with i586 packages on 64 bit installations
4-whatever speed advantage 64 bit may have over 32 bit has never become apparent to me
A good easy to test example is openSSL.. where the most commonly used algorithms are significantly faster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 06, 2015 at 04:19:17PM -0400, Robert Schweikert wrote:
So far I do not think there has been a big rush to actually build packages from the released SLE sources. Maybe this also says something about the overall "excitement" over this event ;) .
Seriously? It's exactly one week since the announcement - which was really terse and kind of cryptic and didn't even mention _where_ in OBS the sources are. And openSUSE conference fell into that week. BtW, do you have something to support that? Or do you just _assume_ it? Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek composed on 2015-05-07 00:27 (UTC+0200):
6-killing off 32 bit means inducing otherwise premature migration of working machines into landfills, econeg.
I wouldn't be so sure about that, old machines tend to be less energy efficient.
It isn't just about cost to use. It's also about raw materials depletion and other costs to manufacture and ship, and about the relationships among them. It makes little sense to incur cost to acquire greater efficiency if actual use is low making payback negative or long. Increasing MPG from 15MPG to 30MPG will not recover a $20,000 investment through fuel cost savings at North American fuel prices if driving only 3,300 miles/year. Newer == better is a theory that doesn't necessarily work out as expected. Energy Star has been around a long time, well over a decade, maybe two by now. Suspend and sleep are old too. Assuming all else equal, new that is 4X as fast and 2X as efficient probably still consumes something approaching 2X the power of the baseline. 20X as fast is unlikely to be 20X as efficient. I've been measuring and recording actual consumption by my own machines for years. They vary all over the place. Age is a poor predictor. The biggest indicator of deviation from average electrical consumption is the size of gfxchip heat sink, and fan vs. no fan, but case size and CPU cooling count too, even need for heat sinks on RAM. Small cases have fewer components and usually less capable power supplies that run at higher fraction of capability, which commonly translates to higher net efficiency. Old machines with tiny or no gfxchip heat sink and no fan aren't big power consumers or heat producers. Newer may well use less electricity, but there's nothing like a guarantee total cost will be less. What ain't broke rarely needs fixin. :-) -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Cristian Rodríguez composed on 2015-05-06 19:40 (UTC-0300):
Felix Miata wrote:
2-64 bit installations take more disk space, which in turn means updates require more time and bandwidth for the duplicative required -32bit packages
The -32bit packages are not required..(maybe except a few small ones) , the problem is that something "recommends" them, either in the patterns or spec files
Not often here. The following are included in all my openSUSE installations, most of which started out as minimal in the first place precisely in order to so configure, and prevent the initial installation process from inflating / beyond necessity: zypp.conf: solver.onlyRequires = true zypper.conf: installRecommends = no -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 06:34 PM, Christian Boltz wrote:
Hello,
Am Mittwoch, 6. Mai 2015 schrieb Robert Schweikert:
We shall see.
Bottom line is that someone has to do the work. Given that we've had like 5 contributors, if that many to get a regular release out the door it will probably be up to those 5 to decide what hey want to do. Those who do, decide.
No objection here, but please keep in mind that replacing the base packages with the SLE packages more or less invalidates the testing done in/with tumbleweed. So if we decide to release 13.3 with the SLE packages as base, we'll need to do quite some additional testing.
In exchange, we get the testing done for SLE and easier/shared maintenance.
So far I do not think there has been a big rush to actually build packages from the released SLE sources. Maybe this also says something about the overall "excitement" over this event ;) .
Maybe. It could also be that people just need some time to do that or are still busy with the things that piled up on their desk while they were at oSC ;-)
I'd also guess that most people here don't know what exactly is in SLES and SLED (speaking of package list, versions and backported patches) and how they differ from openSUSE.
If you or someone else has a list with the major differences between SLE* and openSUSE, it would be more than welcome.
I heard at oSC that the SLE* DVD images are available for download somewhere[tm] to test SLE* (does someone have the download link at hand?), but that's also something the average factory contributor might not know.
download.suse.com One can get the iso without requesting an eval code. Install, run as long as you want without updates. Later, RObert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSrUAAAoJEE4FgL32d2Uk/T0IAI+EOR1klwR6czedXcND1RUt YElDiBtHNBoDBdpHBgaHvJtkC1TDM/oRQhp8WQ+b1lNm4YUlmHOGeae0n2tXYv57 rQTan5wtWE+SBfYId71vykIHdf7I36fRQT4rz4dMG5INzqDY8A/9bICH5k4rI/oK JhrvRBLAW4Q5RDXVjqk0pukzzvZEH4C+ZqvW7bNNLUBsDyNyv83vabm6alT/dbQY +/1bZqTNrD3mPIL8S7VQMl6pWur7XSaXQ4JstMbqzTChczuajd3XcKd5Aj0NKpya 5BrPa11Sx0ATr9qXOjzpp6PXcP15x+B3IQEhUmZa/s39S4E7O8KdksuaLWWtQ+8= =awUv -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 05:14 PM, Stephan Kulow wrote:
Am 06.05.2015 um 22:10 schrieb Robert Schweikert:
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
Robert, I think I heard you say in The Haag that openSUSE needs to stop trying to be everything for everyone and concentrate on a specific user group.
No, sorry you misheard. This statement only concerned the marketing message, not the nature of the way we build our distributions or the overall freedom of letting people do what they like and how they want to scratch their itch. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSraMAAoJEE4FgL32d2UkO8oH/RQ5PCVtRAnxZQKomOlb/uHR qdEeUxltnOix1w50GappMbkYzkwUf6Inf5CcnmIlFfR+E6++/U26d2d0mtQhFOrg 6RDi+t6q6zQKdFqSAzBe4aL+uB/xfMPnwAxAQzjx5WMrm6ULPejltQoOdxNkVN2G b4Dg5Klhu/Xz+Y0dxm0KD6eSG797XUOS7STcOExIaMAIOV4BHH0sRFb7lfpW6JhD +Yt9Z6CFlh0r5BjOlalB8phkPziQJRxRvklFmTaoxtmu9OPIbhHn0Pt0BPCfJtRd 3zI6ejNOtSysdSJ9u9H/YVP31ajj0i/WVElpFr4R/OOV91Z2CXKIqWKcINeOd9E= =MOJG -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/06/2015 06:00 PM, Michal Kubecek wrote:
On Wed, May 06, 2015 at 04:10:56PM -0400, Robert Schweikert wrote:
On 05/06/2015 02:35 PM, Michal Kubecek wrote:
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
And you would (unless we are brave enough to drop i586), they would just be less tested than on x86_64. Which wouldn't really be that much different from today's 13.2 or Tumbleweed. Out of curiosity... the low-end netbook I bought in summer 2010 has a 64-bit processor; how old is yours that it doesn't?
I have no idea. It was one of the first Inspiron 10 that shipped with another distro pre-loaded and help me avoid paying the proprietary pretent OS tax. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVSrfMAAoJEE4FgL32d2Uk4vgIAL7DxsJ6K/HvWwkQxtVyJrLE 9WD3aoremwz1M4FwWUoFwVx118hSV6gVpFveOKx8dXSeNrvgeTmc2WAvSHTRwF0G w2Mt3xIv4R2Cz7V7YdqjCcMr5pVTRP7utSFN0RZwJ+fW4P0InF7Enp6XHL7Ef0wl QgL1ObcLgs5tcBlzDWN5opRel5XwfTHAZTqEgoe1Zc8EWZDrGnWuaun8l62azvTa mlgSc9fcplUEjZLB5MjeJ4HmfezsO/I4v/3z3jsG+Zu9YkzydLKP3h/MOEZvnsMP l6YIE0a38eLIWjQbg5YA867aDNC63qrV5xixeXkx6p+vkCeizMBPQXyc37WxUiw= =n+KU -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
В Thu, 07 May 2015 00:34:56 +0200 Christian Boltz <opensuse@cboltz.de> пишет:
I heard at oSC that the SLE* DVD images are available for download somewhere[tm] to test SLE* (does someone have the download link at hand?),
https://www.suse.com/products/server/download/amd64.html -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wednesday 06 of May 2015 20:24:23 Felix Miata wrote:
Newer may well use less electricity, but there's nothing like a guarantee total cost will be less.
No guarantee, sure. I just wanted to point out that replacing an old machine is not necessarily a bad thing from the "eco" point of view. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 07.05.2015 um 02:49 schrieb Robert Schweikert:
On 05/06/2015 05:14 PM, Stephan Kulow wrote:
Am 06.05.2015 um 22:10 schrieb Robert Schweikert:
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
Robert, I think I heard you say in The Haag that openSUSE needs to stop trying to be everything for everyone and concentrate on a specific user group.
No, sorry you misheard. This statement only concerned the marketing message, not the nature of the way we build our distributions or the overall freedom of letting people do what they like and how they want to scratch their itch.
But that doesn't mean I have to scratch your itches too, right? Because I don't want to. Greetings, Stephan - -- Ma muaß weiterkämpfen, kämpfen bis zum Umfalln, a wenn die ganze Welt an Arsch offen hat, oder grad deswegn. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlVLBSIACgkQwFSBhlBjoJY/OgCfZ8FNxGyW8dVmmB0pU0mZZu2Z MIUAn1BI52RFGtRXLlioZ28BE/LHOVcp =gsir -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
We use 32-bit openSUSE. Mainly because we integrate with lots of devices that are 16/32-bit. We have been careful in our code to try to be sure to state the real size of things that have an explicit size. But we have not had the resources to test a 64-bit compile of everything. It is on our ever-so-long to do list. Also, we use some libraries like the Intel Performance Primitives for which we have purchased the 32-bit versions. All these things, of course, can be solved. As it turns out, we hired a new developer who unknowingly installed the 64-bit openSUSE. I thought: here will be our 64-bit Guinea pig. We discovered that an important proprietary library for which we have the source (from SICK AG) has oodles of 32-bit assumptions. It is the most convoluted C++ code I have had the displeasure to meet. Do we want to try to rewrite their code? Naahh. And they did not really like supporting Linux in the first place. I think the code they provided us was really ported to Qnx by a different customer. We found that it seems to work on Linux. I realize that we need to resolve this issue. But it is not an easy one. I understand the advantage of focusing on one architecture. But is 32-bit really dead? I say: not quite yet... As a minimum, it might be a good idea to decide to eliminate it at some reasonable point in the future, and make an official announcement of that. This will let people plan the transition. The official announcement and a time schedule will let people convince the powers that be that resources are needed within that time frame to resolve any local issues. Roger Oberholtzer Ramböll RST Office: +46 (0)10-615 6020 Mobile: +46 (0)70-815 1696 roger.oberholtzer@ramboll.se ________________________________________ Ramböll Sverige AB Krukmakargatan 21 P.O. Box 17009 SE-104 62 Stockholm, Sweden www.rambollrst.se ________________________________________ From: Michal Kubecek [mkubecek@suse.cz] Sent: Thursday, May 07, 2015 12:00 AM To: opensuse-factory@opensuse.org Subject: Re: [opensuse-factory] Road-map for openSuse 13.3? On Wed, May 06, 2015 at 04:10:56PM -0400, Robert Schweikert wrote:
On 05/06/2015 02:35 PM, Michal Kubecek wrote:
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Well, I still use a netbook every now and then that I like to update to get security fixes etc.
And you would (unless we are brave enough to drop i586), they would just be less tested than on x86_64. Which wouldn't really be that much different from today's 13.2 or Tumbleweed. Out of curiosity... the low-end netbook I bought in summer 2010 has a 64-bit processor; how old is yours that it doesn't?
Try to answer honestly: how much testing and QA did current (13.2 or Tumbleweed) openSUSE kernel get on i586?
Enough that the upgrade from 13.1 to 13.2 on the netbook worked flawlessly. Network, disk and everything I expect works.
And so does everything on mine running 13.1 with SLE12 based kernel. BtW, I have this combination on all my machines except two and all run without problems (I just don't consider it a statistically relevant sample). One of the remaining two has Evergreen 11.4 (with SLE11-SP2 based kernel) and runs fine as well. The only machine I have problems with is one I dared to install 13.2 on (but those are not kernel problems). Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 07:29:32 Roger Oberholtzer wrote:
I understand the advantage of focusing on one architecture. But is 32-bit really dead? I say: not quite yet...
Well it definitely is not dead. :-( I'm afraid 10 years from now, I might still be reading e-mails explaining we can't get rid of i586 (together with those explaining we can't get rid of ifconfig).
As a minimum, it might be a good idea to decide to eliminate it at some reasonable point in the future, and make an official announcement of that. This will let people plan the transition. The official announcement and a time schedule will let people convince the powers that be that resources are needed within that time frame to resolve any local issues.
I should make one thing straight: I did _not_ suggest next openSUSE release shouldn't provide i586 architecture. All I did want to say was that the fact that SLE12 is not provided for (and supported on) i586 shouldn't be seen as relevant for the discussion if we want to base next (or any future) openSUSE release on SLE packages. And that's for the reasons I already mentioned: it does not stop us from providing i586 openSUSE and if we do, i586 openSUSE wouldn't get significantly less testing, attention and support than recent releases had. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 09:44:18 Michal Kubecek wrote:
... it does not stop us from providing i586 openSUSE and if we do, i586 openSUSE wouldn't get significantly less testing, attention and support than recent releases had.
Or more precisely: "... than it would have if we based it on a Factory snapshot." Those are going to decline in time in any case. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 2015-05-07 09:44, Michal Kubecek wrote:
On Thursday 07 of May 2015 07:29:32 Roger Oberholtzer wrote:
I understand the advantage of focusing on one architecture. But is 32-bit really dead? I say: not quite yet...
Well it definitely is not dead. :-( I'm afraid 10 years from now, I might still be reading e-mails explaining we can't get rid of i586 (together with those explaining we can't get rid of ifconfig).
Software is cheap(er) to substitute, as there is no tax (in both a legal and a practical sense) on "disposing" of electrons. Fedora has already gotten rid of ifconfig by default, and in post-13.2 TW, it finally is gone too by default. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 07, 2015 at 12:43:26AM +0200, Michal Kubecek wrote:
On Wed, May 06, 2015 at 04:19:17PM -0400, Robert Schweikert wrote:
So far I do not think there has been a big rush to actually build packages from the released SLE sources. Maybe this also says something about the overall "excitement" over this event ;) .
Seriously? It's exactly one week since the announcement - which was really terse and kind of cryptic and didn't even mention _where_ in OBS the sources are. And openSUSE conference fell into that week.
BtW, do you have something to support that? Or do you just _assume_ it?
Well the sources and binaries are in: SUSE:SLE-12:GA and SUSE:SLE-12:Update And I am building GA in home:msmeissn:sle12 if someone is curious how this is setup with a project link. (Currently also bootstrapping the i586 tree so I can build Wine on the outside.) CIao, Marcus -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, 2015-05-06 at 17:05 +0200, Guido Berhoerster wrote:
AFAICT the kernel and X stack in openSUSE releases are currently maintained by SUSE staff. Especially for the kernel the current resources seem barely sufficient. The plan seems to be to cut these resources back by leveraging the same kernel/X stack that is already maintained in SLE.
The key point to a maintained kernel is to pick one that is in the long term support. Or to take what Canonical produces. Whatever we do will not replace a stable kernel tree. Regards Oliver -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Jan Engelhardt composed on 2015-05-07 09:59 (UTC+0200):
Software is cheap(er) to substitute, as there is no tax (in both a legal and a practical sense) on "disposing" of electrons. Fedora has already gotten rid of ifconfig by default,
But still available. :-)
and in post-13.2 TW, it finally is gone too by default.
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'. It's been conspicuously absent to my eyes every visit to man ip. :-) Is some other command required in TW when ifconfig is not installed? -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 2015-05-07 11:39, Felix Miata wrote:
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'
ip a -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Jan Engelhardt composed on 2015-05-07 11:46 (UTC+0200):
Felix Miata wrote:
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'
ip a
Why is that missing from the usage message and the man page? Most people are not clairvoyant with eidetic memory. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 2015-05-07 12:08, Felix Miata wrote:
Jan Engelhardt composed on 2015-05-07 11:46 (UTC+0200):
Felix Miata wrote:
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'
ip a [== ip addr]
Why is that missing from the usage message and the man page?
Is it? $ ip Usage: ip [ OPTIONS ] OBJECT { COMMAND | help } where OBJECT := { link | addr | addrlabel | route ... The only nitpick bug there is that it ought to say [ COMMAND | help ] instead of { COMMAND | help }. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, 7 May 2015 12:08, Felix Miata wrote:
Jan Engelhardt composed on 2015-05-07 11:46 (UTC+0200):
Felix Miata wrote:
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'
ip a
Why is that missing from the usage message and the man page? Most people are not clairvoyant with eidetic memory.
Quote from the "man 8 ip", see the sentence in the middle? [quote] xfrm - manage IPSec policies. The names of all objects may be written in full or abbreviated form, f.e. address is abbreviated as addr or just a. COMMAND Specifies the action to perform on the object. The set of possible actions depends on the object type. [/quote] and TBH, I've overlooked it, too. Nice to know. Now, for the question: "ifconfig -a" had some statistics, too, how do I get them with "ip"? Answer: "ip -s link". Use more than one "-s" to give more statistics - Yamaban. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Jan Engelhardt composed on 2015-05-07 12:18 (UTC+0200):
Felix Miata wrote:
Jan Engelhardt composed on 2015-05-07 11:46 (UTC+0200):
Felix Miata wrote:
I still have it in TW20150430, and good thing too. I have yet to discover an equivalent to 'ifconfig -a'
ip a [== ip addr]
Why is that missing from the usage message and the man page?
Is it?
$ ip Usage: ip [ OPTIONS ] OBJECT { COMMAND | help } where OBJECT := { link | addr | addrlabel | route ...
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 7 May 2015 at 13:00, Felix Miata <mrmazda@earthlink.net> wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
"ip a" is short for "ip addr" which is in the man page Stop being so pedantic ;) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 12:25:45 Yamaban wrote:
Now, for the question: "ifconfig -a" had some statistics, too, how do I get them with "ip"?
Answer: "ip -s link". Use more than one "-s" to give more statistics
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 07, 2015 at 07:00:35AM -0400, Felix Miata wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
man 8 ip: "The names of all objects may be written in full or abbreviated form, for exampe address can be abbreviated as addr or just a." Petr
Roger Oberholtzer wrote:
As it turns out, we hired a new developer who unknowingly installed the 64-bit openSUSE. I thought: here will be our 64-bit Guinea pig. We discovered that an important proprietary library for which we have the source (from SICK AG) has oodles of 32-bit assumptions. It is the most convoluted C++ code I have had the displeasure to meet. Do we want to try to rewrite their code? Naahh. And they did not really like supporting Linux in the first place. I think the code they provided us was really ported to Qnx by a different customer. We found that it seems to work on Linux. I realize that we need to resolve this issue. But it is not an easy one.
As long as it's user space you can run such legacy 32bit code on an otherwise 64bit system just fine. That's what the *-32bit packages are for. So for this use case there is no need for an i586 installation medium. cu Ludwig -- (o_ Ludwig Nussel //\ V_/_ http://www.suse.de/ SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg) Maxfeldstraße 5; 90409 Nürnberg; Germany -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
As long as it's user space you can run such legacy 32bit code on an otherwise 64bit system just fine. That's what the *-32bit packages are for. So for this use case there is no need for an i586 installation medium.
I think all the issues would be in user space. We do have a device driver (https://build.opensuse.org/package/show/home:rogeroberholtzer:kernel/univers...) that maps the VME bus into an application's memory. I have no idea if that driver is 32-bit safe. This is used on systems where we run a JEOS-type diskless openSUSE (built with kiwi) that boot with PXE on VME-based PCs. It provides network access to a VME-based real-time measurement system. Like I said, we use openSUSE in adventurous ways. We would have to install both the 64-bit and 32-bit versions of libraries and development packages that we use. Seems a recipe for mistakes. We need a plan... -- Roger Oberholtzer-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 06.05.2015 um 23:14 schrieb Stephan Kulow:
specific user group. We can move i586 support into a ports project - just as we have arm7 and ppc and you can be maintainer. But it really shouldn't be in our focus in 2015.
And even though I'm one of the old farts who still has quite some i686 hardware around (and in active use), i have to heartily agree with Stephan here :-) now if we only could also do away the same with that ppc64le stuff that's polluting my obs mailbox with crazy buildfailures all the time... :-) -- Stefan Seyfried "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Richard Feynman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 07.05.2015 um 09:29 schrieb Roger Oberholtzer:
We use 32-bit openSUSE. Mainly because we integrate with lots of devices that are 16/32-bit. We have been careful in our code to try to be sure to state the real size of things that have an explicit size. But we have not had the resources to test a 64-bit compile of everything. It is on our ever-so-long to do list. Also, we use some libraries like the Intel Performance Primitives for which we have purchased the 32-bit versions. All these things, of course, can be solved.
You can happily ever after run 32bit userland on a 64bit kernel / openSUSE installation -- Stefan Seyfried "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Richard Feynman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 07.05.2015 14:49, Stefan Seyfried wrote:
Am 06.05.2015 um 23:14 schrieb Stephan Kulow:
specific user group. We can move i586 support into a ports project - just as we have arm7 and ppc and you can be maintainer. But it really shouldn't be in our focus in 2015.
And even though I'm one of the old farts who still has quite some i686 hardware around (and in active use), i have to heartily agree with Stephan here :-)
now if we only could also do away the same with that ppc64le stuff that's polluting my obs mailbox with crazy buildfailures all the time... :-)
Is it really that bad? I don't see too any ppc64le specific build failures in stagings at least. Greetings, Stephan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlVLcuMACgkQwFSBhlBjoJa5rACfXa+9Cug0Mp05yn3ewYRxQc6t LvcAn1wGJDZwcURjLReN43g/eQrXd0/0 =/31r -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 6, 2015 at 2:35 PM, Michal Kubecek <mkubecek@suse.cz> wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
I think I have 2 valid use cases for 32-bit: - Booting random PCs I carry openSUSE Live Images made via susestudio with me at all times on my job. Sometimes I have to work with clients very old machines, but other times they are very new. I almost always boot them into openSUSE to work with them. 32-bit boot media has served me well with both situations. - VMs I have numerous VMs on my computer. It is my impression they are smaller with 32-bit. Also, I think the free version of VMware only supports 32-bit Greg -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Thu, May 07, 2015 at 12:01:09AM +0200, Jan Engelhardt wrote:
On Wednesday 2015-05-06 20:35, Michal Kubecek wrote:
haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
One does not even need to try hard.
OK, I should have said it explicitely: I meant a new one. I guess if I wanted, I could still get a ZX Spectrum, maybe even ZX-81.
Sure, what are you offering, I'm sure I've got a couple in the attic still :-) Anyway, I thought you could still get some 32-bit SoC for the embedded and mini market - AMD Geode comes to mind, but I don't know if they even still make that. -- Per Jessen, Zürich (20.5°C) http://www.hostsuisse.com/ - dedicated server rental in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
A 32-bit system can run far more postfix and dovecot processes than a 64bit with the same amount of memory. -- Per Jessen, Zürich (20.0°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Specifically, what would be gained by dropping the i586 build? I still have use-cases for 32bit, but I could just remain on the last release with support. -- Per Jessen, Zürich (19.9°C) http://www.dns24.ch/ - free dynamic DNS, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 2015-05-07 17:22, Per Jessen wrote:
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
A 32-bit system can run far more postfix and dovecot processes than a 64bit with the same amount of memory.
If you want to run that many postfixes, you would rather want to use a X32 userspace, and not a i586 one. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Thursday 07 of May 2015 07:29:32 Roger Oberholtzer wrote:
I understand the advantage of focusing on one architecture. But is 32-bit really dead? I say: not quite yet...
Well it definitely is not dead. :-( I'm afraid 10 years from now, I might still be reading e-mails explaining we can't get rid of i586 (together with those explaining we can't get rid of ifconfig).
ifconfig surely has a lot less life than i586 ... I sometimes revert to ifconfig when I can't remember how to display stats with 'ip' :-) -- Per Jessen, Zürich (19.9°C) http://www.hostsuisse.com/ - dedicated server rental in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Stefan Seyfried wrote:
Am 06.05.2015 um 23:14 schrieb Stephan Kulow:
specific user group. We can move i586 support into a ports project - just as we have arm7 and ppc and you can be maintainer. But it really shouldn't be in our focus in 2015.
And even though I'm one of the old farts who still has quite some i686 hardware around (and in active use), i have to heartily agree with Stephan here :-)
Ditto and ditto. We still have an ancient HP Proliant G1 chugging away in a corner somewhere. One of these days it'll be virtualized, but as long as it runs, there's no big need. -- Per Jessen, Zürich (20.0°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Jan Engelhardt wrote:
On Thursday 2015-05-07 17:22, Per Jessen wrote:
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
A 32-bit system can run far more postfix and dovecot processes than a 64bit with the same amount of memory.
If you want to run that many postfixes, you would rather want to use a X32 userspace, and not a i586 one.
Completely agree, but I have yet to try that out, one of these days I will. I can't help wondering if there are any issues in 32bit apps communicatiung with 64bit apps? E.g. a 32bit postfix talking to a 64bit mysql? -- Per Jessen, Zürich (20.1°C) http://www.hostsuisse.com/ - dedicated server rental in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Thursday 07 of May 2015 12:25:45 Yamaban wrote:
Now, for the question: "ifconfig -a" had some statistics, too, how do I get them with "ip"?
Answer: "ip -s link". Use more than one "-s" to give more statistics
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once.
Must be more like brand new versions - I don't get any stats when I tried it on 13.2. -- Per Jessen, Zürich (20.1°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 17:48:27 Per Jessen wrote:
Michal Kubecek wrote:
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once.
Must be more like brand new versions - I don't get any stats when I tried it on 13.2.
Just checked the git: should be there since iproute2 3.17. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek <mkubecek@suse.cz> writes:
On Thursday 07 of May 2015 17:48:27 Per Jessen wrote:
Michal Kubecek wrote:
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once.
Must be more like brand new versions - I don't get any stats when I tried it on 13.2.
Just checked the git: should be there since iproute2 3.17.
Only ip li, not ip ad. Andreas. -- Andreas Schwab, SUSE Labs, schwab@suse.de GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE 1748 E4D4 88E3 0EEA B9D7 "And now for something completely different." -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 17:33:05 Per Jessen wrote:
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Specifically, what would be gained by dropping the i586 build? I still have use-cases for 32bit, but I could just remain on the last release with support.
I already tried to explain it once but, OK, again: I did _not_ propose to drop i586 build. I just wanted to say that the fact that SLE is not released for and supported on i586 shouldn't be seen as an argument against basing next (or any future) openSUSE release on SLE packages. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Richard Brown composed on 2015-05-07 13:21 (UTC+0200):
Felix Miata wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
"ip a" is short for "ip addr" which is in the man page
Where in the man page exactly? If it was apparent I wouldn't have had reason to ask. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 18:01:42 Andreas Schwab wrote:
Michal Kubecek <mkubecek@suse.cz> writes:
On Thursday 07 of May 2015 17:48:27 Per Jessen wrote:
Michal Kubecek wrote:
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once.
Must be more like brand new versions - I don't get any stats when I tried it on 13.2.
Just checked the git: should be there since iproute2 3.17.
Only ip li, not ip ad.
"ip -s link show" works since "forever". Since iproute2 3.17, "ip -s addr show" works as well: unicorn:~ # rpm -q iproute2 iproute2-3.19-108.1.x86_64 unicorn:~ # ip -s addr show 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 brd 127.255.255.255 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever RX: bytes packets errors dropped overrun mcast 71282 689 0 0 0 0 TX: bytes packets errors dropped carrier collsns 71282 689 0 0 0 0 ... Added by commit 5d5cf1b43706 ("ip address: print stats with -s") Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am Donnerstag 07 Mai 2015, 09:44:18 schrieb Michal Kubecek:
On Thursday 07 of May 2015 07:29:32 Roger Oberholtzer wrote:
I understand the advantage of focusing on one architecture. But is 32-bit really dead? I say: not quite yet...
Well it definitely is not dead. :-( I'm afraid 10 years from now, I might still be reading e-mails explaining we can't get rid of i586 (together with those explaining we can't get rid of ifconfig).
my 2 cents and factoids: I'm running X86_64 where possible, but on my end user desktop and laptop I do have quite a few applications (games) that are distributed in binary form and only as 32bit, so please do NOT eliminate the ability of running 32bit apps, whatever else you guys decide about the 32bit architectures. cheers MH -- gpg key fingerprint: 5F64 4C92 9B77 DE37 D184 C5F9 B013 44E7 27BD 763C -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/07/2015 12:39 AM, Michal Kubecek wrote:
On Wednesday 06 of May 2015 20:24:23 Felix Miata wrote:
Newer may well use less electricity, but there's nothing like a guarantee total cost will be less.
No guarantee, sure. I just wanted to point out that replacing an old machine is not necessarily a bad thing from the "eco" point of view.
I volunteer one day a week for a non-profit organization in Kansas City (www.connectingforgood.org). This group accepts donations from anybody and refurbishes the computers that are suitable. If a desktop has a CPU > Pentium 4, we ensure that it has at least 2 GB RAM and an 80 GB hard drive. After installation of Windows 7 Pro, the machine is sold to low-income people for $75. That includes everything but speakers. Those are rarely donated. If the user wants a full Office license, that is an additional $20. Everything has a one-year, no questions asked, warranty. All computer equipment that is unsuitable for refurbishing is properly recycled so that very little ends up in a landfill. New may be more economical to operate, but the price may be too high for some users. Larry -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thursday 07 of May 2015 11:19:09 Larry Finger wrote:
New may be more economical to operate, but the price may be too high for some users.
My understanding from the start was that the term "econeg" Felix used meant ecology point of view, not economy. Michal Kubecek -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, 07 May 2015, 18:02:19 +0200, Felix Miata wrote:
Richard Brown composed on 2015-05-07 13:21 (UTC+0200):
Felix Miata wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
"ip a" is short for "ip addr" which is in the man page
Where in the man page exactly? If it was apparent I wouldn't have had reason to ask.
Quoted from "man 8 ip": The names of all objects may be written in full or abbreviated form, for example address can be abbreviated as addr or just a. And, for further information try "man 8 ip-address" (from SEE ALSO). HTH, cheers. l8er manfred
Petr Gajdos composed on 2015-05-07 13:48 (UTC+0200):
On Thu, May 07, 2015 at 07:00:35AM -0400, Felix Miata wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
man 8 ip:
"The names of all objects may be written in full or abbreviated form, for exampe address can be abbreviated as addr or just a."
That doesn't look anything like a translation of 'ifconfig -a(ll)' to me. It's like most man pages, void of examples that give some meaning to the list of possible options, and reduce need to Google for a usable explanation or HOWTO. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, 7 May 2015 17:57, Michal Kubecek <mkubecek@...> wrote:
On Thursday 07 of May 2015 17:48:27 Per Jessen wrote:
Michal Kubecek wrote:
Recent versions also support "ip -s addr show" so you can have both addresses and statistics at once.
Must be more like brand new versions - I don't get any stats when I tried it on 13.2.
Just checked the git: should be there since iproute2 3.17.
FYI: Versions of iproute2 in openSUSE: OSS 13.2 offical update: 3.16, thats a no for plain 13.2 + upd. OSS 13.1 offical release: 3.9.0 Michal Kubecek has 3.19 in his "home" repo [1] Factory is on version 4.0, from "security:netfilter" [2] repo. Version 4.0 is available there for Factory, 13.2, 13.1, SLE-12. [1] http://download.opensuse.org/repositories/home:/mkubecek:/utils/openSUSE_13.... [2] http://download.opensuse.org/repositories/security:/netfilter/openSUSE_13.1/ - Yamaban. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 07.05.2015 um 16:12 schrieb Stephan Kulow:
On 07.05.2015 14:49, Stefan Seyfried wrote:
now if we only could also do away the same with that ppc64le stuff that's polluting my obs mailbox with crazy buildfailures all the time... :-)
Is it really that bad? I don't see too any ppc64le specific build failures in stagings at least.
No, it's always the same multimedia packages that probably nobody is ever going to use on such niche hardware anyway :-) -- Stefan Seyfried "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Richard Feynman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 7, 2015 at 12:44 PM, Per Jessen <per@computer.org> wrote:
Completely agree, but I have yet to try that out, one of these days I will. I can't help wondering if there are any issues in 32bit apps communicatiung with 64bit apps? E.g. a 32bit postfix talking to a 64bit mysql?
Network protocols are architecture-independent unless they're broken. MySQL uses a network protocol. Other things might break (shmem stuff mostly), but certainly not that. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 7, 2015 at 1:46 PM, Felix Miata <mrmazda@earthlink.net> wrote:
Petr Gajdos composed on 2015-05-07 13:48 (UTC+0200):
On Thu, May 07, 2015 at 07:00:35AM -0400, Felix Miata wrote:
Nothing in the entirety of the usage message says anything about "a", same omission as the man page. Mere mortals don't all know what man page writers and programmers assume they know.
man 8 ip:
"The names of all objects may be written in full or abbreviated form, for exampe address can be abbreviated as addr or just a."
That doesn't look anything like a translation of 'ifconfig -a(ll)' to me. It's like most man pages, void of examples that give some meaning to the list of possible options, and reduce need to Google for a usable explanation or HOWTO.
This is because this tool is terminally user-unfriendly..and just like ifconfig its interface cannot be changed or removed or the whining stream will overflow the pacific ocean. It is best to work on something that makes it unlikely you will ever need to use it. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, May 7, 2015 at 12:44 PM, Per Jessen <per@computer.org> wrote:
Jan Engelhardt wrote:
On Thursday 2015-05-07 17:22, Per Jessen wrote:
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
A 32-bit system can run far more postfix and dovecot processes than a 64bit with the same amount of memory.
If you want to run that many postfixes, you would rather want to use a X32 userspace, and not a i586 one.
Completely agree, but I have yet to try that out, one of these days I will. I can't help wondering if there are any issues in 32bit apps communicatiung with 64bit apps? E.g. a 32bit postfix talking to a 64bit mysql?
Unless one of those components is doing something terrible wrong, there should be no problem.. in your example, a postfix plugin is probably sending SQL queries to the database using the mysql client library. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Michal Kubecek wrote:
On Thursday 07 of May 2015 17:33:05 Per Jessen wrote:
Michal Kubecek wrote:
On Wed, May 06, 2015 at 02:06:20PM -0400, Robert Schweikert wrote:
Well, I am not convinced about this. Especially that SLE no longer has an x86 implementation.
I'm going to say something that will be very unpopular but I feel it really needs to be said. Should we really care so much about i586? I don't have any 32-bit system since something like 2008, I definitely haven't seen any 32-bit x86 CPU in usual e-shops for at least 5 years and I'm not sure I would be able to buy one today if I tried hard.
Specifically, what would be gained by dropping the i586 build? I still have use-cases for 32bit, but I could just remain on the last release with support.
I already tried to explain it once but, OK, again: I did _not_ propose to drop i586 build. I just wanted to say that the fact that SLE is not released for and supported on i586 shouldn't be seen as an argument against basing next (or any future) openSUSE release on SLE packages.
Apologies, I misunderstood. -- Per Jessen, Zürich (13.6°C) http://www.hostsuisse.com/ - dedicated server rental in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 07.05.2015 20:30, Stefan Seyfried wrote:
Am 07.05.2015 um 16:12 schrieb Stephan Kulow:
On 07.05.2015 14:49, Stefan Seyfried wrote:
now if we only could also do away the same with that ppc64le stuff that's polluting my obs mailbox with crazy buildfailures all the time... :-)
Is it really that bad? I don't see too any ppc64le specific build failures in stagings at least.
No, it's always the same multimedia packages that probably nobody is ever going to use on such niche hardware anyway :-)
Snipe them out with a well placed ExclusiveArch :) Greetings, Stephan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlVMeF4ACgkQwFSBhlBjoJZR9ACdFXjRjsKJi3NtnpXb3OCAw7GQ 0usAn0zJ6dBoKigOAOw+fg5a9YAl5bGn =zshn -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Onsdag den 6. maj 2015 20:53:10 skrev Michal Kubecek:
I don't think marketing is the problem. The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive.
Richard himself showed about 11.000 Tumbleweed users, and ~150.000 13.2 downloads in the presentation. The presentation even showed growth for 13.2 compared to 13.1.
The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him.
Which group of users is this? The people running CentOS and Debian Stable today? I.e. a few server users who don't need contemporary applications and hardware support (I'm still not convinced about the SLE base supporting recent consumer desktop hardware, especially not 2-4 years from now). The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months). I really doubt too many users will be willing to trade package availability, up-to-date applications and hardware support for the bit of extra stability from the SLE base. Particularly if the difference in lifetime is not huge (26 months vs. "at least 36 months")
"The 13.2 release required a large amount of work from a small number of people to make it a success."
will not be resolved just because we pick a different base.
I disagree. People have free will, they won't come just because we need them. They will if we do something to attract them. I believe offering a solution in the area where we have been failing to provide one for years is more likely to attract users and contributors than offering two versions of a distribution that are not too far from each other.
I doubt you can attract packagers to work on this thing. Not Tumbleweed packagers. In my experience even Debian packagers aren't to enthusiastic about working on Debian Stable or backporting stuff to such an old base. And you also have to take into account the less technical contributions of marketing, artwork, translations, testing, writing howtos, supporting people on forums and irc etc. (this group includes me). I just can't see the attraction. As far as I can tell it'll be like Mint/Ubuntu LTS, only more outdated, and with similar or less lifetime or like CentOS, with some more packages available, but with much shorter lifetime. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Saturday 09 May 2015 14:42:08 Martin Schlander wrote:
The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months).
In my opinion we are all making assumptions here without actually knowing what is required by the group of users (Future and Current), that is not attracted to the openSUSE Tumbleweed releases. I can imagine that there might be a group of users, that wants to have a stable release that is support for a certain period, but I wonder like Martin if they would opt-in for a 3-year release cycle.
I doubt you can attract packagers to work on this thing. Not Tumbleweed packagers. In my experience even Debian packagers aren't to enthusiastic about working on Debian Stable or backporting stuff to such an old base.
I guess it depends on the amount of work involved. If I look at the situation with KDE, then we have here two parallel processes for openSUSE releases. The first process was to update the release with the minor KDE releases as that these would not cause any dependency issues and the second process was to provide a separate OBS project for newer KDE releases for supported openSUSE releases. We needed this separate project as that the experience learned us that with each newer KDE release newer versions of dependent packages were required. If I look now at SLE12, then we wouldn't be able to build Plasma 5 on top of it as that SLE12 ships Qt 5.3.1 and Plasma 5 needs Qt 5.4. If SLE12 now would be chosen as the base for the next openSUSE release, then it would mean that we release a new product with "old" packages. Maybe preferred by some users, but I guess that most users still would like to have something newer. Why otherwise to upgrade from 13.2, if you get nothing newer in return. just my five cents. Raymond -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/09/2015 03:42 PM, Martin Schlander wrote:
Onsdag den 6. maj 2015 20:53:10 skrev Michal Kubecek:
I don't think marketing is the problem. The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. Richard himself showed about 11.000 Tumbleweed users, and ~150.000 13.2 downloads in the presentation. The presentation even showed growth for 13.2 compared to 13.1.
The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him. Which group of users is this? The people running CentOS and Debian Stable today? I.e. a few server users who don't need contemporary applications and hardware support (I'm still not convinced about the SLE base supporting recent consumer desktop hardware, especially not 2-4 years from now).
The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months).
I really doubt too many users will be willing to trade package availability, up-to-date applications and hardware support for the bit of extra stability from the SLE base. Particularly if the difference in lifetime is not huge (26 months vs. "at least 36 months")
"The 13.2 release required a large amount of work from a small number of people to make it a success."
will not be resolved just because we pick a different base. I disagree. People have free will, they won't come just because we need them. They will if we do something to attract them. I believe offering a solution in the area where we have been failing to provide one for years is more likely to attract users and contributors than offering two versions of a distribution that are not too far from each other. I doubt you can attract packagers to work on this thing. Not Tumbleweed packagers. In my experience even Debian packagers aren't to enthusiastic about working on Debian Stable or backporting stuff to such an old base.
And you also have to take into account the less technical contributions of marketing, artwork, translations, testing, writing howtos, supporting people on forums and irc etc. (this group includes me). I just can't see the attraction.
As far as I can tell it'll be like Mint/Ubuntu LTS, only more outdated, and with similar or less lifetime or like CentOS, with some more packages available, but with much shorter lifetime. Excellent points Martin,
I also feel that expanding the gap between Tumbleweed packages and stable release, will make users move away from openSUSE. In the end it has to do with the project's vision: Do we want to be a stable alternative to Debian and CentOS for server usage (as SLE is) or do we want to be something stable but also appealing to the Desktop users, who are willing to upgrade package versions once a year. Hitting both targets with one distribution is a long shot. And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market. Best, Angelos -- Angelos Tzotsos Remote Sensing Laboratory National Technical University of Athens http://users.ntua.gr/tzotsos -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi, Am 09.05.2015 um 17:56 schrieb Angelos Tzotsos:
I also feel that expanding the gap between Tumbleweed packages and stable release, will make users move away from openSUSE.
In the end it has to do with the project's vision: Do we want to be a stable alternative to Debian and CentOS for server usage (as SLE is) or do we want to be something stable but also appealing to the Desktop users, who are willing to upgrade package versions once a year. Hitting both targets with one distribution is a long shot. And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market.
In a perfect world we probably want three distributions: Tumbleweed openSUSE as of today/yesterday openSUSE "LTS" At least I would want this (I could live w/o Tumbleweed actually). But this is totally out of the scope of we can maintain at the moment. (Actually not that totally because with Evergreen we have such a thing but it suffers from too few contributors.) Now the question really comes down to "closer to our current users". As far as I can remember we had stats that a _lot_ of people are still on old outdated, unmaintained openSUSE versions. These users are just not reflected here on opensuse-_factory_. Sorry, I'm too lazy now to dig up presentations or list posts having some stats on that. Wolfgang -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 2015-05-09 14:42, Martin Schlander wrote:
Which group of users is this? The people running CentOS and Debian Stable today? I.e. a few server users who don't need contemporary applications and hardware support (I'm still not convinced about the SLE base supporting recent consumer desktop hardware, especially not 2-4 years from now).
The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months).
Yes. True.
I really doubt too many users will be willing to trade package availability, up-to-date applications and hardware support for the bit of extra stability from the SLE base. Particularly if the difference in lifetime is not huge (26 months vs. "at least 36 months")
Yes.
I doubt you can attract packagers to work on this thing. Not Tumbleweed packagers. In my experience even Debian packagers aren't to enthusiastic about working on Debian Stable or backporting stuff to such an old base.
And you also have to take into account the less technical contributions of marketing, artwork, translations, testing, writing howtos, supporting people on forums and irc etc. (this group includes me). I just can't see the attraction.
As a translator, I don't touch tumbleweed. I don't know what will happen with the translation, but one thing we need is a string freeze period in which we do most of the translation. It just doesn't work if the developers change things a day before or after we translated them: they break. As for supporting people on forums or lists, we need time to learn new features (developers don't usually do the supporting, and not many of them write good docs). I don't see this happening with tumbleweed, specially with novices. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iEYEARECAAYFAlVORO8ACgkQtTMYHG2NR9WoSgCdHe6sUW7OySVh1ylkf70qUDNV D5MAn15TTwn3nV2MIHM8Jz7cbOS2JX6k =hc1v -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 2015-05-09 15:42, Raymond Wooninck wrote:
On Saturday 09 May 2015 14:42:08 Martin Schlander wrote:
In my opinion we are all making assumptions here without actually knowing what is required by the group of users (Future and Current), that is not attracted to the openSUSE Tumbleweed releases.
I can imagine that there might be a group of users, that wants to have a stable release that is support for a certain period, but I wonder like Martin if they would opt-in for a 3-year release cycle.
Not me. I might do it on my server, but not on my desktop machines, unless there is some problem impeding me upgrade.
If SLE12 now would be chosen as the base for the next openSUSE release, then it would mean that we release a new product with "old" packages. Maybe preferred by some users, but I guess that most users still would like to have something newer. Why otherwise to upgrade from 13.2, if you get nothing newer in return.
That's right... - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iEYEARECAAYFAlVORecACgkQtTMYHG2NR9XIjgCfepzPswb4bVh0cs5zWDz9MSrp eDkAn0LbnTz2RQshLm0r+M2GAoOkDcFh =WMoF -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Sat, May 9, 2015 at 1:10 PM, Wolfgang Rosenauer <wolfgang@rosenauer.org> wrote:
Hi,
Am 09.05.2015 um 17:56 schrieb Angelos Tzotsos:
I also feel that expanding the gap between Tumbleweed packages and stable release, will make users move away from openSUSE.
In the end it has to do with the project's vision: Do we want to be a stable alternative to Debian and CentOS for server usage (as SLE is) or do we want to be something stable but also appealing to the Desktop users, who are willing to upgrade package versions once a year. Hitting both targets with one distribution is a long shot. And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market.
In a perfect world we probably want three distributions: Tumbleweed openSUSE as of today/yesterday openSUSE "LTS"
At least I would want this (I could live w/o Tumbleweed actually).
But this is totally out of the scope of we can maintain at the moment. (Actually not that totally because with Evergreen we have such a thing but it suffers from too few contributors.)
Now the question really comes down to "closer to our current users". As far as I can remember we had stats that a _lot_ of people are still on old outdated, unmaintained openSUSE versions.
https://lizards.opensuse.org/wp-content/uploads/2013/08/uuid_month.png From: https://lizards.opensuse.org/2013/08/23/more-on-statistics/ Unfortunately they are a couple years old. And yes they are surprising. Greg -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 9 May 2015 at 14:42, Martin Schlander <martin.schlander@gmail.com> wrote:
Richard himself showed about 11.000 Tumbleweed users, and ~150.000 13.2 downloads in the presentation. The presentation even showed growth for 13.2 compared to 13.1.
Users, yes. But the graph for 'contributors to the Regular Release' is a very different story. and its the desire and momentum of contributors actually contributing to getting a Regular Release out that we need to consider when we talk about doing one.
Which group of users is this? The people running CentOS and Debian Stable today? I.e. a few server users who don't need contemporary applications and hardware support (I'm still not convinced about the SLE base supporting recent consumer desktop hardware, especially not 2-4 years from now).
The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months).
I really doubt too many users will be willing to trade package availability, up-to-date applications and hardware support for the bit of extra stability from the SLE base. Particularly if the difference in lifetime is not huge (26 months vs. "at least 36 months")
The majority of our user base - look at Alberto's presentation from oSC 13 that shows the actual user numbers of our repositories, and how lots of people keep on older versions for a lot longer than we say they should https://www.youtube.com/watch?v=NwfohZ8RBd8
I doubt you can attract packagers to work on this thing. Not Tumbleweed packagers. In my experience even Debian packagers aren't to enthusiastic about working on Debian Stable or backporting stuff to such an old base.
I think you're wrong here.. I was having some very interesting conversations with the KDE team last night, and they were coming up with all kind of ideas of doing very exciting things ontop of this. We have the tools, talent, and expetise to do things that no other distribution can even consider - openSUSE-based-on-SLE with official 'add-on modules' for Desktops for example (eg. KDE:Current, or GNOME:Stable)? So the desktops can move at a pace that suits the community, while the core doesn't change? These are exciting ideas which I think might work to address your doubts and turn this idea for a Regular Release into something exceptional - We need to avoid falling into the trap of thinking of this 'like a Debian' or 'like a CentOS' - We're openSUSE, let's build something which only we can do. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 9 May 2015 at 15:42, Raymond Wooninck <tittiatcoke@gmail.com> wrote:
Why otherwise to upgrade from 13.2, if you get nothing newer in return.
My longer answer to this is in a thread in -project. But my short answer is "because it works" People don't just upgrade because it's new and shiny, some people upgrade because something works better than it used to, even if its ultimately the same version underneath. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 9 May 2015 at 19:10, Wolfgang Rosenauer <wolfgang@rosenauer.org> wrote:
Hi,
Am 09.05.2015 um 17:56 schrieb Angelos Tzotsos:
And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market.
In a perfect world we probably want three distributions:
We have the capacity to maintain two distributions 1 Tumbleweed Rolling Distribution 1 openSUSE Stable Distribution However, we are very limited on the number of people actually working on the Stable distribution, and so I think embrasing the opportunity to use the SLE sources for all their worth will do a great deal of good to get new people involved there, while simultaniously reducing the work required as we'd be benefiting from the work already done by SUSE.
Now the question really comes down to "closer to our current users". As far as I can remember we had stats that a _lot_ of people are still on old outdated, unmaintained openSUSE versions. These users are just not reflected here on opensuse-_factory_. Sorry, I'm too lazy now to dig up presentations or list posts having some stats on that.
https://www.youtube.com/watch?v=NwfohZ8RBd8 Here they are :) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
In a perfect world we probably want three distributions: We have the capacity to maintain two distributions 1 Tumbleweed Rolling Distribution 1 openSUSE Stable Distribution
However, we are very limited on the number of people actually working on the Stable distribution, and so I think embrasing the opportunity to use the SLE sources for all their worth will do a great deal of good to get new people involved there, while simultaniously reducing the work required as we'd be benefiting from the work already done by SUSE. As discussed yesterday, if we are using the tools correctly, then we could
On Saturday 09 May 2015 19:50:06 Richard Brown wrote: maintain the two distributions. But that would also mean that in my opinion we need to split the distribution in multiple rings/modules. E.g. We were delivering newer version for the KDE 4 desktop environment by utilizing the power of OBS and created a separate repository that was building for multiple openSUSE releases (KDE:Current). Utilizing this was beneficial for a lot of users, but also for the team itself as that we only needed to maintain a single set of packages. However with this approach we always had to be careful that a newer version could also mean newer version for buildrequirements. Raymond -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/09/2015 09:42 AM, Raymond Wooninck wrote:
On Saturday 09 May 2015 14:42:08 Martin Schlander wrote:
The main complaint about the current openSUSE releases from the "stability crowd", is the lifetime in my opinion, and that would largely adressed by going to a 12 month release cycle, equaling 26 month lifetime (2 releases+2 months).
In my opinion we are all making assumptions here without actually knowing what is required by the group of users (Future and Current), that is not attracted to the openSUSE Tumbleweed releases.
Well if you consider only users that are not currently on this list and not users of openSUSE I would tend to agree with your statement. However, I think in this thread we have seen statements from 3 distinct groups. From my point of view there is no reason not to believe that each of these groups would not be represented within the group of people not participating in the discussion. - - We have heard from the Tumbleweed users and those that are considering potential use under certain circumstances. - - We have heard from those that like the current snapshot based releases, ~12 months cycle. Primary statements revolve around reasonable stability, not too far behind upstream, reasonably current HW support etc. - - We have heard from those that like longer cycles and more stability. These statements often focus on production use and server deployments. I think each of the voices heard is a representation of the separation of users. We can argue about percentages and how this would divide and we'd probably never reach agreement. But I think we can probably all see these 3 distinct tendencies.
I can imagine that there might be a group of users, that wants to have a stable release that is support for a certain period, but I wonder like Martin if they would opt-in for a 3-year release cycle.
Yes, from those big three categories we can probably come up with an infinite number of further distinction. Later, Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVT03CAAoJEE4FgL32d2UkqpEIAL20+mOaN6p9QLptjFW6Ilm0 1bqZ6hVuzemkpbKA7UzCVq+GkRuD1CBde/0YkHm9nsa1LIQ00+zXz2mUUDdXFzLG oTLnh/hD5u34RQ0ZpQ2ASJPIHwkGGZNFMfAvDfqz3ZcPabGJSjcr6CdQQii6Zkrm hmOnV8IIHbNEQCoJC6Qk875IISo/KR/tsKKUdO0xXND9GOLm1c9ttr9UC958YB+T OHz8hRDAgWi5d/9Qoywp6zjrYJIZbQmpA/UqMA2UZjjFtw4N2YTOWVykhP7bD8qW kqDZmN9qhSg7GCK/huRSCm5DN7Fz9k//2nX/N3yhlosKbXr/9zU6dBV0hnsDlzk= =beqP -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 05/09/2015 01:10 PM, Wolfgang Rosenauer wrote:
Hi,
Am 09.05.2015 um 17:56 schrieb Angelos Tzotsos:
I also feel that expanding the gap between Tumbleweed packages and stable release, will make users move away from openSUSE.
In the end it has to do with the project's vision: Do we want to be a stable alternative to Debian and CentOS for server usage (as SLE is) or do we want to be something stable but also appealing to the Desktop users, who are willing to upgrade package versions once a year. Hitting both targets with one distribution is a long shot. And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market.
In a perfect world we probably want three distributions: Tumbleweed openSUSE as of today/yesterday openSUSE "LTS"
At least I would want this (I could live w/o Tumbleweed actually).
But this is totally out of the scope of we can maintain at the moment. (Actually not that totally because with Evergreen we have such a thing but it suffers from too few contributors.)
Agreed. Robert - -- Robert Schweikert MAY THE SOURCE BE WITH YOU Public Cloud Architect LINUX rjschwei@suse.com IRC: robjo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAEBAgAGBQJVT04CAAoJEE4FgL32d2Ukvw4H/3jBB4vekSCvH0A/tT7sFuVR OGykfXtmGdRW4MFRxDH6/annIk3tbZfZcRQ6fXveO21HuZfZs3QCXnX3HWagw3pr 5Rbpo0ljK4CQzi5aO5YbMwWS2oh0t/G/0NlM4mHys1Np8pFXHtqpgO82BjXf/atx WjZGI+EnsDOaewLTqwimr3Bc7aSXL3WZjf1izS2GTSEQFLmB7WY/8uTABESQnPTM 3FzepWbvFPqsgA8gz1ZyL2cP6RL9mjrRevv08eGD7HWhL7I6qgSP8lNnWeW+25qE IkeTZXnlv0bAZf+qNioFFipVObeP/h4BWqnTsGxy187SfIzrU4EX7ahbRULipe0= =XF5h -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
* Wolfgang Rosenauer <wolfgang@rosenauer.org> [2015-05-09 19:11]:
Hi,
Am 09.05.2015 um 17:56 schrieb Angelos Tzotsos:
I also feel that expanding the gap between Tumbleweed packages and stable release, will make users move away from openSUSE.
In the end it has to do with the project's vision: Do we want to be a stable alternative to Debian and CentOS for server usage (as SLE is) or do we want to be something stable but also appealing to the Desktop users, who are willing to upgrade package versions once a year. Hitting both targets with one distribution is a long shot. And if we don't have the resources to maintain 2 distributions (SLE-based and Factory-based), lets chose one closer to our current users and not enter the server market.
In a perfect world we probably want three distributions: Tumbleweed openSUSE as of today/yesterday openSUSE "LTS"
At least I would want this (I could live w/o Tumbleweed actually).
But this is totally out of the scope of we can maintain at the moment. (Actually not that totally because with Evergreen we have such a thing but it suffers from too few contributors.)
Actually it should be possible if Evergreen were to be based on the SLE sources, in combination with the new SLE backports that should eventually cover the LTS server use case and be less work than today since you wouldn't have to monitor and backport security fixes yourself any more. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 09.05.2015 um 19:45 schrieb Richard Brown:
We're openSUSE, let's build something which only we can do. This should be a slogan for openSUSE T-Shirts. ;-)
Stefan -- www.invis-server.org Stefan Schäfer Ludwigstr. 1-3 63679 Schotten -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Mon, May 11, 2015 at 2:10 AM, Stefan Schäfer <ml@fsproductions.de> wrote:
Am 09.05.2015 um 19:45 schrieb Richard Brown:
We're openSUSE, let's build something which only we can do.
This should be a slogan for openSUSE T-Shirts. ;-)
I'm totally onboard with that. Actually, I'd love to see some openSUSE polos that I could purchase and then wear to work.
Stefan
-- www.invis-server.org
Stefan Schäfer Ludwigstr. 1-3 63679 Schotten
-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-- Sincerely, Bob Martens Webmaster/Technician Martin Luther College http://mlc-wels.edu -- This electronic communication, including any attached documents, may contain confidential and/or legally privileged information that is intended only for use by the recipient(s) named above. If you have received this communication in error, please notify the sender immediately and delete the communication and any attachments. Views expressed by the author do not necessarily represent those of Martin Luther College. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Robert Martens wrote:
On Mon, May 11, 2015 at 2:10 AM, Stefan Schäfer <ml@fsproductions.de> wrote:
Am 09.05.2015 um 19:45 schrieb Richard Brown:
We're openSUSE, let's build something which only we can do.
This should be a slogan for openSUSE T-Shirts. ;-)
I'm totally onboard with that. Actually, I'd love to see some openSUSE polos that I could purchase and then wear to work.
I saw these guys mentioned recently: https://www.hellotux.com/opensuse -- Per Jessen, Zürich (19.3°C) http://www.hostsuisse.com/ - virtual servers, made in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Per Jessen wrote:
Robert Martens wrote:
On Mon, May 11, 2015 at 2:10 AM, Stefan Schäfer <ml@fsproductions.de> wrote:
Am 09.05.2015 um 19:45 schrieb Richard Brown:
We're openSUSE, let's build something which only we can do.
This should be a slogan for openSUSE T-Shirts. ;-)
I'm totally onboard with that. Actually, I'd love to see some openSUSE polos that I could purchase and then wear to work.
I saw these guys mentioned recently:
Plenty of them: http://shop.opensuse.org/ http://www.amazon.de/Fruit-Loom-Polo-Shirt-openSUSE-Schriftzug/dp/B003FYCAL0 http://www.linux-onlineshop.de/Fanartikel/Shirts-/-Kleidung/Polo-Shirts/Dist... -- Per Jessen, Zürich (19.7°C) http://www.hostsuisse.com/ - dedicated server rental in Switzerland. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, 2015-05-06 at 15:36 +0200, Frederic Crozat wrote:
Le mercredi 06 mai 2015 à 08:29 -0500, Malcolm a écrit :
On Wed 06 May 2015 01:45:06 PM CDT, Richard Brown wrote:
<snip>
You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
<snip>
Hi The only thing I've had to compile for the kernel was the it87 module for SLES 12, it is present for SLED 12...
This means it is part of kernel-*-extra package (which is only available on SLED and not on SLES).
My proposal would be for an openSUSE kernel based on SLE kernel, we could have a single kernel package which would include the content on kernel + kernel-extra (if kernel team is fine with that)..
Converging would be nice, but I had the impression that (at least with sles) the kernel is lagging very much? I remember once hearing (could be mistaken) that @maxfeld, they had to backport all sorts of recent kernel-bugfixes to ancient sles-kernels.... hw -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Le mercredi 13 mai 2015 à 08:55 +0200, Hans Witvliet a écrit :
On Wed, 2015-05-06 at 15:36 +0200, Frederic Crozat wrote:
Le mercredi 06 mai 2015 à 08:29 -0500, Malcolm a écrit :
On Wed 06 May 2015 01:45:06 PM CDT, Richard Brown wrote:
<snip>
You also have to consider that SUSE is currently considering the possibility of not backporting fixes/features but actually changing the Kernel version they use in SLE 12 SP2 (assuming they can maintain API/ABI stability of course), so in our case with openSUSE we'd only have this kernel for about a year.. well maintained, with hardware enablement, and proper security patches.. sure if people can raise actual legitimate problems, I'm happy to hear them but I would prefer if the discussion continued on the basis of facts, not peoples theories about what does or doesn't work in the SLE codebase we now get to play with ;)
<snip>
Hi The only thing I've had to compile for the kernel was the it87 module for SLES 12, it is present for SLED 12...
This means it is part of kernel-*-extra package (which is only available on SLED and not on SLES).
My proposal would be for an openSUSE kernel based on SLE kernel, we could have a single kernel package which would include the content on kernel + kernel-extra (if kernel team is fine with that)..
Converging would be nice, but I had the impression that (at least with sles) the kernel is lagging very much?
I remember once hearing (could be mistaken) that @maxfeld, they had to backport all sorts of recent kernel-bugfixes to ancient sles-kernels....
Again, it depends on your vision of "lagging" ;) Of course, sles kernel are getting a lot of backports (for instance, the btrfs stack in SLES kernel is way more recent than the one from vanilla 3.12) but OTOH, each new upstream kernel is introducing regressions, which are not always fixed (I remember reading about this on LWN recently but couldn't find the url). The point of using a "SLE" kernel would be to leverage work being done by SUSE kernel team for SLE in "less regression caused by new kernel" but still getting some new features. And kernel version updates are scheduled for some Service Packs in SLE (not just every service pack, unlike what we are currently doing on openSUSE regular release).. -- Frederic Crozat Enterprise Desktop Release Manager SUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 09.05.15 Richard Brown wrote:
We have the capacity to maintain two distributions 1 Tumbleweed Rolling Distribution 1 openSUSE Stable Distribution
Some people mentioned Evergreen. Why not think about using the man power going into evergreen (that AFAIK is a user driven group effort) for this LTS release. But make it 'more official'? Note: I'm not sure what the evergreen group think about this. Johannes -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with SeaMonkey - http://www.enigmail.net/ iEYEARECAAYFAlVTalIACgkQzi3gQ/xETbKTzgCfWHbLU4WMM9UL3OlMUA75pxZA Yp4AnR7QxCKdyXGU4H0dbshvlqmVzRkN =R1bF -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Wed, May 13, 2015 at 11:14 AM, Johannes Kastl <mail@ojkastl.de> wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 09.05.15 Richard Brown wrote:
We have the capacity to maintain two distributions 1 Tumbleweed Rolling Distribution 1 openSUSE Stable Distribution
Some people mentioned Evergreen. Why not think about using the man power going into evergreen (that AFAIK is a user driven group effort) for this LTS release. But make it 'more official'?
Note: I'm not sure what the evergreen group think about this.
Johannes
The "Evergreen group" is 2 or 3 dedicated souls. For userspace they basically watch the security patches coming out for officially maintained releases and backport them if appropriate. For the kernel they have migrated the evergreen release to the closest SLE kernel which is newer than the original evergreen kernel and migrated to it. Greg -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 06.05.15 Michal Kubecek wrote:
I don't think marketing is the problem.
Communicating the openSUSE way to the public surely is.
The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him.
As an end user, running openSUSE Tumbleweed on my work laptop and 'fool around' machines and running 13.1 in >10 VMs all over the place I agree. I would suggest the following approach for the future: - ---------- Citation ----------- Tumbleweed is for people with new laptops. Drivers have been mentioned . Releases are for long running machines, servers, etc. - ---------- Citation ----------- If I can run my servers on openSUSE for more than the lifetime releases had up til now, I would get rid of having to upgrade 'frequently'. And VMs and servers rarely need the latest software (drivers might be an issue with new servers and new hardware).
It would be even nicer if we could do three distributions: Tumbleweed, something similar to openSUSE <= 13.1 and some "LTS" version (possibly SLE based). But I'm afraid we don't have resources to do all three and there is no easy way to change that fact in short timeframe.
I would rather pick two flavors and do them well. Johannes -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with SeaMonkey - http://www.enigmail.net/ iEYEARECAAYFAlVTpbYACgkQzi3gQ/xETbK/uQCeOiWRskDXIEpfx12TaTFME5sx IX4An1Soce+CTwDHJ8WsgxdZd92PAhNy =YL1I -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Hi, Am 13.05.2015 um 17:14 schrieb Johannes Kastl:
On 09.05.15 Richard Brown wrote:
We have the capacity to maintain two distributions 1 Tumbleweed Rolling Distribution 1 openSUSE Stable Distribution
Some people mentioned Evergreen. Why not think about using the man power going into evergreen (that AFAIK is a user driven group effort) for this LTS release. But make it 'more official'?
as Greg (and myself earlier) pointed out, Evergreen has very limited manpower. But indeed we are watching closely (and I'm participating in) the discussion since the outcome will directly affect the future of Evergreen. As we are all volunteers and have our own requirements currently bundled in Evergreen we most likely will join forces for creating a different kind of openSUSE _if_ it fulfills our needs. Richard's proposal comes pretty close to mine at least. A lot of details to be sorted out though. Wolfgang -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Am 13.05.2015 um 21:27 schrieb Johannes Kastl:
The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him. As an end user, running openSUSE Tumbleweed on my work laptop and 'fool around' machines and running 13.1 in >10 VMs all over the place I agree.
I would suggest the following approach for the future:
- ---------- Citation ----------- Tumbleweed is for people with new laptops. Drivers have been mentioned .
Releases are for long running machines, servers, etc. - ---------- Citation -----------
If I can run my servers on openSUSE for more than the lifetime releases had up til now, I would get rid of having to upgrade 'frequently'. And VMs and servers rarely need the latest software (drivers might be an issue with new servers and new hardware).
100% agreement! Stefan -- www.invis-server.org Stefan Schäfer Ludwigstr. 1-3 63679 Schotten -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 13.05.15 Wolfgang Rosenauer wrote:
As we are all volunteers and have our own requirements currently bundled in Evergreen we most likely will join forces for creating a different kind of openSUSE _if_ it fulfills our needs. Richard's proposal comes pretty close to mine at least. A lot of details to be sorted out though.
This sounds to me like 'the evergreen maintainers' are not afraid that their baby might be taken away. If the requirements match, that is. Thanks for your answers, guys. Johannes -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with SeaMonkey - http://www.enigmail.net/ iEYEARECAAYFAlVU5K4ACgkQzi3gQ/xETbJxIwCdG/DJqnJRz6woAjB313xOSLWg drAAoIKt12S3hr2ElWjW4ZsiSUlsr/86 =+ZCR -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-05-09 19:45, Richard Brown wrote:
On 9 May 2015 at 14:42, Martin Schlander <martin.schlander@gmail.com> wrote:
Richard himself showed about 11.000 Tumbleweed users, and ~150.000 13.2 downloads in the presentation. The presentation even showed growth for 13.2 compared to 13.1.
Users, yes. But the graph for 'contributors to the Regular Release' is a very different story. and its the desire and momentum of contributors actually contributing to getting a Regular Release out that we need to consider when we talk about doing one.
In that case, openSUSE would be a distribution for developers only, no users. For and by devs. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVZPm8ACgkQja8UbcUWM1wfGAD+N2qDulKA+YbSJgFZu0PyV0Kh e/MRgDyrdp9KSgwi6Q8A/ReE0QcoQBrtbFAS2yozhtsAJYUimGEr1yaNSxNzccuy =2OnM -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-05-14 08:34, Stefan Schäfer wrote:
Am 13.05.2015 um 21:27 schrieb Johannes Kastl:
The problem is that with the change of approach to Tumbleweed and Factory, current Tumbleweed attracted substantial part of those openSUSE users who focus on new versions and new features. As a result of these changes, 13.2 lost big part of its "added value" and became less attractive. The way I understand Richard's arguments, he believes one way to regain some of the sense to the releases is to move them more away from the Tumbleweed so that they could target a different group of users, in particular group openSUSE failed to target for long time. And I agree with him. As an end user, running openSUSE Tumbleweed on my work laptop and 'fool around' machines and running 13.1 in >10 VMs all over the place I agree.
I would suggest the following approach for the future:
- ---------- Citation ----------- Tumbleweed is for people with new laptops. Drivers have been mentioned .
Releases are for long running machines, servers, etc. - ---------- Citation -----------
If I can run my servers on openSUSE for more than the lifetime releases had up til now, I would get rid of having to upgrade 'frequently'. And VMs and servers rarely need the latest software (drivers might be an issue with new servers and new hardware).
100% agreement!
You forget the middle ground. Plain desktop/laptop users, wanting reasonably recent software, just not wanting to work the installation or upgrade every some months. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVZQKoACgkQja8UbcUWM1y2tQD+IN+XR54kutp+eQOn2TUyVK7D G+AOUBfQevKBt11a9QcBAIq9Lc7s8ukKsuJOLrTHGjmpNLVtbXyPQtLm6wu7p9ld =4Q8U -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 2015-05-10 14:23, Robert Schweikert wrote:
Well if you consider only users that are not currently on this list and not users of openSUSE I would tend to agree with your statement. However, I think in this thread we have seen statements from 3 distinct groups. From my point of view there is no reason not to believe that each of these groups would not be represented within the group of people not participating in the discussion.
- We have heard from the Tumbleweed users and those that are considering potential use under certain circumstances.
- We have heard from those that like the current snapshot based releases, ~12 months cycle. Primary statements revolve around reasonable stability, not too far behind upstream, reasonably current HW support etc.
- We have heard from those that like longer cycles and more stability. These statements often focus on production use and server deployments.
I concur with that. We are three groups.
I think each of the voices heard is a representation of the separation of users. We can argue about percentages and how this would divide and we'd probably never reach agreement. But I think we can probably all see these 3 distinct tendencies.
I think so. - -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iF4EAREIAAYFAlVZQWYACgkQja8UbcUWM1wchAD+OPcFh/EW8X01lYoaONl87oNa jA5iERkPo5WC/RxguUkA/2p1irzmRbXNtsTMORHRH7y4N4ZWTkYeURxx6tvoRii5 =SF2b -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 18.05.15 Carlos E. R. wrote:
You forget the middle ground. Plain desktop/laptop users, wanting reasonably recent software, just not wanting to work the installation or upgrade every some months.
No, I did not forget them. I deliberately left them out. If there is enough manpower to create TW, openSUSE LTS and 'normal' releases, then ok. Yeah. If there is only manpower to get two products (and it seems like it), I would vote for TW and LTS. For users wanting the latest software without the rolling-release-thingy I would recommend LTS with add-on repositories to get the latest XYZ. Johannes -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 Comment: Using GnuPG with SeaMonkey - http://www.enigmail.net/ iEYEARECAAYFAlVbglgACgkQzi3gQ/xETbKosACfWQVch+36CLDw8hsn10pmR8Oz hzgAn2kd1Rcz3wW9Hk3npGH+tTVX2IyR =aFmr -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
participants (55)
-
Andreas Schwab
-
Andrei Borzenkov
-
Angelos Tzotsos
-
blue hut
-
Bob Williams
-
Bruno Friedmann
-
Carlos E. R.
-
Christian Boltz
-
Christoph Grüninger
-
Christoph Grüninger
-
Claudio Freire
-
Cristian Rodríguez
-
Felix Miata
-
Frederic Crozat
-
Graham P Davis
-
Greg Freemyer
-
greg.freemyer@gmail.com
-
Guido Berhoerster
-
Hans Witvliet
-
James Mason
-
Jan Engelhardt
-
Johannes Kastl
-
Larry Finger
-
Lars Müller
-
Ludwig Nussel
-
Malcolm
-
Manfred Hollstein
-
Marcus Meissner
-
Martin Schlander
-
Mathias Homann
-
Michal Kubecek
-
Oliver Neukum
-
Patrick Shanahan
-
Pawel Wieczorkiewicz
-
Per Jessen
-
Petr Gajdos
-
Raymond Wooninck
-
Richard Brown
-
Robert Martens
-
Robert Schweikert
-
Roger Oberholtzer
-
Roman Bysh
-
Ronan Arraes Jardim Chagas
-
Sid Boyce
-
Simon Lees
-
Stanislav Baiduzhyi
-
Stefan Brüns
-
Stefan Schäfer
-
Stefan Seyfried
-
Stephan Kulow
-
Torsten Gruner
-
Uzair Shamim
-
Wilhelm Boltz
-
Wolfgang Rosenauer
-
Yamaban