Guido Berhoerster wrote:

* Frederic Crozat [2011-06-10 19:04]:

...

systemd is coming for next openSUSE (12.1) scheduled next fall.

As far as I'm aware there was only one recent discussion on this this list about it [1] which started with the premise that systemd will be the default for 12.1. I'd like to know who has decided when and for what reasons that systemd will be the default for 12.1? More specifically, what alternatives were considered and why and how is systemd serving the openSUSE project better in the long term?

Same here. I would also like to say that I think going for systemd as default in 12.1 is very much bleeding edge, and maybe not entirely in line with our strategy. -- Per Jessen, Zürich (13.6°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

jdd wrote:

Le 11/06/2011 10:49, Per Jessen a écrit :

...

Same here. I would also like to say that I think going for systemd as default in 12.1 is very much bleeding edge, and maybe not entirely in line with our strategy.

may be I'm wrong, but I had the exact inverse feeling, that is that systemd would *not* be the default for 12.1.

Look at Frederics proposed plan, phase 2: * phase 2: systemd-sysvinit package installed by default and replace sysvinit. -- Per Jessen, Zürich (15.2°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Le 11/06/2011 18:39, Greg KH a écrit :

It should "just work" so there should not be anything you have to learn about, right?

wrong. I had quite often to edit init scripts for adapt to local situations, so I don't see why systemd could be any better in this respect (it's not it's fault), so if these scripts are no more used I will have to learn new things The moves from lpr to cups and to come from grub to grub2 or xorg.conf removed did no good for my configs, hal to udev also not to say we don't have to make moves, but often it is moving from something simple to something complicated that don't always work by the way I have a bug against systemd with no usefull answer (and only one answer, by the way), so I wonder if the debugging is already done if you remember, openSUSE did some bad moves in the past, so don't be surprised if people are frightened. Better doc, good debug are the key, and long life to systemd... jdd Nb: https://bugzilla.novell.com/show_bug.cgi?id=696917 right now 5 minutes wait during boot -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sat, Jun 11, 2011 at 08:29:17PM +0200, jdd wrote:

Le 11/06/2011 18:39, Greg KH a écrit :

...

It should "just work" so there should not be anything you have to learn about, right?

wrong.

I had quite often to edit init scripts for adapt to local situations, so I don't see why systemd could be any better in this respect (it's not it's fault), so if these scripts are no more used I will have to learn new things

Of course if you need to customize things you will have to adapt, to think otherwise is crazy. But for the majority of people out there, they never have to edit an init script. Heck, I haven't had to touch one myself for many many years now and I use Linux in all sorts of different situations.

The moves from lpr to cups and to come from grub to grub2 or xorg.conf removed did no good for my configs, hal to udev also

That is application configurations, and have nothing to do with systemd.

not to say we don't have to make moves, but often it is moving from something simple to something complicated that don't always work

systemv simple? Hah.

by the way I have a bug against systemd with no usefull answer (and only one answer, by the way), so I wonder if the debugging is already done

if you remember, openSUSE did some bad moves in the past, so don't be surprised if people are frightened.

People are always worried about change, I understand. Personally I think they should go use slackware if they are that concerned about change :) It's as if people here don't trust the developers of the system that they use to do the right thing. And that's an attitude that is not very wise, or useful. greg k-h -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sat, Jun 11, 2011 at 10:45:01PM +0200, Per Jessen wrote:

Greg KH wrote:

...

On Sat, Jun 11, 2011 at 08:29:17PM +0200, jdd wrote:

...

Le 11/06/2011 18:39, Greg KH a écrit :

...

It should "just work" so there should not be anything you have to learn about, right?

wrong.

I had quite often to edit init scripts for adapt to local situations, so I don't see why systemd could be any better in this respect (it's not it's fault), so if these scripts are no more used I will have to learn new things

Of course if you need to customize things you will have to adapt, to think otherwise is crazy.

But for the majority of people out there, they never have to edit an init script. Heck, I haven't had to touch one myself for many many years now and I use Linux in all sorts of different situations.

From which one can only deduce that you do not use it in real life.

I'm sorry that you feel my life is somehow not "real".

For instance, "for many many years" you've never had the need to deploy your own daemon for which openSUSE did not provide a suitable init-script.

That is exactly true.

...

It's as if people here don't trust the developers of the system that they use to do the right thing. And that's an attitude that is not very wise, or useful.

Oh, please, get off your high horse. Not trusting developers is indeed very wise, because developers most often live in ivory towers, blissfully unaware. I've spent a significant part of the last 25 years as a software engineer, I know only too well.

Then why are you participating here if you don't trust us? The developers are the ones doing the work, and if you don't trust them, well, we have larger issues. greg k-h -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Saturday, 2011-06-11 at 12:59 -0700, Greg KH wrote:

But for the majority of people out there, they never have to edit an init script. Heck, I haven't had to touch one myself for many many years now and I use Linux in all sorts of different situations.

Just for the record, I have, several times.

It's as if people here don't trust the developers of the system that they use to do the right thing. And that's an attitude that is not very wise, or useful.

ROTFL! Famous last words :-P I must have written somewhere the many times they failed. Now, where did I put that list... :-P - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk35CTkACgkQtTMYHG2NR9XymACfSeh+6bW1Ca/bL8xdBIaxXKhc cOcAn2mKgX9+bh57Tx5oJWtw8XeUT75l =rhxe -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sat, Jun 11, 2011 at 09:04:15PM +0200, Per Jessen wrote:

Greg KH wrote:

...

On Sat, Jun 11, 2011 at 11:31:45AM +0200, jdd wrote:

...

Le 11/06/2011 10:49, Per Jessen a écrit :

...

Same here. I would also like to say that I think going for systemd as default in 12.1 is very much bleeding edge, and maybe not entirely in line with our strategy.

By the time 12.1 comes out, we would not be the only distro shipping this, so it's not that "bleeding edge" at all.

systemd probably not, but our implementation/integration certainly.

And that is exactly what is being worked on here. How do you expect this to happen, unless people do the work? Are they not to do the work until it is magically done already for them?

After all, judging by e.g. Frederics proposed plan as well as previous discussions, we're not exactly talking plug-compatible here.

What do you mean by "plug"?

Given that our strategy says we aim for a "stable core", changing the default for a significant core component with a bare minimum of testing doesn't seem quite the right approach.

So that means we can never change anything in the core at all? That's nonsense.

...

...

I had many difficulties understanding inits, I'm not that impatient to learn yet an other tool (this and grub2 makes it rather frightening :-)

It should "just work" so there should not be anything you have to learn about, right?

Two many "should"s in that paragraph. IMHO.

Ok, how about I say, "it will just work so there is nothing you have to learn about". Unless you want to take advantage of the new tools and control systemd offers you. You can learn about those if you want to, but for everyone else, it will "just work". If not, file a bug. greg k-h -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sat, Jun 11, 2011 at 10:26:04PM +0200, Per Jessen wrote:

Greg KH wrote:

...

On Sat, Jun 11, 2011 at 09:04:15PM +0200, Per Jessen wrote:

...

Greg KH wrote:

...

On Sat, Jun 11, 2011 at 11:31:45AM +0200, jdd wrote:

...

Le 11/06/2011 10:49, Per Jessen a écrit :

...

Same here. I would also like to say that I think going for systemd as default in 12.1 is very much bleeding edge, and maybe not entirely in line with our strategy.

By the time 12.1 comes out, we would not be the only distro shipping this, so it's not that "bleeding edge" at all.

systemd probably not, but our implementation/integration certainly.

And that is exactly what is being worked on here. How do you expect this to happen, unless people do the work? Are they not to do the work until it is magically done already for them?

Did I suggest that? No, I was merely suggesting that making systemd the default was not the right approach wrt our strategy. If there are good reasons for moving to systemd, we should certainly work on it, but that is, IMHO, not sufficient for making it the default in 12.1. Once it has been available for a while, received sufficient testing (by the many people who are eagerly awaiting the new features), then we should make it the default.

So what would define "available for a while" sufficiently? It's available now, in 11.4 and Tumbleweed (well, in Tumbleweed:Testing at the moment, it will move to Tumbleweed proper next week). We have many months before 12.1 is out, shouldn't that be enough time to get this work done (testing, etc.)?

...

Ok, how about I say, "it will just work so there is nothing you have to learn about".

If it "will just work", I certainly have no problem with changing it - I just have a slight problem believing that it "will just work" with only a bare minimum of testing.

Then let's get more testers involved. greg k-h -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Markus Slopianka wrote:

Am Samstag 11 Juni 2011, 22:26:04 schrieb Per Jessen:

...

Once it has been available for a while, received sufficient testing (by the many people who are eagerly awaiting the new features), then we should make it the default.

12.1 will be released in November. By then a whole Fedora 15 life cycle with systemd is completed (Fedora 16 will also be released in November). That certainly counts as “available for a while, received sufficient testing”.

But does that include the many changes we need to make, or is systemd widely plug compatible that we don't need very many? If that is the case, then yes, that's plenty of testing. -- Per Jessen, Zürich (19.3°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Markus Slopianka wrote:

Am Sonntag 12 Juni 2011, 13:54:40 schrieb Per Jessen:

...

plug compatible

What?

http://en.wikipedia.org/wiki/Plug_compatible -- Per Jessen, Zürich (21.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sunday, June 12, 2011 11:08:41 AM Markus Slopianka wrote:

...

http://en.wikipedia.org/wiki/Plug_compatible

No, systemd is not hardware which allows you to plug new CPUs in an old socket

Software is plugable too, but I don't think that plan is to ensure that init and systemd are that way compatible. Plan is to have them both available with systemd being default, and for people with dependency on init, it will be there as an option. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Le 14/06/2011 10:48, Andreas Jaeger a écrit :

systemd will support the old initscripts, see the plan given.

We do not need to rewrite all packages and add new service files, we can keep the old system V ones - at least for most of the packages,

so typing "1" on the boot line will still log in maintenance mode? do you have a link to a not too difficult to read systemd info page? thanks jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Tuesday, June 14, 2011 11:09:53 Per Jessen wrote:

Andreas Jaeger wrote:

...

On Sunday, June 12, 2011 15:44:18 Per Jessen wrote:

...

Markus Slopianka wrote:

...

Am Sonntag 12 Juni 2011, 13:54:40 schrieb Per Jessen:

...

plug compatible

What?

http://en.wikipedia.org/wiki/Plug_compatible

I'm not sure what you mean with plug compatibel - so let me interpret it:

systemd will support the old initscripts, see the plan given.

We do not need to rewrite all packages and add new service files, we can keep the old system V ones - at least for most of the packages,

I'm surprised that quite a few people have asked about what it means to be plug compatible, I thought it was a common expression.

If you would have worded it as "Is it a plug-in replacement", I think more would have got what you meant. Plug compatibel can have different meanings as you noticed.

As far as my own tests have shown, systemd is largely plug compatible, although with exceptions such as acpid, syslog, cron.

ok, it's a plug-in replacement ;) Thanks for your tests, Andreas -- Andreas Jaeger, Program Manager openSUSE aj@{novell.com,suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126 -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

* Greg KH [2011-06-11 18:39]:

On Sat, Jun 11, 2011 at 10:41:56AM +0200, Guido Berhoerster wrote:

...

* Frederic Crozat [2011-06-10 19:04]:

...

systemd is coming for next openSUSE (12.1) scheduled next fall.

As far as I'm aware there was only one recent discussion on this this list about it [1] which started with the premise that systemd will be the default for 12.1. I'd like to know who has decided when and for what reasons that systemd will be the default for 12.1? More specifically, what alternatives were considered and why and how is systemd serving the openSUSE project better in the long term?

There is only one alternative, and that package is no longer being maintained or developed, so there really isn't anything else to choose from.

As for why to switch, you did read the long series of posts about systemd for admins, right? All of those things are stuff that users want, and care about, why would we not provide them?

This was not about offering systemd as an option, but part of the proposal ("phase 3") was to replace SysV init files with native systemd files. And an init daemon is not any arbitrary package, fully comitting to an implementation has long-term consequences. Hence, I would expect some kind of decision-making based on some real considerations rather than just following the latest buzz and the quite vocal promotion of its author.

...

I admit that I disapprove of its approach to cram everything but the kitchen sink into an init daemon (including stuff completely unrelated to init such as (auto)mounting, handling LUKS volumes, controlling the system locale, time, and hostname, replacing ConsoleKit, or the planned per-user session-startup functionality) rather thank keeping it simple and doing one thing well (a design philosophy which has served Un*x systems rather well in terms of functionality, security, and sustainability of codebases). So far it's not even clear where this will end.

It does one thing well, the rest is supported by helper scripts and plugins.

Do you have an alternative that you think should be used instead?

I'd really like to have an answer to my initial questions (including why we have to switch the default init system at all right now). But the alternatives I am aware of are sysvinit which we are using right now and upstart, given the (im)maturity of systemd, the lack of clarity where the scope of systemd will eventually end, and my aforementioned concerns I'd consider both of these the lesser evil.

And note, we aren't talking about taking away the choice to use systemv from what I can tell, you can always go back to that if you want to for some reason.

Following, see above. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

* Greg KH [2011-06-11 22:00]:

On Sat, Jun 11, 2011 at 08:28:39PM +0200, Guido Berhoerster wrote:

...

* Greg KH [2011-06-11 18:39]:

...

On Sat, Jun 11, 2011 at 10:41:56AM +0200, Guido Berhoerster wrote:

...

* Frederic Crozat [2011-06-10 19:04]:

...

systemd is coming for next openSUSE (12.1) scheduled next fall.

As far as I'm aware there was only one recent discussion on this this list about it [1] which started with the premise that systemd will be the default for 12.1. I'd like to know who has decided when and for what reasons that systemd will be the default for 12.1? More specifically, what alternatives were considered and why and how is systemd serving the openSUSE project better in the long term?

There is only one alternative, and that package is no longer being maintained or developed, so there really isn't anything else to choose from.

As for why to switch, you did read the long series of posts about systemd for admins, right? All of those things are stuff that users want, and care about, why would we not provide them?

This was not about offering systemd as an option, but part of the proposal ("phase 3") was to replace SysV init files with native systemd files. And an init daemon is not any arbitrary package, fully comitting to an implementation has long-term consequences. Hence, I would expect some kind of decision-making based on some real considerations rather than just following the latest buzz and the quite vocal promotion of its author.

And how do you know that was not done already?

I don't, it was and still is my initial question.

And also, please always remember, that changes happen here by people doing the work, not by people sitting around and discussing things, or dissing things.

That's not a valid argument when it comes to init which is probably the most important component of the system after the kernel. For such a change which directly and indirectly affects every other userspace component and, as I already said, has long-term consequences for the project (cost of implementation, cost of switching later etc.) I would expect a little more legitimacy. Apart from that a lot of packagers (probably including me) are expected to do work here.

Do you not trust the developers involved to get this working correctly? If so, offer to help out. If not, well, that's a different problem...

No I don't and I'm not interested in helping out with it since I disapprove with the direction it has taken (even in Redhat people seem to begin to disapprove if you look at fedora-devel). I would be interested in helping out with either improving the existing system or switching to something which is an init daemon and not a jack-of-all-trades attempting to replace half of the system.

...

...

...

I admit that I disapprove of its approach to cram everything but the kitchen sink into an init daemon (including stuff completely unrelated to init such as (auto)mounting, handling LUKS volumes, controlling the system locale, time, and hostname, replacing ConsoleKit, or the planned per-user session-startup functionality) rather thank keeping it simple and doing one thing well (a design philosophy which has served Un*x systems rather well in terms of functionality, security, and sustainability of codebases). So far it's not even clear where this will end.

It does one thing well, the rest is supported by helper scripts and plugins.

Do you have an alternative that you think should be used instead?

I'd really like to have an answer to my initial questions (including why we have to switch the default init system at all right now).

Because what we have right now sucks.

Seriously, it does, it was great for the 70's and 80's when things were static, but now, it makes absolutly no sense whatsoever. Linux has been evolving to support this type of dynamic, use only what you need when you need it, type of a system for a very long time now, and this is just one piece of that progression that has been needing to change for a very long time.

Yeah it sucks, and it has for a long time compared to its alternatives and I'm also familiar with SMF on Solaris, rc on FeeBSD and its modernization efforts[1], as well as upstart on Ubuntu. I'd still be interested in the reasoning why we have to switch right now and why it has to be systemd.

...

But the alternatives I am aware of are sysvinit which we are using right now and upstart, given the (im)maturity of systemd, the lack of clarity where the scope of systemd will eventually end, and my aforementioned concerns I'd consider both of these the lesser evil.

So what would make systemd somehow "mature" in your eyes? A major distro shipping it as their default init system? The developers working on it paid to do nothing else but support it and guarantee that it works properly for everyone?

I don't consider Fedora exactly as the standard for maturity (see fedora-devel and the Redhat bugtracker). Furthermore, it's not clear what direction it will eventually take and where it scope will end, what will come after the planned handling of user-sessions, a mail client[2]?

Or something else?

Oh, and upstart is a dead-end project, so that's not even an option, sorry.

Oh right, a year ago it was supposed to be the future. [1] http://people.freebsd.org/~trhodes/fsc/ [2] http://www.catb.org/jargon/html/Z/Zawinskis-Law.html -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

* Markus Slopianka [2011-06-12 13:49]:

...

Yeah it sucks, and it has for a long time compared to its alternatives and I'm also familiar with SMF on Solaris, rc on FeeBSD and its modernization efforts[1], as well as upstart on Ubuntu. I'd still be interested in the reasoning why we have to switch right now and why it has to be systemd.

Upstart contributions require copyright assignments to Canonical. That rules Upstart out.

Last year that was not a problem. (And I'm also not advertising upstart.)

And that you seriously propose init systems from completely different operating systems which therefore did never ever received any form of testing or development on Linux makes me wonder how serious to take your arguments about maturity.

I did not propose that, that would be complete nonsense anyway.

systemd is largely compatible with Mac OS X's launchd -- which means that Apple already did a big part patching deamons out there for launchd and systemd only requires a few additional tweaks.

It is not compatible to launchd, systemd is only copycatting parts of launchd's design. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sun, Jun 12, 2011 at 09:52:45AM +0200, Guido Berhoerster wrote:

Yeah it sucks, and it has for a long time compared to its alternatives and I'm also familiar with SMF on Solaris, rc on FeeBSD and its modernization efforts[1], as well as upstart on Ubuntu. I'd still be interested in the reasoning why we have to switch right now and why it has to be systemd.

Why now? Upstream claimed it is enough stable and feature complete, it has been already released in Fedora 15, so will be here any better time in future? And of course, having it enabled by default earlier in Factory would help with bugfixing a lot. From administrator point of view is the most important it provides the simple access to majority Linux features - cpu schedulers, io scheduling, cgroups, pam, process namespaces and so and so [1] With systemd you can disable, change or override the vendor suppliend service. Nothing will prevents you to create your own sshd.service, which will be never replaced on package upgrade like /etc/init.d/sshd. Having the same way how the service is started independently on the way how the start is triggered is unique - now you have to maintain init script and (x)inetd configuration file separate. With systemd whatever-activation will turn on the same .service file. Usage of cgroups makes the system much robust and prevents you from mistakes like killall in one init script will kill all instances of daemon. And with the easy of creating your own variants of init scripts, you can very easy maintain more instances of one service. And yes, when the whole logic of service run/start/restart is now a part of systemd, so no need to reimplement it again and again. There are of course more features, like socket-based activation, paralel start, no need of fork, clean execution environment, having a great cross-distro consolidation [2] (because it contains tools like) and so, but I hope this is enough for you. [1] http://0pointer.de/public/systemd-man/systemd.exec.html [2] http://0pointer.de/blog/projects/the-new-configuration-files.html Regards Michal Vyskocil

On Mon, Jun 13, 2011 at 06:45:33PM +0200, Guido Berhoerster wrote:

Anyway, since such a decision has implications for the whole distribution and its direction due to the associated costs of switchting to it and possibly away from it in the future I would expect some deliberation based on technical grounds in public _before_ making that decision including an evaluation of the alternatives, reasoning why it is desirable to fully commit to it right now (as opposed to providing optional support), associated costs, reasoning why it is in the interest of openSUSE etc. So far I haven't been able to find any such deliberation or even a simple discussion about making a decision in whichever form, there are just several statements by SUSE employees which take it for granted that systemd will be the default in 12.1, this was my main point and is what irks me the most in this case.

Hi Guido, first of all - I wrote only about fact it's good to have systemd in **Factory** enabled by default earlier than later. This will give us more time for testing and on the end we will have the hard facts to decide if it will make a sense enable it by default on 11.2 or not. This is similar to Linus's talk is cheap approach. Of course all people involved in systemd hopes it will be default in 11.2 - it does not make a sense to work on a project, when you don't trust it will be succesfull, right? For the rest, you wrote "there are just several statements by SUSE employees". Please be aware of that despite the part behind @ in our email, we are just the part of community like any other. At least me is telling only my POW here, but I'm sure that the rest of colleagues are doing the same. And my POW is simple - I am convinced systemd is a) better than any of sysvinit alternatives like upstart, runit, init-ng or whatever else I checked and read some documentation, b) is the future of Linux world. I don't care about costs (if you mean the mapower as a cost), because it will take a little time from me to support that. So I'm open to support it. But if you see a need for such decission making process before, then it's up to you as a part of community, because I (and probably others) see no need for it. IOW if you want such process, then you have make it. I would say you should ask openSUSE Board [1] for help. [1] http://en.opensuse.org/openSUSE:Board_2010 Best Regards Michal Vyskocil

Why not splitting Factory here? There´s the factory-tested tree, right? Why not using systemd in Factory, and the old start environment in factory-tested?

because there is a rule that says new commits to factory should not break it too bad... check the openssl case few weeks ago... factory should be used for testing... a broken factory defies the aim... Alin -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Adam Tauno Williams wrote:

On Mon, 2011-06-13 at 18:45 +0200, Guido Berhoerster wrote:

...

* Michal Vyskocil [2011-06-13 14:25]:

...

On Sun, Jun 12, 2011 at 09:52:45AM +0200, Guido Berhoerster wrote: Anyway, since such a decision has implications for the whole distribution and its direction due to the associated costs of switchting to it and possibly away from it in the future I would expect some deliberation based on technical grounds in public _before_ making that decision including an evaluation of the alternatives,

And the implied statement in this statement is that there has not been public deliberation. FACT: there has been, lots of it. systemd is a widely discussed topic, there have been articles on LWN, there is the systemd-for-admins PDF, and there has been no shortage of related traffic here and on other lists.

I haven't seen much discussion of the technical merits of systemd on any of the opensuse lists. Late last year it was mentioned that it could be tested (I remember trying it out), but that's about it.

Personally as a desktop/laptop user of openSUSE [and openSUSE is *the* platform I use] and as a professional LINUX administrator.... systemd is long overdue. Control of modern systems through a stack of shell scripts is a dreadful, tedious, and prone-to-failure hack.

Despite which it hasn't caused much of an issue here for the last ten-or-so years. -- Per Jessen, Zürich (24.1°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Rajko M. wrote:

On Tuesday, June 14, 2011 08:25:02 AM Per Jessen wrote:

...

Adam Tauno Williams wrote:

...

... there has been no shortage of related traffic here and on other lists.

I haven't seen much discussion of the technical merits of systemd on any of the opensuse lists.

It would be interesting if you can explain how average Linux user can discuss systemd vs. init, when majority have no idea that they exist, even lesser how they work.

Noone has suggested that, Rajko - we are talking about community members. What is the community here for if it is not for instance to discuss such important changes?

What would bring public discussion except usual bikeshedding (like this).

There is nothing trivial about this, imho. -- Per Jessen, Zürich (16.9°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Alberto Passalacqua wrote:

Hi,

2011/6/15 Per Jessen :

...

Noone has suggested that, Rajko - we are talking about community members.  What is the community here for if it is not for instance to discuss such important changes?

honestly I do not think members/non-members make a difference when it comes to technical aspects, especially of this kind.

I beg to differ. I really don't think you can dismiss the community just like that when it comes to technical questions. If the community cannot make a difference or have a valued opinion on technical aspects, why do we bother with the community at all?

The community can of course discuss things, but it should be done with some reason, and not just for the sole purpose of discussing, as it too often happens on these mailing list. Over-talking and over-discussing have been a problem for a long time, since it is a waste of time and resources, especially if, as it often happens, nothing new comes out from the discussion.

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

This also keeps active people, those who want to see things done and not over-talked, away from the mailing lists, with the most evident consequence that lists become less useful.

I'm sorry, but how do those "active people" know what to get done if they don't discuss it with the community? (obviously there is lots of stuff that does not warrant much discussion). -- Per Jessen, Zürich (17.9°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wednesday 15 June 2011 09:08:20 Per Jessen wrote:

Alberto Passalacqua wrote:

...

Hi, <snip>

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

I'm sure that is not what Alberto meant. I wonder what the value of this meta-discussion is. I don't see how we have to ask for discussion? The maintainers of systemd and the maintainers of sysvinit want to have systemd. That's quite clear. Nobody has stepped up with an alternative proposal and offered to do better. So the decision has been made implicitly by those who do the work. That is how a FOSS community works. Anyone can step in and change the outcome and that has always been the case. I agree with Alberto that you seem to want to discuss everything. Is that just to wave your flag around and keep others from working?!?

...

This also keeps active people, those who want to see things done and not over-talked, away from the mailing lists, with the most evident consequence that lists become less useful.

I'm sorry, but how do those "active people" know what to get done if they don't discuss it with the community? (obviously there is lots of stuff that does not warrant much discussion).

"discussion" is useless if it's for the sake of discussion. If anyone has objections, he/she can bring them up any time and come with a better proposal and do the work to prove it's possible. In absense of that, what is there to discuss? Yes, that is a rethorical question and the only valid answer is "nothing". Nobody wants to stop anyone from discussing anything if there are valid arguments. But complaining about a lack of discussion while you have nothing of any value to say is just wasting time. I did blog about that recently, see http://blog.jospoortvliet.com/ - read it and comment there if you disagree.

On Wednesday 15 June 2011 14:04:57 Per Jessen wrote:

Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 09:08:20 Per Jessen wrote:

...

Alberto Passalacqua wrote:

...

Hi,

<snip>

...

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

I'm sure that is not what Alberto meant.

I wonder what the value of this meta-discussion is. I don't see how we have to ask for discussion?

The maintainers of systemd and the maintainers of sysvinit want to have systemd. That's quite clear. Nobody has stepped up with an alternative proposal and offered to do better. So the decision has been made implicitly by those who do the work. That is how a FOSS community works.

Every now and then I see references to "Product Management" - don't they have a say?

Not over openSUSE, no.

...

Anyone can step in and change the outcome and that has always been the case. I agree with Alberto that you seem to want to discuss everything. Is that just to wave your flag around and keep others from working?!?

Ridiculous suggestion. You misunderstand, I have no interest in discussing everything.

...

...

...

This also keeps active people, those who want to see things done and not over-talked, away from the mailing lists, with the most evident consequence that lists become less useful.

I'm sorry, but how do those "active people" know what to get done if they don't discuss it with the community? (obviously there is lots of stuff that does not warrant much discussion).

"discussion" is useless if it's for the sake of discussion.

Certainly. Maybe we should have said "deliberation" instead, as I think Guido Berhoerster also did initially.

Deliberation suggests the people involved haven't thought about it and I think that is not a fair assesment. Really, if anyone has objections, he/she should bring them up. In absense of those it makes sense to trust the experts, don't you think?

/Per

Hey, On 15.06.2011 18:10, Greg Freemyer wrote:

On Wed, Jun 15, 2011 at 10:58 AM, Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 14:04:57 Per Jessen wrote:

...

Jos Poortvliet wrote:

Every now and then I see references to "Product Management" - don't they have a say?

Not over openSUSE, no.

I thought Coolo had the final say on what went into factory and what didn't?

From http://en.opensuse.org/openSUSE:Factory_development_model#Escalations Most decisions are taken by the maintainer of the package. If the maintainer can't take a decision or if a conflict arises between maintainers the devel project maintainers decide together. If the devel project maintainers also don't come to a conclusion or a conflict between two devel projects arises the openSUSE release team takes the decision. If the decision can't be taken by the release team they appeal to the openSUSE Board which tries to facilitate the decision with all the involved people. If that also is not successful the board puts up the issue to a vote among the members. Maintainer ==> Devel project maintainers ==> openSUSE:Release team ==> openSUSE:Board ==> openSUSE:Members Vote Henne -- Henne Vogelsang, openSUSE. Everybody has a plan, until they get hit. - Mike Tyson -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Am Mittwoch, 15. Juni 2011, 20:36:47 schrieb Wolfgang Rosenauer:

- If not, everything done to Factory in that respect until that happens needs an easy fallback plan to old behaviour. Is that made sure? - If yes, we are fine - If not, something is broken in the process above

I was under the assumption that we only have to take a decision once there is no way back. That means level 3 of Frederic's plan. Greetings, Stephan -- Sent from openSUSE -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Hey, On 15.06.2011 20:36, Wolfgang Rosenauer wrote:

Am 15.06.2011 20:05, schrieb Henne Vogelsang:

...

...

I thought Coolo had the final say on what went into factory and what didn't?

From

http://en.opensuse.org/openSUSE:Factory_development_model#Escalations

Maintainer ==> Devel project maintainers ==> openSUSE:Release team ==> openSUSE:Board ==> openSUSE:Members Vote

We are talking about replacing a central piece of the system. So Maintainer does not fit

Both maintainers agree as Jos wrote. No conflict, no escalation. Done :) Henne -- Henne Vogelsang, openSUSE. Everybody has a plan, until they get hit. - Mike Tyson -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wednesday 15 June 2011 21:51:42 Wolfgang Rosenauer wrote:

Am 15.06.2011 20:50, schrieb Henne Vogelsang:

...

Hey,

On 15.06.2011 20:36, Wolfgang Rosenauer wrote:

...

Am 15.06.2011 20:05, schrieb Henne Vogelsang:

...

...

I thought Coolo had the final say on what went into factory and what didn't?

From

http://en.opensuse.org/openSUSE:Factory_development_model#Escalations

Maintainer ==> Devel project maintainers ==> openSUSE:Release team ==> openSUSE:Board ==> openSUSE:Members Vote

We are talking about replacing a central piece of the system. So Maintainer does not fit

Both maintainers agree as Jos wrote. No conflict, no escalation. Done :)

No comment. If you think that's enough for that kind of change I'm out.

As I stated, if you have problems with the decision and want help make a better one, you can say so. Greg also mentioned he is not happy with the decision. I would expect a fact-based response from him (and you?) to the proposal from Frederic so you guys can discuss this as adults. I'm sorry but this is how it works in pretty much every FOSS community. Good examples are KDE and GNOME. Exceptions are Ubuntu ('Mark decides') and Fedora ('Red Hat decides') but I suppose you don't want to go that route, right?

Wolfgang

On Wed, 15 Jun 2011 22:59:17 +0200, jdd wrote:

...

I'm sorry but this is how it works in pretty much every FOSS community.

yes, but I've seen many post asking for bugreports. If the maintainers do decide themselves, they will have to make the bugreports themselve.

Now c'mon, jdd, you know that's not what was meant, and this kind of hyperbole really doesn't help advance the discussion. Obviously, maintainers have to make decisions based on the available information. Part of that available information is community-contributed bug reports. Part of it is asking (as the original message in this thread attempted to do) for those whom it would impact to recognize that it's not a trivial task and that there is some work necessary to make the change happen. Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Thu, 16 Jun 2011 09:22:30 +0200, jdd wrote:

Le 15/06/2011 23:17, Jim Henderson a écrit :

...

Obviously, maintainers have to make decisions based on the available information.

I think you underestimate the burden on members when a bad decision is taken.

No, I don't. But in order to make a good decision, *data* is needed. Not hyperbole, not guesses, but data.

Howmany times we had to explain "yes openSUSE sucks, but..." without having any mean to take part of the decision.

Welcome to the discussion, that's what this thread was started for, was it not? Frederic proposed a plan, and then the world exploded with people claiming there was no discussion about it. THIS is the discussion, people. We need less shouting and more factual data.

It's not possible to say "maintainer decides and others follow". The risk is to have no more followers.

Then provide some hard data points so a better decision can be made.

we have to understand than as openSUSE grow the non completely technical work grow, when the technical don't, and is also important.

Do also understand than of course, during discussion, most member do considere the developpers opinion as very important, but others opinions are also of some importance

To me, the opinions that matter are the ones that are backed up by facts and data, not the ones that seemingly say "the sky is falling" and seem to think that just yelling louder will get people's attention. The thing that will get people's attention is factual data. So test the thing in factory and report back on what works and what doesn't. At some point, we will get to a "go/no go" decision point for 12.1, and if the data at that point in the 12.1 release cycle shows that moving to systemd is a bad idea, then it seems to me those making the decision to include it should decide to hold off. But note that that doesn't mean it has to work 100% (nobody can test every possible test case, sorry Per, 100% doesn't generally happen *today* with any software product on the face of the planet), but it has to work well enough that the trade off between the problems and the advantages is acceptable. Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-06-16 at 09:29 +0200, jdd wrote:

may I try to summarize the systemd situation?

...

* testers used to tweek with init files are invited to open bugreports each time systemd do not fit they needs. May be some more documentation and some help can solve the problem

We need to be able to modify ourselves, and easily, the scripts or whatever systemd uses for each service, same as we do now with init.d. It is impossible to foreview any future tweek we may need in the future. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk3/wm4ACgkQtTMYHG2NR9XCkgCeLeFvr3qqYI2KvO3C6dWQFMh7 lLoAn3t+ffLloi/zKflgliBpNFa8VMe2 =a0/L -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

What I _want_ is that I have a working 12.1 without regressions caused by the fact we switched init. If that doesn't work I expect that a switch is flipped before the release to go back to what we had. I haven't heard other comments so I just expect this is how it will be done. This is exactly what Frederic proposed from the start - first we prepare, then we test and then we switch. And those that didn't test, have no right to

Am Mittwoch, 15. Juni 2011 schrieb Wolfgang Rosenauer: protest after we switched - IMO. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Am Donnerstag, 16. Juni 2011 schrieb jdd:

I know for sure that not all the programms in openSUSE have been tested (some don't even install)

Well, show me a decision to switch to an untested program for default and I'll buy this argument. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-06-16 at 10:14 +0200, Stephan Kulow wrote:

And those that didn't test, have no right to protest after we switched - IMO.

That is, at least, unkind. Not everybody can test, for many reasons: not having a second system, or the space, or the skills, or the time, for example. You know that the real testing phase of a version happens in the first month or two after release, not before. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk3/xCoACgkQtTMYHG2NR9XYbQCgioMGEvo4p+ssl48O7xsPSvki 9dsAoIKv/I7asQ7P9wOQBY25byx2i2XN =Qc2i -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Hey, On 15.06.2011 21:51, Wolfgang Rosenauer wrote:

Am 15.06.2011 20:50, schrieb Henne Vogelsang:

...

On 15.06.2011 20:36, Wolfgang Rosenauer wrote:

...

Am 15.06.2011 20:05, schrieb Henne Vogelsang:

...

...

I thought Coolo had the final say on what went into factory and what didn't?

From

http://en.opensuse.org/openSUSE:Factory_development_model#Escalations

Maintainer ==> Devel project maintainers ==> openSUSE:Release team ==> openSUSE:Board ==> openSUSE:Members Vote

We are talking about replacing a central piece of the system. So Maintainer does not fit

Both maintainers agree as Jos wrote. No conflict, no escalation. Done :)

No comment.

That kind of defeats the purpose of this list dude ;-)

If you think that's enough for that kind of change I'm out.

Of course everybody can always chime in with arguments. You can, the Release Team can, anyone. What do you want to organize? The decision making process (a.k.a. Escalations) or the how/what/when of the implementation? If you think Frederics proposal lacks then tell him. He'll listen and incorporate your feedback. That's what this thread initially was for right? Henne -- Henne Vogelsang, openSUSE. Everybody has a plan, until they get hit. - Mike Tyson -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Hey, On 15.06.2011 22:27, Greg Freemyer wrote:

On Wed, Jun 15, 2011 at 2:50 PM, Henne Vogelsang wrote:

...

On 15.06.2011 20:36, Wolfgang Rosenauer wrote:

...

Am 15.06.2011 20:05, schrieb Henne Vogelsang:

...

...

I thought Coolo had the final say on what went into factory and what didn't?

From

http://en.opensuse.org/openSUSE:Factory_development_model#Escalations

Maintainer ==> Devel project maintainers ==> openSUSE:Release team ==> openSUSE:Board ==> openSUSE:Members Vote

We are talking about replacing a central piece of the system. So Maintainer does not fit

Both maintainers agree as Jos wrote. No conflict, no escalation. Done :)

I personally want to see the change, but I don't like the decision process at this point.

And what you wrote is not how I read the escalation policy. It says

"if a conflict arises between maintainers the devel project maintainers decide together".

In this case, every package that has a init script is potentially impacted. All of the maintainers that maintain a init script need to be part of the decision.

Now go back to the first mail in this thread and read it again. This is exactly what Frederic tried. He said he wants to do it, he said he has support from upstream, he provided all the necessary info for script maintainers, he proposed a plan how to do it smoothly, even with keeping an exit strategy, he said where and how he would like to discuss this. What else do you want? Henne -- Henne Vogelsang, openSUSE. Everybody has a plan, until they get hit. - Mike Tyson -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Am Donnerstag, 16. Juni 2011 schrieb Greg Freemyer:

It starts with "systemd is coming for next openSUSE (12.1) scheduled next fall."

And then it follows with a roll-out plan.

That's what it is: a plan. A plan Frederic needs support(ers) for to succeed and so he posted it not only on -factory but even on -project to invite everyone to this big project. And blaming him for believing that openSUSE is about doing something instead of talking about it is lame. The posting of the plan was the perfect time to raise issues and points that you want to be checked in the testing phase. General "SL 10.3 was the best, no reason to change anything" discussions belong on opensuse@ I believe ;( And systemd (and the associated plymouth) is something that generated a lot of "me too" mails in my fate mail folder, but for some reason it dried out. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Hey, On 15.06.2011 22:32, jdd wrote:

Le 15/06/2011 20:05, Henne Vogelsang a écrit :

openSUSE:Release team

http://en.opensuse.org/openSUSE:Release_team

coolo and darix. Did this change since the community is said to have a major importance in the project?

Coolo and Darix are exemplary pillars of our community. What counts for openSUSE is the contribution someone brings, not who his employer is.

I think there should be some more people (not saying these two are not good, they are!)

How about you read that page to the end? :) How to join Are you an openSUSE Member and want to join us? Just head over to one of our communication channels and talk to us, there are plenty of things to do! But please be aware that maintaining a distribution is a very time consuming and technically challenging task. We appreciate if you bring something to the table but we most likely won't find the time to teach you a lot. Henne -- Henne Vogelsang, openSUSE. Everybody has a plan, until they get hit. - Mike Tyson -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Le 15/06/2011 16:58, Jos Poortvliet a écrit :

Deliberation suggests the people involved haven't thought about it and I think that is not a fair assesment. Really, if anyone has objections, he/she should bring them up. In absense of those it makes sense to trust the experts, don't you think?

experts are not experts on anything like marketing, for example There should be a formal way to make sure a decision is made on important subjects in a known way by known people. A "technical staff" with experts from various domain, including marketing and project, in charge of saying the final "yes" or "no". A way not to do again some odd decision we suffered from in the past witout even knowing who decided to do so. This staff could include the board (but not only) and hold an IRC meeting at some important period of the distribution lifetime. Just before RC and feature freeze could be a good moment. Asking some questions like: * systemd and inits at still available, what one do we make default for the Gold master? * Gnome 4 is nearly ready, and us are we ready? you see the point? Like we work now, nobody is responsible (who choosed between kde3 and kde4 as default?) - for the good or the bad jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wed, 15 Jun 2011 12:45:07 -0400, Patrick Shanahan wrote:

This is becoming a *government/politics* issue. Have we reached the point of "leaership by committee"? At some point, *someone* must be empowered to make rational decisions as the world will wait for no-one and we *will* be left in the lurch.

I *trust* those with knowledge to make intelligent decisions about my software platform the same as I trust my doctor/lawyer/financier/... When I find that the trust is not warranted, I *will* make changes.

I will say that from the "discussion" here that I feel (I have sufficient lack of expertise to do other than "feel") systemd is the future.

Well stated, Patrick. It's useful to discuss things to a point, but a decision needs to be made (regardless of the topic) and we need to move on to the next thing. Trying to come to a group consensus is a useful decision making process, but not something that should be employed for every single decision being made - the community can't micromanage every aspect of the project - that's part of the reason we elected a leadership team in the project. I trust them to make the right decisions for the project overall, even if I don't always agree with every individual small decision. If I didn't trust them, I wouldn't vote for them. Yes, I even trust the ones I didn't vote for, because they represent a majority in the community - and if the community starts going in a direction I can't agree with, then the best remedies I have are either to find a different community to put my efforts into, or to run for the board myself. With regards to the topic under discussion, maybe someone could (if it hasn't already been done) post a high-level comparison of the two components and identify what a migration path would look like to get from sysvinit to systemd. If we are going to switch, we need to provide a migration path for those who have created custom init scripts. Change is rarely easy. So if a change is to be made, the focus should be on making the change as easy/painless as possible, and we should highlight the advantages that the change brings. Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wed, 15 Jun 2011 20:34:57 +0200, Per Jessen wrote:

Jim Henderson wrote:

...

It's useful to discuss things to a point, but a decision needs to be made (regardless of the topic) and we need to move on to the next thing. Trying to come to a group consensus is a useful decision making process, but not something that should be employed for every single decision being made - the community can't micromanage every aspect of the project - that's part of the reason we elected a leadership team in the project.

I was not aware that we had elected a leadership team.

They're called the "openSUSE Board". Surely you've heard of them.

...

With regards to the topic under discussion, maybe someone could (if it hasn't already been done) post a high-level comparison of the two components and identify what a migration path would look like to get from sysvinit to systemd.

I think that is what part of what we are asking. Overall, switching to systemd sounds like a good idea to me, although I don't think we should make it default in 12.1. systemd comes with certain new features (e.g. cgroup support) which will no doubt be attractive to some more than others. Let the keen, early adopters iron out whatever wrinkles might be left after Factory, then let the regular user get a system (in 12.2) that "just works".

I would guess that most users probably won't care which one is in place as long as the stuff they're using works properly. Most people don't care about the init scripts until such time as they fail to actually start a needed service. We're still a ways off from 12.1, I think far enough that there's time to iron the wrinkles out in Factory and have it ready for the 12.1 release.

...

If we are going to switch, we need to provide a migration path for those who have created custom init scripts.

systemd is, as far as my own tests have shown, virtually sysvinit compatible, i.e. custom init scripts will quite likely continue to work.

Well, then, why the objections to the switch? Why all the discussion? Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wed, 2011-06-15 at 19:33 +0200, Per Jessen wrote:

...

I will say that from the "discussion" here that I feel (I have sufficient lack of expertise to do other than "feel") systemd is the future. I agree, but the discussion (as I see it) has been more about whether to push it into to being the default when it being the default is not really needed nor warranted. (as far as the facts have been presented to us).

I think it should become the default; because maintaining two stacks is just wasted effort [systemd systems and non-systemd systems]. Make a change and make it clean and sharp. If dynamic workstation / laptops require it [or it at least makes life easier] then just deploy it as the default; servers don't much care and sys-admins will adapt. And there are advantages to systemd for servers as described in the systemd-for-admins PDF.

My position is to push back until our integration or systemd has received sufficient testing. It is _easy_ to test and measure for anyone who boots regularly.

There is nobody arguing for including it prior to "sufficient testing". -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Jos Poortvliet wrote:

On Wednesday 15 June 2011 14:04:57 Per Jessen wrote:

...

Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 09:08:20 Per Jessen wrote:

...

Alberto Passalacqua wrote:

...

Hi,

<snip>

...

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

I'm sure that is not what Alberto meant.

I wonder what the value of this meta-discussion is. I don't see how we have to ask for discussion?

The maintainers of systemd and the maintainers of sysvinit want to have systemd. That's quite clear. Nobody has stepped up with an alternative proposal and offered to do better. So the decision has been made implicitly by those who do the work. That is how a FOSS community works.

Every now and then I see references to "Product Management" - don't they have a say?

Not over openSUSE, no.

Okay, I was not aware we had progressed so far. Can I quote you?

...

Certainly. Maybe we should have said "deliberation" instead, as I think Guido Berhoerster also did initially.

Deliberation suggests the people involved haven't thought about it

Not really, but I won't bore you with explaining it. Google is your friend.

and I think that is not a fair assesment. Really, if anyone has objections, he/she should bring them up. In absense of those it makes sense to trust the experts, don't you think?

(please stop being condescending, it doesn't suit you). No, I don't think so. It doesn't always make sense to leave decision making to the experts. In particular when it comes to controversial or non-trivial decisions, I think it makes sense to ask the experts to deliberate with the community first. Not seek a consensus, just inform people about pros and cons and whys and hows. Like I suggested to Greg KH, developers (and experts) often live in ivory towers, blissfully unaware. This does not, imho, make them qualified to make decisions for the greater good. -- Per Jessen, Zürich (26.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Greg KH wrote:

...

No, I don't think so. It doesn't always make sense to leave decision making to the experts. In particular when it comes to controversial or non-trivial decisions, I think it makes sense to ask the experts to deliberate with the community first. Not seek a consensus, just inform people about pros and cons and whys and hows. Like I suggested to Greg KH, developers (and experts) often live in ivory towers, blissfully unaware. This does not, imho, make them qualified to make decisions for the greater good.

I think you will find that all of the openSUSE developers are not "blissfully unaware" and are rather pragmatic people who are trying to do the right thing.

Thanks for proving me right, Greg.

And to tell someone who is actually doing the work that what they are doing is wrong and should be debated more, by people who don't do the work at all, is being condescending.

Actually, that is called management. Something we (as a project) lack.

Again, trust the people doing the work. If you don't, then there are lots of other distros out there to try out, but again, you are going to have to trust the work that they do, or help do it yourself.

Can someone tell me what the point of the openSUSE community is then? If we, the community members, are not supposed to (at least try to) exercise our influence on where openSUSE is going and what it is going to be, what's the point of the community? If the direction and the contents of openSUSE are being dictated to us anyway, why bother taking part in the community? -- Per Jessen, Zürich (23.7°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Robert Schweikert wrote:

On 06/15/2011 03:00 PM, Per Jessen wrote:

...

Can someone tell me what the point of the openSUSE community is then?

As a community we come together to work on the distribution, tools, etc. Discussing is not work.

It is a precursor - if we don't discuss and agree on an approach, we have anarchy.

As pointed out by Coolo, the maintainers of a package have ultimate say what gets done. How does it help to have an init system in 12.1 that no one is willing to maintain?

It doesn't help, but it's not a problem either. However, a halfway finished init-system does not help either, but it would be a problem.

This should be treated no differently then any other package.

You don't think a complete change of the init system is just a little more critical than upgrades of kedit or jfsutils?

Following your arguments we should have a gazillion message discussion about what kernel version we should have in 12.1, we should have also had a lengthy discussion about updating glibc, and what about GNOME, maybe we should discuss the use of GNOME 3 until the end of time.

Robert, you haven't been paying attention - we are talking about deliberation, i.e. informing people about pros and cons and whys and hows, encouraging them to voice their concerns and/or opinions. Some of that is slowly beginning now, but the approach has had to be forced and has been completely backward, imho.

I personally was very happy to see a plan put forth for the switch over, rather than being surprised one day that systemd "magically" showed up. The reward for those putting in the effort of doing the work and trying to keep people informed? Messages from naysayers that do not participate in the actual work.

Surely it is a reward that people care enough to speak up. An indifferent community would kill the project.

If you want to influence where things are going, use your fingers to write code, spec files etc. rather than a stream of messages to the ML.

I do too, but I happen to think both are important for the project. There are no doubt lots of openSUSE community members who do not "do the work" - I for one regard their thoughts and opinions to be as important as those of any other member. -- Per Jessen, Zürich (19.7°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Sun, Jun 19, 2011 at 01:59:56PM +0200, Jos Poortvliet wrote:

On Thursday 16 June 2011 18:33:56 Per Jessen wrote:

...

Robert Schweikert wrote:

...

On 06/15/2011 03:00 PM, Per Jessen wrote:

...

Can someone tell me what the point of the openSUSE community is then?

As a community we come together to work on the distribution, tools, etc. Discussing is not work.

It is a precursor - if we don't discuss and agree on an approach, we have anarchy.

But how can you claim there is no discussion? Frederic brought it up to start one, and that is exactly what is happening. The whole whining about "but we don't discuss things" is bull****

Jos, his initial e-mail read like "this will be done for 12.1" perhaps it was incorrectly worded, but it read to me like "these are the facts". There was no invitation for discusion. Ciao, Marcus -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Monday 20 June 2011 10:58:42 Per Jessen wrote:

Marcus Meissner wrote:

...

On Sun, Jun 19, 2011 at 01:59:56PM +0200, Jos Poortvliet wrote:

...

On Thursday 16 June 2011 18:33:56 Per Jessen wrote:

...

Robert Schweikert wrote:

...

On 06/15/2011 03:00 PM, Per Jessen wrote:

...

Can someone tell me what the point of the openSUSE community is then?

As a community we come together to work on the distribution, tools, etc. Discussing is not work.

It is a precursor - if we don't discuss and agree on an approach, we have anarchy.

But how can you claim there is no discussion? Frederic brought it up to start one, and that is exactly what is happening. The whole whining about "but we don't discuss things" is bull****

Jos,

his initial e-mail read like "this will be done for 12.1"

perhaps it was incorrectly worded, but it read to me like "these are the facts". There was no invitation for discusion.

Exactly.

He could've done better. But he presented the plan the maintainers had, if it wasn't open for discussion, why present it? This is how it works everywhere in FOSS so don't get upset about that. Yes, he could've added a "thoughts anyone?" at the bottom but the lack of that doesn't warrant this huge bikeshedding discussion, nor the offensive responses to the plan by some.

Can someone tell me what the point of the openSUSE community is then? If we, the community members, are not supposed to (at least try to) exercise our influence on where openSUSE is going and what it is going to be, what's the point of the community?

You're confusing community members with consumers. No, FOSS consumers have no say in FOSS projects. At best they can say "I disagree for the following reasons: ..." and the people who are actually doing the work then decide in part on those reasons. As a consumer you have the choice to use another product but no consumer can dictate from the outside what the developers inside do. That system is not perfect but it gets stuff done. If you want to keep SysV, you have to maintain it yourself (or hire someone to do it for you). Then you can make a CD image yourself in SUSE Studio with SysV as default. And since that is your project, nobody from the outside has any right to interfere. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wednesday 15 June 2011 19:20:55 Per Jessen wrote:

Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 14:04:57 Per Jessen wrote:

...

Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 09:08:20 Per Jessen wrote:

...

Alberto Passalacqua wrote:

...

Hi,

<snip>

...

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

I'm sure that is not what Alberto meant.

I wonder what the value of this meta-discussion is. I don't see how we have to ask for discussion?

The maintainers of systemd and the maintainers of sysvinit want to have systemd. That's quite clear. Nobody has stepped up with an alternative proposal and offered to do better. So the decision has been made implicitly by those who do the work. That is how a FOSS community works.

Every now and then I see references to "Product Management" - don't they have a say?

Not over openSUSE, no.

Okay, I was not aware we had progressed so far. Can I quote you?

Sure. I'm being a bit black and white here, AJ (who is Product Manager for openSUSE) probably could influence what SUSE employees want. But they can't dicatate what happens in openSUSE other than by doing the work - which puts them on equal foot with others doing work. And obviously way above people who just talk and talk (but that's how it should be, right?).

...

...

Certainly. Maybe we should have said "deliberation" instead, as I think Guido Berhoerster also did initially.

Deliberation suggests the people involved haven't thought about it

Not really, but I won't bore you with explaining it. Google is your friend.

...

and I think that is not a fair assesment. Really, if anyone has objections, he/she should bring them up. In absense of those it makes sense to trust the experts, don't you think?

(please stop being condescending, it doesn't suit you). No, I don't think so. It doesn't always make sense to leave decision making to the experts. In particular when it comes to controversial or non-trivial decisions, I think it makes sense to ask the experts to deliberate with the community first. Not seek a consensus, just inform people about pros and cons and whys and hows. Like I suggested to Greg KH, developers (and experts) often live in ivory towers, blissfully unaware. This does not, imho, make them qualified to make decisions for the greater good.

I think Grek and Robert have answered this better than I could. I also talked about this in my blog, which I pointed you to before. A convenient link for you: http://bit.ly/lZu4fT BTW the discussion in public that Guido asked for is happening, or rather, it was until it got derailed in this silly meta discussion. Reading through the tread again, Guido has concerns which were being addressed. In the end, maybe the problems he mentions can be solved. Maybe not and in that case, the release team makes the decision. If they (ok, it's just Coolo right now) don't feel comfortable with that they can ask the Board. See: http://en.opensuse.org/openSUSE:Factory_development_model This is a complicated issue and it is OK and understandable that some people have concerns. We discuss them, that is how it should be and that is how decisions are made. Let's please stop the meta-discussion about having discussions, it gets us nowhere. If Guido and others feel systemd is problematic, they should try to come up with an alternative solution. If they can't it's a choice between an arguably bad thing and nothing - no choice at all... It sucks but that is how it is.

* Jos Poortvliet [2011-06-15 16:58]:

On Wednesday 15 June 2011 14:04:57 Per Jessen wrote:

...

Jos Poortvliet wrote:

...

On Wednesday 15 June 2011 09:08:20 Per Jessen wrote:

...

Alberto Passalacqua wrote:

...

Hi,

<snip>

...

Sorry, that's apparently how the openSUSE community works - if we're not happy with it, it's up to us to change it. And I don't think it's better done by not discussing changes at all.

I'm sure that is not what Alberto meant.

I wonder what the value of this meta-discussion is. I don't see how we have to ask for discussion?

The maintainers of systemd and the maintainers of sysvinit want to have systemd. That's quite clear. Nobody has stepped up with an alternative proposal and offered to do better. So the decision has been made implicitly by those who do the work. That is how a FOSS community works.

Every now and then I see references to "Product Management" - don't they have a say?

Not over openSUSE, no.

AFAIK SUSE is not a charity but has a business interest in protoyping and evaluating features before including them in their flagship product. In this case they seem to even sponsor some of its development which of course makes sense from theri POV if there are customers demanding a feature which systemd provides (as it was mentioned in another mail).

...

...

Anyone can step in and change the outcome and that has always been the case. I agree with Alberto that you seem to want to discuss everything. Is that just to wave your flag around and keep others from working?!?

Ridiculous suggestion. You misunderstand, I have no interest in discussing everything.

...

...

...

This also keeps active people, those who want to see things done and not over-talked, away from the mailing lists, with the most evident consequence that lists become less useful.

I'm sorry, but how do those "active people" know what to get done if they don't discuss it with the community? (obviously there is lots of stuff that does not warrant much discussion).

"discussion" is useless if it's for the sake of discussion.

Certainly. Maybe we should have said "deliberation" instead, as I think Guido Berhoerster also did initially.

Deliberation suggests the people involved haven't thought about it and I think that is not a fair assesment. Really, if anyone has objections, he/she should

As if that hadn't happened before in the Linux sphere. And when things do not turn out as anticipated we end up with code churn, endless wasted development hours, and regressions when switching to something supposedly better (remember the hal->DeviceKit->u* mess when hal was found to have become too complex and unmaintainable?).

bring them up. In absense of those it makes sense to trust the experts, don't you think?

No, I do have concerns and I have mentioned them, most importantly it is not even clear yet where systemd's aspirations actually end, ie. what it will attempt to replace next. I've also mentioned before that an init system, or better a "platform" as its main developer refers to it, is not an arbitrary package but affects the direction this project takes. In addition there are costs associated with switching which are spread among many oS developers. Since SUSE is investing into systemd development it apparently has a vested interest in forcing its usage in oS, so I guess it makes no sense to further discuss this here. If there is an executive managment decision that requires no further discussion then lets please clearly name it as such in the future. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Am Dienstag, 14. Juni 2011 schrieb Per Jessen:

Adam Tauno Williams wrote:

...

On Mon, 2011-06-13 at 18:45 +0200, Guido Berhoerster wrote:

...

* Michal Vyskocil [2011-06-13 14:25]:

...

On Sun, Jun 12, 2011 at 09:52:45AM +0200, Guido Berhoerster wrote: Anyway, since such a decision has implications for the whole distribution and its direction due to the associated costs of switchting to it and possibly away from it in the future I would expect some deliberation based on technical grounds in public _before_ making that decision including an evaluation of the alternatives,

And the implied statement in this statement is that there has not been public deliberation. FACT: there has been, lots of it. systemd is a widely discussed topic, there have been articles on LWN, there is the systemd-for-admins PDF, and there has been no shortage of related traffic here and on other lists.

I haven't seen much discussion of the technical merits of systemd on any of the opensuse lists. Late last year it was mentioned that it could be tested (I remember trying it out), but that's about it.

It could be tested since then and its availablity was even part of the 11.4 announcement.

...

Personally as a desktop/laptop user of openSUSE [and openSUSE is *the* platform I use] and as a professional LINUX administrator.... systemd is long overdue. Control of modern systems through a stack of shell scripts is a dreadful, tedious, and prone-to-failure hack.

Despite which it hasn't caused much of an issue here for the last ten-or-so years.

You don't seem to maintain these scripts. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Am 11.06.2011 21:53 schrieb Greg KH:

[...] Because what we have right now sucks.

Seriously, it does, it was great for the 70's and 80's when things were static, but now, it makes absolutly no sense whatsoever. Linux has been evolving to support this type of dynamic, use only what you need when you need it, type of a system for a very long time now, and this is just one piece of that progression that has been needing to change for a very long time.

So if I'm installing a web server machine with apache, it will boot quickly, but the first access to the server will take almost forever because nobody connected to port 80 before and thus apache wasn't started? Or does "use only what you need when you need it" not apply here (and if not, why)? Looking at http://0pointer.de/blog/projects/socket-activation.html it seems that systemd socket activation adds an additional overhead of a few context switches for every single connect. How big is the performance degradation for short-lived TCP and UDP connections in that case? Regards, Carl-Daniel -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

On Wed, 2011-06-15 at 09:53 +0200, Frederic Crozat wrote:

Le mercredi 15 juin 2011 à 09:08 +0200, Carl-Daniel Hailfinger a écrit :

...

Am 11.06.2011 21:53 schrieb Greg KH:

...

[...] Because what we have right now sucks.

Seriously, it does, it was great for the 70's and 80's when things were static, but now, it makes absolutly no sense whatsoever. Linux has been evolving to support this type of dynamic, use only what you need when you need it, type of a system for a very long time now, and this is just one piece of that progression that has been needing to change for a very long time.

So if I'm installing a web server machine with apache, it will boot quickly, but the first access to the server will take almost forever because nobody connected to port 80 before and thus apache wasn't started? Or does "use only what you need when you need it" not apply here (and if not, why)?

Socket activation is mainly used to simplify service startup and to avoid to express any dependencies, and not about on-demand start. On a common systemd setup, there are almost zero dependencies to resolve, because all communication channels between services are established before any service is ever started. The kernel will just suspend the requesters until the real service is running. The same logic applies to filesystems access when systemd's automounter is used. Most common services are started just unconditionally, like they have been started since forever, regardless if there are traffic on the socket or not.

...

Looking at http://0pointer.de/blog/projects/socket-activation.html it seems that systemd socket activation adds an additional overhead of a few context switches for every single connect. How big is the performance degradation for short-lived TCP and UDP connections in that case?

Are you running apache behind xinetd ATM ? I don't think so. So I don't see why you would use socket-activation for apache server.

Right, if you run a real webserver, you don't run it inetd mode. Also, the overhead for native/non-inetd systemd socket activation is zero. Apache listens itself, it only gets the first connection passed, if configured to run with socket activation. SShd is a fine example, and there are not many, to run by default from systemd's inetd mode. On most workstations nobody ever logs-in over ssh, and it can be started on demand just fine. If you run a server, where users log-in over ssh all the time, you wouldn't run sshd in inetd mode there, but just enable the service unconditionally, so it is immediately ready when users connect. Kay -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Kay Sievers wrote:

On Wed, 2011-06-15 at 11:48 +0200, Kay Sievers wrote:

...

...

Le mercredi 15 juin 2011 à 09:08 +0200, Carl-Daniel Hailfinger a écrit :

...

...

...

So if I'm installing a web server machine with apache, it will boot quickly, but the first access to the server will take almost forever because nobody connected to port 80 before and thus apache wasn't started? Or does "use only what you need when you need it" not apply here (and if not, why)?

Socket activation is mainly used to simplify service startup and to avoid to express any dependencies, and not about on-demand start.

The other example of socket activation is package upgrades. When pid 1 keeps the socket established, you can replace packages without ever losing any new connection.

You don't need systemd for that though. Keeping sockets open across exec is an operating system feature. The hard part is restoring the application's state and that's nothing systemd can help with. From the top of my head I only remember irssi which can re-exec itself since ages without losing connections or state. It's embarassing that e.g. dbus updates require a system reboot but that's nothing systemd can magically solve.

With systemd you can even replace services like syslogd with a different implementation without ever losing a single message.

Well, I doubt that's relevant in practice and would only work for the local socket anyways. Kind of funny that you take syslog as example though since you only care about rsyslogd.

Services are also automatically restarted if a service crashes, without losing any new message submitted during the time the service was not available. Clients will never see connection refused.

These are mostly features enterprise users asked for, and what no other init can really provide or work around.

The plain old inetd could already do local socket activation¹. Maybe the name prevented people from using it for that purpose. Unfortunately xinetd lost the ability to use local sockets. Implementing it wouldn't be rocket science though. cu Ludwig [1] http://www.freebsd.org/cgi/man.cgi?query=inetd&sektion=8 -- (o_ Ludwig Nussel //\ V_/_ http://www.suse.de/ SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org

Kay Sievers wrote:

On Thu, 2011-06-16 at 13:18 +0200, Ludwig Nussel wrote:

...

Kay Sievers wrote:

...

The other example of socket activation is package upgrades. When pid 1 keeps the socket established, you can replace packages without ever losing any new connection.

You don't need systemd for that though. Keeping sockets open across exec is an operating system feature. The hard part is restoring the application's state and that's nothing systemd can help with. From the top of my head I only remember irssi which can re-exec itself since ages without losing connections or state. It's embarassing that e.g. dbus updates require a system reboot but that's nothing systemd can magically solve.

We don't want any service besides pid 1 to re-ecec itself. It's just to fragile to get right. [...] You can upgrade/restart many services without interruption.

Maybe you didn't run "systemctl restart dbus.service" today to watch your login session collapse :-) No doubt that systemd taking care of listening sockets is useful for simple, more or less stateless services. Starting udevd that way isn't too impressive though. As soon as you have more connections and state information associated with the connections the application needs to do the hard work though, no matter whether started via systemd or not. Next dbus security update is already in the queue btw. cu Ludwig -- (o_ Ludwig Nussel //\ V_/_ http://www.suse.de/ SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org