* Richmond [05-06-17 19:41]:

Why is there an i586 of TW, when there is no Leap i586? What's the point of testing it if no one can use it?

TW is not "testing", that is factory. And TW is not Leap XX.X. why do you think no one can use it? -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri Registered Linux User #207535 @ http://linuxcounter.net Photos: http://wahoo.no-ip.org/piwigo -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

* Richmond [05-06-17 20:03]:

Patrick Shanahan wrote:

...

* Richmond [05-06-17 19:41]:

...

Why is there an i586 of TW, when there is no Leap i586? What's the point of testing it if no one can use it? TW is not "testing", that is factory. And TW is not Leap XX.X.

why do you think no one can use it?

I thought that TW was the testing place for packages which became part of the release version eventually.

I guess I am confused.

I guess. I have been using Tw for quite a few years as a productive environment, on 4 boxes. I have leap 42.2 as a server. factory is the development environment, precursor to Tw. But basically they are the same. many packages appear on factory before release and then to Tw. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri Registered Linux User #207535 @ http://linuxcounter.net Photos: http://wahoo.no-ip.org/piwigo -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Tumbleweed and Leap are two different distributions with discreet code bases. Tumbleweed is not the testing branch for Leap. The long explanation for that is my video from FOSDEM on the topic: https://youtu.be/GoKYpj6LuJg Tumbleweed has an i586 version because we have contributors maintaining it. Leap does not because we do not have contributors maintaining it. The sad truth though is that even with the support of our contributors for i586 Tumbleweed, the architecture itself is DEAD and unmaintainable in the long term. The amount of packages that just can not build for i586 is increasing. Things like Firefox are becoming unfeasible to build for i586. The capability for our maintainers to fix them are decreasing. It is just a matter of time before i586 Tumbleweed is either dropped, or at least retired as an official part of the release (i.e. It will no longer be tested, or at least the test results will be ignored when deciding wheather to release). When that time comes it could be very sudden, we're basically one developers decision away from potential major breakage that could force our hand. And we've been in that state for well over a year, sooner or later i586 Tumbleweeds luck is going to run out. On 7 May 2017 at 04:05, Carlos E. R. wrote:

On 2017-05-07 02:06, Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 20:03]:

...

Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 19:41]:

...

Why is there an i586 of TW, when there is no Leap i586? What's the point of testing it if no one can use it? TW is not "testing", that is factory. And TW is not Leap XX.X.

TW = Factory

...

...

...

why do you think no one can use it?

I thought that TW was the testing place for packages which became part of the release version eventually.

It is, but not directly.

But both TW and Leap are releases. One rolling, the other fixed.

...

...

I guess I am confused.

Your confusion is normal and typical.

...

I guess. I have been using Tw for quite a few years as a productive environment, on 4 boxes. I have leap 42.2 as a server.

factory is the development environment, precursor to Tw. But basically they are the same. many packages appear on factory before release and then to Tw.

Factory is just an internal stage, it is not published. If RB is reading, he will soon tell you that factory does not exist.

The reason that there is no 32 bit Leap is because Leap takes its kernel from SLE, and there is no 32 bit SLE.

SLE derives directly from Factory after much testing and fixing. Leap takes the core from SLE, and adds about 60% from TW.

-- Cheers / Saludos,

Carlos E. R.

(from 42.2 x86_64 "Malachite" (Minas Tirith))

-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Op zondag 7 mei 2017 04:05:09 CEST schreef Carlos E. R.:

On 2017-05-07 02:06, Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 20:03]:

...

Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 19:41]:

...

Why is there an i586 of TW, when there is no Leap i586? What's the point of testing it if no one can use it?

TW is not "testing", that is factory. And TW is not Leap XX.X.

TW = Factory

No, it's not. Factory is the development, TW the rolling release.

...

...

...

why do you think no one can use it?

I thought that TW was the testing place for packages which became part of the release version eventually.

It is, but not directly.

Wrong. It isn't. Not indirectly either

But both TW and Leap are releases. One rolling, the other fixed.

...

...

I guess I am confused.

Your confusion is normal and typical.

...

I guess. I have been using Tw for quite a few years as a productive environment, on 4 boxes. I have leap 42.2 as a server.

factory is the development environment, precursor to Tw. But basically they are the same. many packages appear on factory before release and then to Tw.

Factory is just an internal stage, it is not published. If RB is reading, he will soon tell you that factory does not exist.

The reason that there is no 32 bit Leap is because Leap takes its kernel from SLE, and there is no 32 bit SLE.

That's not the reason. Fact that so far no one has stepped up to take the challenge and do the job is.

SLE derives directly from Factory after much testing and fixing. Leap takes the core from SLE, and adds about 60% from TW.

-- Gertjan Lettink, a.k.a. Knurpht openSUSE Board Member openSUSE Forums Team -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On 2017-05-07 16:26, Knurpht - Gertjan Lettink wrote:

Op zondag 7 mei 2017 04:05:09 CEST schreef Carlos E. R.:

...

On 2017-05-07 02:06, Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 20:03]:

...

Patrick Shanahan wrote:

...

* Richmond <> [05-06-17 19:41]:

...

Why is there an i586 of TW, when there is no Leap i586? What's the point of testing it if no one can use it?

TW is not "testing", that is factory. And TW is not Leap XX.X.

TW = Factory

No, it's not. Factory is the development, TW the rolling release.

Depends on who says it.

...

...

...

...

why do you think no one can use it?

I thought that TW was the testing place for packages which became part of the release version eventually.

It is, but not directly.

Wrong. It isn't. Not indirectly either

In your opinion.

...

The reason that there is no 32 bit Leap is because Leap takes its kernel from SLE, and there is no 32 bit SLE.

That's not the reason. Fact that so far no one has stepped up to take the challenge and do the job is.

That's the official explanation now. The explanation at the time was that SLES had dropped 32 bit support, so there was no 32 bit kernel ready, and it had to be done by somebody. -- Cheers / Saludos, Carlos E. R. (from 42.2 x86_64 "Malachite" at Telcontar)

On 2017-05-07 23:49, Patrick Shanahan wrote:

* Carlos E. R. <> [05-07-17 17:11]:

will make a believer of you if you use factory repos insead of Tw repos.

Oh, I know about that. But most of those are simply TW repos that have not changed the name. -- Cheers / Saludos, Carlos E. R. (from 42.2 x86_64 "Malachite" at Telcontar)

* Carlos E. R. [05-07-17 18:25]:

On 2017-05-07 23:49, Patrick Shanahan wrote:

...

* Carlos E. R. <> [05-07-17 17:11]:

...

will make a believer of you if you use factory repos insead of Tw repos.

Oh, I know about that. But most of those are simply TW repos that have not changed the name.

then you *must* agree that Tw != factory if 1 != 2 then they are different, ie: not the *same* -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri Registered Linux User #207535 @ http://linuxcounter.net Photos: http://wahoo.no-ip.org/piwigo -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Le 02/06/2017 à 15:16, Carlos E. R. a écrit :

On 2017-05-08 00:34, Knurpht - Gertjan Lettink wrote:

why resurrect this :-) I installed TW 32 bits recently on a 12 years old computer I have in my linux user group with 1250 MB ram, with absolutely no problem, and with an other same but with only 512Mb, it was long but manageable. else I would have to use debian. I read also that some applications do not anymore compile easily with 32 bits and this is bad news. Of course I know this hardware is becoming rarer with time, but I also have new hardware with 64 bit processor and windows 32 bits, which is difficult to get linux on it, but this gives me the feeling that 32 bits applications are not to die if windows 32 bits still exists jdd -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On 2017-06-02 15:36, jdd@dodin.org wrote:

Le 02/06/2017 à 15:16, Carlos E. R. a écrit :

...

On 2017-05-08 00:34, Knurpht - Gertjan Lettink wrote:

why resurrect this :-)

Because he said that what I said was false. I had to wait till I could search my archive to prove my recollection was correct. No, I do not want to resurrect a thread about 32 bit been dead or not ;-) However, I keep a 32 bit machine on 13.1, isolated. For my home server I had to purchase new hardware. -- Cheers / Saludos, Carlos E. R. (from 42.2 x86_64 "Malachite" (Minas Tirith))

Le 02/06/2017 à 16:30, Carlos E. R. a écrit :

However, I keep a 32 bit machine on 13.1, isolated. For my home server I had to purchase new hardware.

of find one in a trash can (I did once :-)) jdd -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On 02/06/17 01:15 PM, Carlos E. R. wrote:

On 2017-06-02 18:00, jdd@dodin.org wrote:

...

Le 02/06/2017 à 16:06, Per Jessen a écrit :

...

jdd@dodin.org wrote:

...

Of course I know this hardware is becoming rarer with time, but I also

Running 32bit in virtual environments is useful too.

could you elaborate? other than test box?

Because the 32 bit virtual machines use less resources than the 64 bit ones, with the same software.

:-) Maybe we should resurrect some 16-bit Linux :-) -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Le 02/06/2017 à 19:15, Carlos E. R. a écrit :

On 2017-06-02 18:00, jdd@dodin.org wrote:

...

Le 02/06/2017 à 16:06, Per Jessen a écrit :

...

jdd@dodin.org wrote:

...

Of course I know this hardware is becoming rarer with time, but I also

Running 32bit in virtual environments is useful too.

could you elaborate? other than test box?

Because the 32 bit virtual machines use less resources than the 64 bit ones, with the same software.

is that significant? ressources are cheap nowaday. I would like to see numbers jdd -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

jdd@dodin.org wrote:

Le 02/06/2017 à 19:15, Carlos E. R. a écrit :

...

On 2017-06-02 18:00, jdd@dodin.org wrote:

...

Le 02/06/2017 à 16:06, Per Jessen a écrit :

...

jdd@dodin.org wrote:

...

Of course I know this hardware is becoming rarer with time, but I also

Running 32bit in virtual environments is useful too.

could you elaborate? other than test box?

Because the 32 bit virtual machines use less resources than the 64 bit ones, with the same software.

is that significant? ressources are cheap nowaday. I would like to see numbers

It is significant. I can run a lot more 32bit processes on a 32bit oS, than 64bit processes on a 64bit oS on the same hardware. With 32bit, you get more bang for the buck - under the right circumstances. For e.g. virtual hosting, you get more xen domUs, but also for a postfix mailserver running thousands of smtpd processes. Quite possibly similar for a dovecot machine with thousands of users. I have some 40 machines running mail servers - typically 4 cores with HT, 12G memory, it varies. In 32 bit they can easily run 2000 smtpd processes each, in 64bit a lot less. I haven't actually tried or measured, just rule of thumb. There have been suggestions to just run postfix 32bit on 64bit oS, I'll be trying that out at some point, when we move to Leap in production. -- Per Jessen, Zürich (20.6°C) http://www.hostsuisse.com/ - virtual servers, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

jdd@dodin.org wrote:

Le 02/06/2017 à 20:16, Per Jessen a écrit :

...

It is significant. I can run a lot more 32bit processes on a 32bit oS, than 64bit processes on a 64bit oS on the same hardware.

if so, may be somebody have to sponsor a 32 bit team for tw :-(

Well, yes and no. For virtual hosting, at the level where it matters, the problem is easily solved with hardware, as I think you also suggested. FOr me runing thousands of smtpd processes, I think the better solution is to run a 32-bit postfix on 64bit oS. -- Per Jessen, Zürich (19.2°C) http://www.cloudsuisse.com/ - your owncloud, hosted in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

jdd@dodin.org wrote:

Le 03/06/2017 à 09:22, Per Jessen a écrit :

...

Well, yes and no. For virtual hosting, at the level where it matters, the problem is easily solved with hardware, as I think you also suggested. FOr me runing thousands of smtpd processes, I think the better solution is to run a 32-bit postfix on 64bit oS.

but may be you wont anymore have 32 bits postfix...

I don't have one right now, there is no 32-bit postfix package being built. There are other 32bit packages though, so it's only a matter of creating one. -- Per Jessen, Zürich (21.3°C) http://www.dns24.ch/ - your free DNS host, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On 02/06/17 01:40 PM, James Knott wrote:

On 06/02/2017 01:25 PM, Anton Aylward wrote:

...

On 02/06/17 01:15 PM, Carlos E. R. wrote:

...

On 2017-06-02 18:00, jdd@dodin.org wrote:

...

Le 02/06/2017 à 16:06, Per Jessen a écrit :

...

jdd@dodin.org wrote:

...

Of course I know this hardware is becoming rarer with time, but I also Running 32bit in virtual environments is useful too.

could you elaborate? other than test box? Because the 32 bit virtual machines use less resources than the 64 bit ones, with the same software. :-) Maybe we should resurrect some 16-bit Linux :-)

I'm sure there's some 4 bit Intel 4004 CPUs in service somewhere. ;-)

https://en.wikipedia.org/wiki/Intel_4004

Ok, lets cut to the chase: https://en.wikipedia.org/wiki/1-bit_architecture And yes, you could have run a form of "unix" on one of them. Back in the mid 1970s I worked for a short while with a MIL-SPEC rated 1-bit CPU. It was supposed to be a 16-bit 'computer" in that it had 16-bit words and addressing, but the "mill" was 1-bit. It was simple and !FAST!, fast enough to keep up with the "real" 16-bit processors of that day, the early Intel 8086, the Zilog Z-8000 and the early Motorola, as well as some fringe machines that implemented Pascal or FORTH directly. Strictly speaking, it was a 'bit-serial' machine, the type of architecture that was common in the 1950s when hardware was !expensive!. Sorry, I can't recall it's name. it was a evaluation unit and in a couple of days I decided it was a no-go. -- A: Yes. > Q: Are you sure? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting frowned upon? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Anton Aylward wrote:

On 03/06/17 05:11 AM, Per Jessen wrote:

...

Per Jessen wrote:

...

jdd@dodin.org wrote:

...

Le 03/06/2017 à 09:22, Per Jessen a écrit :

...

Well, yes and no. For virtual hosting, at the level where it matters, the problem is easily solved with hardware, as I think you also suggested. FOr me runing thousands of smtpd processes, I think the better solution is to run a 32-bit postfix on 64bit oS.

but may be you wont anymore have 32 bits postfix...

I don't have one right now, there is no 32-bit postfix package being built.

... being built for openSUSE 64bit, i.e. for Leap.

What about "Docker" as an alternative to a VM? I'm not knowledgeable but the couple of magazine style articles I've read lead me to believe that Docker offers a light-weight alternative to a VM.

It's a possibility, I don't know about Docker to say if it's suitable for virtual hosting. I looked at Linux containers some years back, they were okay for e.g. separating webhosts.

And is it really this simple? https://www.slideshare.net/dotCloud/postfix-tuto4

Quite likely, but way overkill for running 2000 smtpd processes. -- Per Jessen, Zürich (24.2°C) http://www.hostsuisse.com/ - virtual servers, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org