Mailinglist Archive: opensuse-project (88 mails)

< Previous Next >
Re: [opensuse-project] openSUSE 13.1 - Preparation for Evergreen
  • From: Richard Brown <RBrownCCB@xxxxxxxxxxxx>
  • Date: Sat, 12 Dec 2015 18:39:05 +0100
  • Message-id: <CAA0b23xFA3y2tbEVa9JdzJau167107=zvNqM9Yv5NsrW5BUaiw@mail.gmail.com>
Hi Axel,

On 12 December 2015 at 17:16, Axel Braun <axel.braun@xxxxxx> wrote:

Sorry to say, but this is quite an arrogant answer. I will tell you why:

I'm hurt by you saying this. My statement is not intended to be
arrogant, but a reflection of the reality as I see it.

I'm maintaining packages and demo-CD for some medical software, which has some
interests in the developing countries around the world. It can be used by
single doctors, but as well in Hospitals.

Many of those guys are happy to have some hardware at all, no matter how old
it it, and it is most likely 32bit. And there is no point (-> money <-) in
walking around the corner to the next hardware store to buy an upgrade.

The 32-bit demo on 13.1 has by far the most downloads, the template on Studio
is IMHO more reliable than 13.2.

The openSUSE Project and the distributions we produce have a broader
application than your medical software in developing countries

Looking at the big picture, 32 bit downloads have not been a majority
since openSUSE 12.2 which was released in Jan 2012 and reached end of
life in 2014

12.3 released in March 2013 - Support Ended Jan 2015
64-bit downloads 45521
32-bit downloads 28196
Proportion of 32-bit downloads = 37%

13.1 released in November 2013 - Support is what we're discussing here
13.1 64-bit downloads 60138
13.1 32-bit downloads 39836
Proportion of 32-bit downloads = 39%

13.2 released in November 2014
13.2 64-bit downloads 103429
13.2 32-bit downloads 27727
Proportion of 32-bit downloads = 20%

Your use case may buck this trend, but the Projects decision to not
support the 32-bit Intel architecture in the future is supported by
the numbers.

Though ultimately the numbers are somewhat meaningless - we're a
volunteer organisation, we support whatever our volunteers,
collectively, decide to support

And I think it's safe to say that the openSUSE community, as a whole,
has demonstrated a lack of desire to support 32-bit Intel going
forward - if there was volunteers willing to do it, we'd be doing it
:)

I don't say this to prevent or discourage anyone from stepping up and
doing the work despite this.
But I think it is the correct and honest thing to do is make the
situation very clear, so if you or anyone else decides to do the work
you realise the scale of the mountain you need to climb.

And either encouraging repository maintainers to continue to support
an obsolete architecture on an 'out of main support' distribution, or
doing all that work yourself, is a lot of work.. but that's what needs
to be done if what you are asking for is going to be viable..

Thats why I would like to transfer it to
Evergreen, and thats why I'm asking for a build target fo OBS.
Of course it is up to each maintainer to support this, but we have at least
the basis on OBS.

I do not see how a Build Target for Evergreen is any better, or worse,
than a Build Target for openSUSE:13.1

Therefore, as creating a Build target for Evergreen is more work than
keeping the existing 13.1 Build Target, I really don't see the point..

If Evergreen had any sign of continuing after 13.1 then I'd see the
benefit, but Evergreen 13.1 is expected to be the last Evergreen
because of Leap, so I really do not see why we should be doing this.

Giving away the 32bit market is not a bright idea at all, but a consequence of
the decision to reduce efforts on SUSE side and move everything to SLES basis.
I feel it will reduce the overall importance and acceptance of openSUSE. But
thats a different story.

Ending 32-bit support is a consequence of the decisions by the current
maintainers of the openSUSE distributions to not bother doing it for
Leap.

It is a consequence of no other maintainers in the openSUSE community
volunteering to do that work

Please do not blame the corporate SUSE for a decision which we, the
openSUSE Project, are collectively responsible for.

It's within our power to change, if there are people within this
community willing to do the work, build, test, and maintain a 32-bit
distribution.

That said, I'm not going to help, I'd like to see 32-bit Intel die,
and as I do not have any 32-bit hardware available I am not in a
position to help even if I felt differently on this topic
--
To unsubscribe, e-mail: opensuse-project+unsubscribe@xxxxxxxxxxxx
To contact the owner, email: opensuse-project+owner@xxxxxxxxxxxx

< Previous Next >
Follow Ups