Mailinglist Archive: opensuse-factory (1134 mails)

< Previous Next >
Re: [opensuse-factory] Calling for a new openSUSE development model
  • From: "M. Edward (Ed) Borasky" <znmeb@xxxxxxxxx>
  • Date: Thu, 14 Jun 2012 14:42:10 -0700
  • Message-id: <>
On Thu, Jun 14, 2012 at 1:46 AM, Stephan Kulow <coolo@xxxxxxx> wrote:

It's time we realize delaying milestones is not a solution. Instead,
let's use the delay of 12.2 as a reason to challenge our current
development model and look at new ways. Rather than continue to delay
milestones, let's re-think how we work.

openSUSE has grown. We have many interdependent packages in Factory. The
problems are usually not in the packages touched, so the package updates
work. What's often missing though is the work to fix the other packages
that rely on the updated package. We need to do a better job making sure
bugs caused by updates of "random packages" generate a working system.
Very fortunately we have an increasing number of contributors that
update versions or fix bugs in packages, but lately, the end result has
been getting worse, not better. And IMO it's because we can’t keep up in
the current model.

I don't remember a time during 12.2 development when we had less than
100 "red" packages in Factory. And we have packages that fail for almost
five months without anyone picking up a fix. Or packages that have
unsubmitted changes in their devel project for six months without anyone
caring to submit it (even ignoring newly introduced reminder mails).

So I would like to throw in some ideas to discuss (and you are welcome
to throw in yours as well - but please try to limit yourself to things
you have knowledge about - pretty please):

1. We need to have more people that do the integration work - this
 partly means fixing build failures and partly debugging and fixing
 bugs that have unknown origin.
 Those will get maintainer power of all of factory devel projects, so
 they can actually work on packages that current maintainers are unable
2. We should work way more in pure staging projects and less in develop
 projects. Having apache in Apache and apache modules in
 Apache:Modules and ruby and rubygems in different projects may have
 appeared like a clever plan when set up, but it's a nightmare when it
 comes to factory development - an apache or ruby update are a pure
 game of luck. The same of course applies to all libraries - they never
 can have all their dependencies in one project.
 But this needs some kind of tooling support - but I'm willing to
 invest there, so we can more easily pick "green stacks".
 My goal (a pretty radical change to now) is a no-tolerance
 strategy about packages breaking other packages.
3. As working more strictly will require more time, I would like to
 either ditch release schedules all together or release only once a
 year and then rebase Tumbleweed - as already discussed in the RC1

Let's discuss things very openly - I think we learned enough about where
the current model works and where it doesn't so we can develop a new one

Greetings, Stephan

To unsubscribe, e-mail: opensuse-factory+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse-factory+owner@xxxxxxxxxxxx

Project management, software or otherwise, is a well-established
discipline. Yes, sometimes stuff happens and dates slip. But you work
within the constraints of the time, money, people and hardware you
have. My specific recommendations for 12.2 are:

1. Fix all the show-stopper bugs. If that means dropping back to an
older KDE because the one on the beta is sending people to the KDE bug
tracker, then do that. If we have to delete packages to get the GNOME
and KDE LiveCDs to fit on 700 MB media, delete the least popular
packages until you have the sizes down. If GRUB2 can't be made to
work, drop back to GRUB1.

2. Slip the date just enough to deliver a release free of critical
distro-killing bugs. If that means axing features, axe them.

Going forward, I proposed a three-layer strategy.

1. Switch to a Debian-like model for servers. Have a stable server
core (kernel, bootloader, installer, LAMP stack, PostgreSQL,
Perl/CPAN, Ruby/Rails/Sinatra, Python/Django and WebYast). This should
look like Debian stable except a *lot* smaller. It might even fit on a
220 MB LiveCD and certainly should fit on a CD. Release this only when
it's solid and stable and supply security and bug fixes via updates.
This should be something that can compete with CentOS, Debian stable
and Ubuntu Long-Term-Support in level of available support,
documentation, security, performance and stability. Bonus points for a
Platform-as-a-Service like Cloud Foundry's MicroCloud or Red Hat's
OpenShift Origin. But don't have a fixed "release schedule" - release
when all software quality gates have been met.

I can go over to SUSE Studio and build the media - minus the PaaS - in
half an hour. I'm guessing nearly all of the QA and release
engineering can be automated using existing infrastructure if we
reduce the scope to the most commonly used and mature server stacks
and change the commitment / social contract to Debian stable's "we
will ship no distro before its time." ;-)

2. Move everything else: compilers, desktops, applications, advanced
server technologies like Node.js and NoSQL databases and
virtualization infrastructure - to a rolling release similar to
Tumbleweed or Gentoo. And I think the kernel needs to be *out* of
Tumbleweed. If you're doing kernel testing / development, you're not a
user any more. ;-) I've pissed away *way* too many hours recompiling
VMware Workstation, VirtualBox and NVidia driver modules.

3. Acquire our own community instance of SUSE Studio and make it the
main distribution channel. It's extremely well-constructed and
integrated with OBS, it seems to be stable, and the current instance
seems little-used. You can make *anything* there - except a PaaS. ;-)

Twitter: Computational Journalism Server

Data is the new coal - abundant, dirty and difficult to mine.
To unsubscribe, e-mail: opensuse-factory+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse-factory+owner@xxxxxxxxxxxx

< Previous Next >
This Thread