Mailinglist Archive: opensuse-project (57 mails)

[Martin Schlander <martin.schlander@xxxxxxxxx>]

Torsdag 18 september 2008 00:10:43 skrev Boyd Lynn Gerber:
> On an other list, it was suggested that this list would be the best place
> to have a discussion on how we as a community may work together to make
> openSUSE the best and most stable dstribution.

"Best" is pretty vague, people have very different criteria for what they 
think is good. I don't think I'd use openSUSE if I didn't think it was the 
best distro already.

We also have to ask ourselves if we really want openSUSE to be the most 
stable. You can't have both the latest greatest bleeding edge technology and 
rock solid stability. 

Maybe it would be better to start by aligning expectations for openSUSE. 

In my opinion openSUSE delivers a very healthy balance of up-to-date 
technology and rather good stability. This is what I expect and I think that's 
what _should_ be expected. You shouldn't expect a conservative, enterprise 
grade product, with zero risk-taking and where the release is delayed for 
months until it's just right etc. 

When you accept the concept of openSUSE, the limitations in resources, and 
that openSUSE is not highest priority for Novell, then you can start to 
discuss how to make it better, realistically within these boundaries.

> Some things AJ suggested are...
>
> "IMHO the things the items where most help by everybody is needed are:
>
> * early testing with good bugreports in bugzilla.novell.com
> * help in bug triage so that the engineers can work on real bugs
> * fixing bugs

AJs points are good, but maybe we should wonder how to facilitate these 
things. For example it seems to me that development releases are often rushed 
out the door in a horribly broken state, just to meet the roadmap - if alphas 
are hardly installable, that's not conducive to early testing. Maybe spending 
one extra day on alpha releases would lead to substantially more early 
testing.

> I agree with this, but I think we should look at things.  The issues I see
> with the 11.0 release is Wireless Network issues, KDE 4.0.X where from
> what I see using 4.1.1 it would have been a lot better to have had it.  

Yes, many problems are created "by choice". At least I often have the feeling 
when you look at the roadmap and the major new features of the next release, 
you already pretty much know in advance, what will be the FAQs on IRC and 
forums for the next 7-9 months.

When you hear "GNOME will use pulseaudio and switch to packagekit updater" or 
"only KDE 4.0 will be available on livecd" or "we'll ship Firefox beta", you 
know there will be problems, and no amount of testing or bug reporting, 
however early, will prevent that - cuz the stuff is so immature upstream 
that's it not feasible for openSUSE developers to make enough of a difference.

But how do you avoid this? By always being conservative? I don't think so, 
that's what SLE, RHEL, CentOS or Debian are for. Delaying release for 2-4 
months to fix bugs or wait for upstream to do it?  Not realistic.

I think all you can do is work with the relevant teams and try to help them 
make the right decisions on a case by case basis. Of course often times, these 
types of decisions are a matter of being "damned if you do, and damned if you 
don't".

Of course having a little bit of consistency across releases would help these 
matters also. Hopefully our era of ever changing package management backends 
and updater applets etc. is over, and we're moving into a period of less 
zigzag and more incremental improvements holding a more steady course.
---------------------------------------------------------------------
To unsubscribe, e-mail: opensuse-project+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse-project+help@xxxxxxxxxxxx