Re: Old versions of GNOME [was: Re: gtk 2.8 for gnome 2.12]

[Luis Villa <luis villa gmail com>]

On 7/22/05, Glynn Foster <Glynn Foster sun com> wrote:
> Heya,
> 
> > Please bear with me along the following rant.
> 
> Okay, so this rant covers a lot of ground, but I have specific comments
> from a Sun perspective -
> 
> o As a Sun developer, I'd much rather the community focus on
>   churning out the next release of GNOME. Which is pretty much what
>   the average hacker wants to do, right - be innovative, develop
>   new features and generally get the desktop moving forward. Bug fixing
>   gets boring, and bug fixing on stable branches even more so ;)
> 
> o I think it should be up to the various distributions to put their
>   bug fixing patches upstream, and onto the branches ASAP - so that
>   other distributions can also use them. Let's face it - there's
>   no value add in bug fixes, and if they don't get pushed upstream,
>   it makes GNOME look bad rather than other distributions. I'd
>   very much welcome a 'free for all' on the stable branches, past
>   the 2 or 3 official releases we do.


Once upon a time Nat and I talked about having a centrally
located/funded coordinator for the distros, known to be reasonably
neutral, whose job it was to track bugs[1] in older versions, test
patches against multiple versions, etc.- basically do the
coordination/testing/release work that would solve some of the
problems Federico  very correctly highlighted. I still think it would
likely be a good investment for the distros (who will all soon
maintain multiple old versions that none of the developers want to
touch) to pool some money and hire such a person.

[1] At the time (right after the glow of 2.0) I was interested; most
definitely am not these days, though I'd certainly lend my advice on
the bug-tracking side.

> o I'm trying to push a change in development process within Sun, so
>   that we can concentrate our core development on HEAD as much as
>   possible. We've been kicking this around internally for the past
>   couple of years, and now with our focus on OpenSolaris, I think it
>   should be more feasible to do than previously. As an added bonus,
>   we hope to be able to throw QA resources into that as well. All
>   this is going to take time though, and won't happen overnight.

I have seen that mentioned in some blogs, and I agree it would be
great to see- I hope it works out.

Luis