A tale of fixed release schedules

[Javier]

On 3/17/06, jorge o. castro <jorge at whiprush.org> wrote:
> javier wrote:
> > The blog points out to other delays (fedora and suse) so I believe it's saying
> > something like fixed released are VERY difficult to accomplish and don't
> > believe anyone saying so.
> >
>
> People have been questioning the validity of time-based releases in
> software development for a long time. You'll find plenty of arguments
> for either opinion on the GNOME lists (if you go back far enough).
>
> I think that GNOME has proven that time based releases can work, and
> that there's a big advantage for an upstream project in delivering
> releases on a predictable schedule that vendors like Ubuntu can count on
> to ship.
>
> One of the major factors that made me love Ubuntu right off the bat was
> the concept of time based releases. The idea of delaying a release after
>   three successful time based releases worried me at first. But after
> attending the #ubuntu-meeting's and getting feedback from people who are
>   very much used to this kind of release management, I have come to
> following conclusions:
>
> a) sabdfl proposed a change to the schedule, and proposed two IRC
> meetings. Both meetings were well attended, and great care was taken to
> hear every concern of those community members willing to state an
> opinion.[1] Look at other delays by other distributions. You're just not
> going to find that level of community involvement and open communication
>  in other projects.
>
> b) The folks on the TB took the comments from everyone and made a
> decision. It's clear that many lessons where learned during this cycle
> on how to approach a "dapper type" long lived release. I'm sure in
> hindsight there's about 10,000 things that everyone wished they could
> have done better. The important thing is that Ubuntu now has identified
> these problems and is taking measures to fix them.
>
> and
>
> c) as Jerome mentioned earlier in this thread, "Best to do is focus on
> signal instead of noise." Signal in this context means a sweet release
> that people will depend on for years.
>
> Like everyone else, I too have my own pet features and bugs that didn't
> make the Dapper cycle, but hey, there are limited resources and plenty
> of bugs to fix. This week has been filled with "the sky is falling"
> negativity since the delay proposal. I'm sure it gets news sites tons of
> web hits. I think quality is worth the extra 6 weeks ... if this Beineri
> dude has a problem with that, then he's just going to have to learn to
> get over it. :)
>
> --
> jorge
>
> [1] Paul Sladen covered this transparency here:
> https://lists.ubuntu.com/archives/sounder/2006-March/004876.html
>
>
>
>

I also think dapper will rock and people will forget this delay. And I
also believe the delay is needed (specially for kubuntu, xubuntu)
since dapper needs to rock for a long time (3 and 5 years).

I would love if KDE had this predicable release schedule... that's one
thing GNOME is better at and I imagine it was the reason number one
for sabdfl to choose it as the default DE.

The god thing about this is that on a 6 weeks time from the 20th of
april no one will remember this situation...

--

Javier