KDE 4.1 planning
Matt Rogers
mattr at kde.org
Tue Jan 8 02:55:07 CET 2008
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Jan 7, 2008, at 7:49 PM, Sebastian Kügler wrote:
> On Tuesday 08 January 2008 02:14:02 Matt Rogers wrote:
>> Why exactly do the distros matter when planning our releases? (I know
>> I've asked this and it's been explained before, but I still don't get
>> it. Perhaps I never will.)
>
> Distros are the ones that get our software to the user. KDE is a 'raw'
> product, we need the distros' integration work (adding an OS, that
> kind of
> details ;-)),
>
> There's a mutual dependency between us and the distros, we need
> them to get to
> software shipped for users, they need us because we provide raw
> material for
> their finished 'product'. Aligning schedules makes it easier for
> distros to
> ship new products based on new KDE versions.
>
>> If you're just proposing a 6 month release
>> schedule because that's what the distros do, well, then that's a load
>> of bunk, IMHO. If you're proposing a 6 month release schedule because
>> you think it provides good sustainable pace for the majority of us,
>> then that's an actual reason.
>
> Both applies. A 6 month cycle makes it easier to get our software
> also to our
> fellow developers. The KDE4 platform is in pretty good shape, so we
> probably
> don't need too much integration work (in the range of 8 - 6 weeks
> to settle
> down a 4.x). We will be building software on top of that, and not
> change the
> underlying system too much (so that we need lots of time before the
> code is
> releasable again). That gives us a good 4 months from release to
> next feature
> freeze. For larger changes, it basically means: "Work in a branch
> until you
> can get your code stabilised within 2 (worst case) to 6 (best case)
> months."
> That sounds reasonable to me.
>
> 6 months also keeps the amount of changes surmountable. With a
> stable basis,
> we can release more often, so the integration work (fixes in A with
> features
> in B to work together seamlessly) is done more 'in the process'. Less
> powerplanting, which I think is good.
> --
This is the best explanation as to why distros matter in our release
planning that I've read. Thanks for that.
I'll draw up a proposed schedule later tonight
- --
Matt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)
iD8DBQFHgtf7A6Vv5rghv0cRAgR1AJ9vICHG89TMkgTfgoXjEALuDhqBjACgneuv
dhscJ01LA5FkIjk9DXfEcWc=
=W1CW
-----END PGP SIGNATURE-----
More information about the release-team
mailing list