On Tue, May 02, 2000 at 02:36:24PM -0400, Andrew Pimlott wrote:
> On Mon, May 01, 2000 at 12:46:16AM +1000, Anthony Towns wrote:
> > Why would we want to release woody less than a year after potato's
> > *frozen* let alone released?
> > We've already tried this with slink, less time spent between freezes
> > doesn't lead to less time within a freeze. All it does is cut out x
> > months of development time.
> It cuts out x months of _Debian_ development time.
Debian development includes:
* making sure upgrades to new upstream releases work more or
less seamlessly (writing wrappers around incompatible formats, eg)
* development of apt, dpkg, debconf, debbugs, package pools, etc
* cross package conformance (policy, emacsen-common, java-common,
doc-base, etc)
At least ocassionally it includes fixing up upstream code too: fixing
portability bugs, writing manpages, redoing configuration code, automating
the build process and so on.
> Remember, Debian's primary function is as an
> integrator; saying, "We need more time developing the Debian
> infrastructure, so let's hold back hundreds of improved packages"
> seems to be giving oneself airs.
As an integrator, saying "we need more time to properly intergrate"
is giving ourselves airs? Interesting conclusion.
> I don't mean to belittle the value or excellence of the Debian
> infrastructure. But perhaps internal Debian projects should be
> thought of as (relatively) orthogonal to the releases. Develop them
> in a way that is not coupled to the release cycle.
For projects mainly worked on by Debian developers that's difficult to
do though, since the developers legitimately want to spend their time
getting the release out, since their testing base is often now focussing
on old code (ie, frozen) rather than keeping up with any new uploads
they might make (ie, unstable), since some of the autobuilders don't
recompile packages for other architectures, and so on.
And note that Debian specific projects often don't really see the light
of day until a Debian release: thanks to SourceForge, Slackware, Red Hat,
Caldera, SuSE, and so on, non-Debian-specific projects have plenty of
chances (and reasons) to decouple themselves from Debian.
Cheers,
aj
--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG encrypted mail preferred.
``We reject: kings, presidents, and voting.
We believe in: rough consensus and working code.''
-- Dave Clark
Attachment:
pgp5aYa978l3e.pgp
Description: PGP signature