Re: Should we divide Debian to usable and unusable
On Sun, Jun 04, 2000 at 11:10:59PM +0200, Robert Bihlmeyer wrote:
> > We simply shouldn't let all the
> > 0.0.0.1c6-prealpha-johnny-be-good-CVS20000604 versions get into
> > stable. That's achieved by filing RC bugs -- they either get fixed,
> > or packages get removed.
> Indeed. Sami can file RC bugs against the packages he deems unusable.
> I hope that version number is not the only criteria, though...
Of course. (I thought it was clear from my example of version what the
program was like)
> Still, by packaging unusable software, one probably makes the release
> manager's life harder. It would be nice if developers marked (with RC
> bugs?) their package as "test balloon, should not go into stable yet"
> so that they are not even considered - that would pull the plug out of
> "why is ... removed, it works for me" whines.
I have seen a few bug reports against ftp.debian.org from maintainers saying
basically `my package isn't meant to be released, please remove it but leave
it in unstable', which I thought was rather considerate of them. We should
all try to do that when each freeze comes, no matter how biased we might be,
or how much time/thought it takes to decide such a thing.
I'm currently tempted to get my package "hatman" removed (perhaps that won't
be possible due to the state of the freeze). It's a Pacman-clone game that
runs under GGI (which makes it special enough, that's why I packaged it), it
only has one game level included, upstream development crawls (although the
author is responsive when it comes to fixing bugs I forward), and I have one
report that it doesn't behave properly on Alpha. I'd appreciate opinions, if
Digital Electronic Being Intended for Assassination and Nullification