Fedora Extras packaging beta software into production repos, why?

[Nicolas Mailhot]

Le lundi 30 octobre 2006 à 16:18 +0100, Michael Schwendt a écrit :

> There is a simple solution: --> All or nothing. <--
> 
> [...]
> 
> Don't build pieces of a package-set just because they are approved
> already. Keep all dependencies in FE-ACCEPT state until the other packages
> in FE-REVIEW are approved, too.

It's simple but stupid.

1. The dependencies won't conflict less when released later, you're only
maximizing the pain by releasing packages which never had exposure
separately.

2. You're needlessly maximising bureaucratic inertia. Dependencies often
have worth by themselves, either for other FE packages or for people
running unpackaged stuff in the wild. For example a few years ago I
packaged most amavisd-new deps which certainly helped when amavisd-new
was packaged by someone else. Your rule would have forced the
amavisd-new packager to start from scratch.

3. You're rewarding the people who don't get their packages reviewed and
penalise people who follow the appropriate process.

At most, 3rd party maintainers could be allowed to ask in a review for a
grace period after a package is approved (this grace-period being
non-reconductible and with a fixed maximum). I say at most because as in
any FLOSS project the people doing the work get to decide, and in this
case the person submitting his package is the one doing the work.

This is the same for Fedora, kernel patches, whatever. And it's *fair*.
You want something packaged a certain way you do the packaging first.
You build a userbase outside of the normal process with packages that
can't be imported as-is (because they fail review rules?) Tough luck,
not FE's problem, you knew the risks.

<rant about bullying/>

-- 
Nicolas Mailhot