Kevin Fenzi <kevin <at> scrye.com> writes:
> Uh... 2 out of 579 is hard to gather a general opinion from.
4 if you count Conrad and me. Out of the few ones who have reacted to those
threads.
> In addition, many folks might have voted for FESCo members to vote for
> the right thing, and not via a popularity contest.
"popularity contest" is an interesting paraphrase for "democracy"...
> No. Only if such a 'draft guideline' was compelling enough for the
> Packging committee to decide that packages in that area should wait for
> it to be finalized. This is not automatic... it requires the packaging
> folks to decide so.
Has FPC actually been consulted in this matter?
> Why are we waiting months? Why aren't people helping finalize
> guidelines? Because there is no incentive to do so?
The only thing this resolution is an incentive for is not to mention guidelines
at all, even in passing, in order not to "give" FPC "ideas", i.e. the exact
opposite of what you're aiming at.
> Once packages have entered the collection it is MUCH harder to change
> them and make them conform to a good packaging guidelines.
Then that's the real problem to solve. Guidelines will always change. Even the
generic ones change, see e.g. License tags. It should be possible for some
maintainers (to be defined, maybe FPC members) to globally fix packages for
guideline issues, just as rel-eng is allowed global write access to bump
packages for mass-rebuilds.
The current "solution" of holding up packages until everything's all set is the
exact opposite of the "release early, release often" principle which is so
common in Free Software (and in particular in the bazaar-style "Open Source
development model") and which is supposed to be part of Fedora's objectives.
> Is it really that hard to come up with good packaging guidelines for
> new tech?
The turnaround times for current drafts clearly show it is.
Kevin Kofler
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
