To quote Robert McQueen <robot101@debian.org>,
> Just read the thread 'Intent to NMU mozilla...' and thought I
> should raise a few points. Firstly, this entire thing has not yet
> had any bearing on any release of Debian. Myth has clearly stated
> that he will resolve it in time for woody release, so I don't think
> there is as much urgency as we all think.
Like most things, it depends on how you look at it. Most people(all?)
will admit that the longer the package is in Sid or Woody, the
better-tested it'll be. So time *IS* a factor. And since the current
packages are six months old, very few people are inclined to say, "sure,
we'll give you another week." Mostly due to the number of "weeks" that
have already been given.
> Secondly, Myth is paid by Ximian to package mozilla anyway, and it
> makes very little sense to duplicate effort and make packages
> alongside his, in some kind of bizzare competition. He has explained
> why he hasn't uploaded anything, but that doesn't mean he isn't
> working on it. I have been keeping myself informed with his progress
> packaging 0.9.1 over the past few days. He has various patches and
> bugfixes added (being in communication with the Mozilla hackers), but
> is having problems with PSM causing Mozilla to hang.
Duplication of effort is indeed a very bad thing. But it's not a sin,
and nobody goes to jail for it. Duplication of effort is actually
warranted, too, depending on the goals of those who are duplicating the
effort. I believe Kitame's current goal(to get a non-broken,
crypto-enabled, recent release of Mozilla into Sid for testing) is
laudable, and warrants duplication of effort, since Myth has not said
publicly, one way or another, what he will do. Sure, Myth has asked
Kitame to wait for the legal ruling(which is, of course, wise), but he
has not commited himself, one way or the other.
> But more importantly, Myth is working on these packages, and will
> work on them anyway. He will certainly upload full packages if/when
> the crypto ruling comes back positive, otherwise he will upload
> crypto-removed packages to main, and work on something for non-US
> that doesn't waste a lot of archive space with duplicate Mozilla
> source.
Others have already commented on this. I'll just add my voice to the
din: If Mozilla with crypto has to go into non-US, so be it. Non-US is
there for a reason. If Myth doesn't like Debian Policy, he can move to
get it changed. But *please*, only after he's uploaded a non-broken
Mozilla.
> That said, it is my understanding that Kitame's interests in a newer
> Mozilla stem (not to say that there are not other reasons) from being
> able to have Nautilus built with a new and reliable Mozilla engine
> embedded. Others want to proceed with packaging of galeon, skipstone,
> light, etc, and others just want to see the latest and greatest
> Mozilla in the archive to play with. So I have a solution to propose.
a) Kitame's intentions for doing the NMU are almost irrelevant at this
point. *Someone* needs to get newer packages in. End of story. b) He's
willing to package *and* release it. And he's doing so. I don't know of
anyone else who is. Isn't that good enough?
> In my opinion, the least confrontational way for this situation to
> be resolved would be a compromise. Myth could (and I'll help if at
> all possible) prepare and upload crypto-removed packages to main
> ASAP. This means everyone gets a newer Mozilla, and the
> Mozilla-based packages can update as necessary. If crypto can't go
> into woody, these packages will remain. If it can, then Myth can
> upload a full set of Mozilla packages, which as far as we can tell
> will not break the Mozilla-based ones.
We should not be worried about the least confrontational way to solve a
problem. We should be concerned with the best solution. While uploading
a crypto-removed Mozilla to main, and a crypto-enabled Mozilla to
non-US/main would definetly be a great solution, Myth has not said he
will do either(at least not publicly).
Anyhow, I've had my chance to speak. Thanks for the opportunity :)
David Barclay Harris, Clan Barclay
Don't panic.
Attachment:
pgptQZqb34xOR.pgp
Description: PGP signature