Gentoo Archives: gentoo-dev

From: Chris Gianelloni <wolf31o2@g.o>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] Policy for retirement of old gentoo 'versions'
Date: Fri, 02 Jul 2004 21:34:20
Message-Id: 1088804249.9271.55.camel@localhost
In Reply to: Re: [gentoo-dev] Policy for retirement of old gentoo 'versions' by Dylan Carlson
1 On Fri, 2004-07-02 at 17:06, Dylan Carlson wrote:
2 > On Friday 02 July 2004 4:29 pm, Chris Gianelloni wrote:
3 > > A "tested profile" would also have to include specific versions,
4 > > otherwise there is no way that a person could properly certify the
5 > > validity of the test.
6 >
7 > I agree. The profiles only list ~70 packages and those versions aren't
8 > pinned. Although maybe they should be. The difference between the
9 > versions in a tested configuration/profile and what ends up getting
10 > installed later should include security updates (backported security
11 > fixes) -- which is not something we do right now...
12
13 ...and I doubt that we ever will. Gentoo tries to remain as much like
14 the upstream packages as possible, which means we're more likely to
15 require an upgrade than to back-port a patch. This is the exact reason
16 why any plans for an enterprise version of Gentoo all focus on being a
17 separate project from Gentoo proper. I know for a fact that I don't
18 want to waste the precious development time that I have doing the
19 mundane task of back-porting patches to some old version of a package
20 that I've long since forgotten.
21
22 > My point is that I believe we could address this (at least in part) by
23 > pinning versions in profiles, and having repoman block commits that
24 > attempt to remove ebuilds that are required by a profile. It's not a new
25 > idea. This, instead of branching CVS. Although I'm not opposed to that
26 > idea either, but IIRC some devs are...
27
28 Pinning versions in the profiles sounds pretty cool, but it turns
29 *every* package maintainer and arch maintainer into a profile
30 maintainer, which I think is a bad idea. It also bloats the portage
31 tree, since there would be multiple versions of every ebuild, compared
32 to the one or two for most packages that we have now. I still think
33 that the "pinned" tree should be a separate branch.
34
35 --
36 Chris Gianelloni
37 Release Engineering QA Manager/Games Developer
38 Gentoo Linux
39
40 Is your power animal a penguin?

Attachments

File name MIME type
signature.asc application/pgp-signature

Replies

Subject Author
Re: [gentoo-dev] Policy for retirement of old gentoo 'versions' Dylan Carlson <absinthe@g.o>