Gentoo Archives: gentoo-dev

From: tigger@g.o
To: Dan Armak <danarmak@g.o>
Cc: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] RFC: GLEP 19 -- Gentoo Stable Portage Tree
Date: Tue, 03 Feb 2004 12:59:26
Message-Id: 20040203124357.GA4152@gentoo.org
In Reply to: Re: [gentoo-dev] RFC: GLEP 19 -- Gentoo Stable Portage Tree by Dan Armak
1 On Tue, Feb 03, 2004 at 02:38:07PM +0200, Dan Armak wrote:
2 > On Tuesday 03 February 2004 12:39, Paul de Vrieze wrote:
3 > > The main issue that the stable keyword is trying to solve is the
4 > > following:
5 > >
6 > > How can we have package maintainers indicate in an easy way those
7 > > packages that are candidates for inclusion in the frozen tree. Those
8 > > ebuilds can then be collected into the stable branch automatically.
9 > >
10 > > One solution is to ask developers to set the stable flag on an ebuild.
11 > > The newest package with this keyword would than be included into the
12 > > tree.
13 > >
14 > > There are however problems with the keywords solution. The big problem is
15 > > that it is not easy to know which auxiliary files (patches) are needed
16 > > for a particular ebuild. I think a better approach would be a staging
17 > > area of some kind where the relevant ebuilds could be entered into
18 > > (possibly in cat/package-version directories). The inclusion script
19 > > could then test building the package and then include it into the
20 > > to-be-frozen tree.
21 > >
22 > > After the frozen tree is released the staging area could be cleaned up
23 > > (or archived for later reference).
24 >
25 > I understand why a keyword might be useful in marking ebuilds that'll go into
26 > a stable branch. However, you haven't ansewred my question - why have a
27 > separate cvs branch at all? Why not just use keywords as we do now for
28 > arch/~arch?
29
30 My own opinion:
31
32 1. The size of the tree. If this is to be an enterprise tree, its likely
33 that it won't have an ebuild for every package as most enterprise people
34 won't want x (name an obscure desktop related package). Quicker
35 synchrosing and easier backup/restore and less clutter is good. This
36 isn't a big issue really though.
37
38 2. Guaranteed someone will "clean" an ebuild thinking its to old and
39 needs to be removed. This is a big issue.
40
41 If its a seperate tree where only people who know the rules of the tree
42 work, its more likely to stay consistant and not have stuff removed that
43 shouldn't be.
44
45 --
46 rob holland - [ tigger@g.o ]
47 irc://irc.freenode.net/#gentoo as tigger^
48 http://dev.gentoo.org/~tigger/tigger@××××××××××.asc

Replies

Subject Author
Re: [gentoo-dev] RFC: GLEP 19 -- Gentoo Stable Portage Tree Dan Armak <danarmak@g.o>
Re: [gentoo-dev] RFC: GLEP 19 -- Gentoo Stable Portage Tree Brett Simpson <simpsonb@××××××××××××××××××.org>
Re: [gentoo-dev] RFC: GLEP 19 -- Gentoo Stable Portage Tree Terje Kvernes <terjekv@××××××××.no>