Contents of /xml/htdocs/proj/en/glep/glep-0040.txt

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1.1 - (hide annotations) (download)
Fri Sep 9 03:06:28 2005 UTC (13 years, 4 months ago) by g2boojum
Branch: MAIN
File MIME type: text/plain
new glep

1 g2boojum 1.1 GLEP: 40
2     Title: Standardizing "arch" keywording across all archs.
3     Version: $Revision: 1.8 $
4     Last-Modified: $Date: 2005/01/09 16:12:40 $
5     Author: Grant Goodyear <g2boojum@gentoo.org>
6     Status: Draft
7     Type: Standards Track
8     Content-Type: text/x-rst
9     Created: 3-Sep-2005
10     Post-History: 6-Sep-2005
12     Credits
13     =======
15     This GLEP originated from a rather contentious discussion_ on gentoo-dev
16     about combining the x86 and amd64 keywords. This GLEP attempts to get at the
17     heart of that discontent. The proposed stable-keyword guidelines have been
18     lifted verbatim from `The Doc`_.
20     .. _discussion: http://tinyurl.com/bp859
21     .. _The Doc: http://dev.gentoo.org/~plasmaroo/devmanual
23     Abstract
24     ========
26     It is time for x86 to no longer be an exception to the standard
27     keywording guidelines. Thus, an x86 arch team should be responsible
28     for moving packages from ~x86 to x86.
30     Motivation
31     ==========
33     The original, informal x86 keywording policy, where almost any x86 dev (which
34     were the vast majority of devs) who used a package could mark it stable, arose
35     from a time when there were relatively few Gentoo devs. Adding packages to
36     the tree was the principal concern, as opposed to maintaining existing
37     packages. QA considerations have since modified that policy slightly, and now
38     it is the package maintainers who should mark an x86 package stable. Of
39     course, that policy presumes that package maintainers are generally x86 devs,
40     which is slowly becoming less and less true.
42     This policy for x86 is quite different from how every other arch marks
43     packages stable. For the non-x86 archs, each arch has a specific "arch team"
44     which is responsible for moving packages from ``~arch`` to ``arch``, although
45     vapier notes that "arch teams generally defer to maintainers (and rightly so)
46     as to *when* newer versions should go stable." This approach has worked quite
47     well for the non-x86 archs, and this GLEP asserts that the same approach would
48     benefit x86 as well.
50     Specification
51     =============
53     Stabling guidelines for all archs
54     ---------------------------------
56     For a package to move to stable, the following guidelines must be met:
58     * The package has spent a reasonable amount of time in ``~arch`` first.
59     Thirty days is the usual figure, although this is clearly only a guideline.
60     For critical packages, a much longer duration is expected. For small
61     packages which have only minor changes between versions, a shorter period
62     is sometimes appropriate.
63     * The package must not have any non-``arch`` dependencies.
64     * The package must not have any severe outstanding bugs or issues.
65     * The package must be widely tested.
66     * If the package is a library, it should be known not to break any package
67     which depends upon it.
68     * The relevant ``arch`` team must agree to it.
70     x86 arch team
71     -------------
73     A robust x86 arch team needs to be created. The x86@gentoo.org alias already
74     exists, and it merely needs to be used. This team, with the aid of potential
75     non-dev ``arch testers``, has the responsibility of stabling all x86 packages.
76     Current x86 devs who wish to mark their own packages stable must therefore
77     either be members of or make individual arrangements with the x86 arch team.
80     Rationale
81     =========
83     There will be a considerable one-time cost involved in establishing a robust
84     x86 arch team--a good number of bodies (the amd64 atch team has 19 active devs
85     and 12 active non-dev arch testers) need to be recruited to be part of the
86     new arch team, and convincing devs that it is in their best interests to work
87     in a new fashion is likely to be even harder. Certainly the benefit of
88     consistency between the various archs is obvious, but is it worth the cost
89     involved? Here are the arguments for enduring the pain involved:
91     * Over time, x86 is likely to become a minority arch as 64-bit systems
92     become the norm. The implicit assumptions that underly the current
93     system (that most devs, users, and package maintainers use x86)
94     will become increasingly less valid.
95     * Markedly improved QA for x86. Assuming that the author's own is
96     behavior is representative, most x86 devs run ``~x86`` systems.
97     Thus, the assumption that devs are good ``x86`` testers is not really
98     valid. One obvious consequence is that packages tend to languish in
99     ``~x86`` for a very long time, since x86 devs running ``~x86`` have little
100     cause to notice that a package has not been marked stable. The much larger
101     effect, though, is that it is rare for ``x86`` packages to be stabled in
102     the context of a full ``x86`` tree, so the big picture of a stable
103     *system*, not just a stable package, is lost. This approach of stabling
104     in the context of a full stable ``arch`` tree, it has been argued_, is
105     the fundamental reason why the non-x86 archs have notably better QA
106     than does the x86 arch.
108     .. _argued: http://thread.gmane.org/gmane.linux.gentoo.devel/30369
110     Implementation
111     ==============
113     Creation of a robust x86 team is already underway. The more vital step
114     is the official change in policy, along with a sustained effort to get
115     existing x86 devs to go along with it.
117     Alternative Ideas
118     =================
120     Stuart_ has suggested the creation of a new arch keyword: "[-]maint", which
121     would exist in tandem with the normal arch keywords, thereby making the
122     package maintainer's intention explicit. Ciaranm has responded that by
123     definition a package in ``~arch`` is a candidate for ``arch``, so a package's
124     mere presence in the tree (without being in ``package.mask``) should indicate
125     the package maintainer's intention. There was a fair bit of discussion about
126     whether the idea should be a "maint" keyword, or named something else, or an
127     entirely different variable, etcetera, but the basic gist didn't change much.
129     Jstubbs notes that it could be a very good idea if all non-arch devs worked in
130     overlays, but that new portage (gensync) support would be needed to make it
131     truly viable. Stuart pointed out that php5 support was handled just that way.
132     One author's view is that this approach would make the "package in ``~arch``
133     means that it's a de-facto candidate for ``arch``" interpretation even more
134     valid.
136     Ciaranm and weeve have noted that it is occasionally necessary for arch teams
137     to override a package maintainer when it comes to stabling a package. Stuart
138     has asserted that in those cases the arch team should be willing to take on
139     the support burden for that package.
141     .. _Stuart: http://thread.gmane.org/gmane.linux.gentoo.devel/31060
143     Backwards Compatibility
144     =======================
146     Not really an issue here.
149     Copyright
150     =========
152     This document has been placed in the public domain.

  ViewVC Help
Powered by ViewVC 1.1.20