/[gentoo]/xml/htdocs/proj/en/glep/glep-0040.txt
Gentoo

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.3 - (show annotations) (download)
Sun Nov 13 17:16:50 2005 UTC (8 years, 10 months ago) by g2boojum
Branch: MAIN
Changes since 1.2: +3 -3 lines
File MIME type: text/plain
Mark GLEP 43 approved and final

1 GLEP: 40
2 Title: Standardizing "arch" keywording across all archs.
3 Version: $Revision: 1.2 $
4 Last-Modified: $Date: 2005/09/15 21:02:11 $
5 Author: Grant Goodyear <g2boojum@gentoo.org>
6 Status: Approved
7 Type: Standards Track
8 Content-Type: text/x-rst
9 Created: 3-Sep-2005
10 Post-History: 6-Sep-2005 15-Sep-2005
11
12 Status
13 ======
14
15 Approved by the Gentoo Council on 15 September 2005.
16
17 Credits
18 =======
19
20 This GLEP originated from a rather contentious discussion_ on gentoo-dev
21 about combining the x86 and amd64 keywords. This GLEP attempts to get at the
22 heart of that discontent. The proposed stable-keyword guidelines have been
23 lifted verbatim from `The Doc`_.
24
25 .. _discussion: http://tinyurl.com/bp859
26 .. _The Doc: http://dev.gentoo.org/~plasmaroo/devmanual
27
28 Abstract
29 ========
30
31 It is time for x86 to no longer be an exception to the standard
32 keywording guidelines. Thus, an x86 arch team should be responsible
33 for moving packages from ~x86 to x86.
34
35 Motivation
36 ==========
37
38 The original, informal x86 keywording policy, where almost any x86 dev (which
39 were the vast majority of devs) who used a package could mark it stable, arose
40 from a time when there were relatively few Gentoo devs. Adding packages to
41 the tree was the principal concern, as opposed to maintaining existing
42 packages. QA considerations have since modified that policy slightly, and now
43 it is the package maintainers who should mark an x86 package stable. Of
44 course, that policy presumes that package maintainers are generally x86 devs,
45 which is slowly becoming less and less true.
46
47 This policy for x86 is quite different from how every other arch marks
48 packages stable. For the non-x86 archs, each arch has a specific "arch team"
49 which is responsible for moving packages from ``~arch`` to ``arch``, although
50 vapier notes that "arch teams generally defer to maintainers (and rightly so)
51 as to *when* newer versions should go stable." This approach has worked quite
52 well for the non-x86 archs, and this GLEP asserts that the same approach would
53 benefit x86 as well.
54
55 Specification
56 =============
57
58 Stabling guidelines for all archs
59 ---------------------------------
60
61 For a package to move to stable, the following guidelines must be met:
62
63 * The package has spent a reasonable amount of time in ``~arch`` first.
64 Thirty days is the usual figure, although this is clearly only a guideline.
65 For critical packages, a much longer duration is expected. For small
66 packages which have only minor changes between versions, a shorter period
67 is sometimes appropriate.
68 * The package must not have any non-``arch`` dependencies.
69 * The package must not have any severe outstanding bugs or issues.
70 * The package must be widely tested.
71 * If the package is a library, it should be known not to break any package
72 which depends upon it.
73 * The relevant ``arch`` team must agree to it.
74
75 x86 arch team
76 -------------
77
78 A robust x86 arch team needs to be created. The x86@gentoo.org alias already
79 exists, and it merely needs to be used. This team, with the aid of potential
80 non-dev ``arch testers``, has the responsibility of stabling all x86 packages.
81 Current x86 devs who wish to mark their own packages stable must therefore
82 either be members of or make individual arrangements with the x86 arch team.
83
84
85 Rationale
86 =========
87
88 There will be a considerable one-time cost involved in establishing a robust
89 x86 arch team--a good number of bodies (the amd64 arch team has 19 active devs
90 and 12 active non-dev arch testers) need to be recruited to be part of the
91 new arch team, and convincing devs that it is in their best interests to work
92 in a new fashion is likely to be even harder. Certainly the benefit of
93 consistency between the various archs is obvious, but is it worth the cost
94 involved? Here are the arguments for enduring the pain involved:
95
96 * Over time, x86 is likely to become a minority arch as 64-bit systems
97 become the norm. The implicit assumptions that underly the current
98 system (that most devs, users, and package maintainers use x86)
99 will become increasingly less valid.
100 * Markedly improved QA for x86. Assuming that the author's own is
101 behavior is representative, most x86 devs run ``~x86`` systems.
102 Thus, the assumption that devs are good ``x86`` testers is not really
103 valid. One obvious consequence is that packages tend to languish in
104 ``~x86`` for a very long time, since x86 devs running ``~x86`` have little
105 cause to notice that a package has not been marked stable. The much larger
106 effect, though, is that it is rare for ``x86`` packages to be stabled in
107 the context of a full ``x86`` tree, so the big picture of a stable
108 *system*, not just a stable package, is lost. This approach of stabling
109 in the context of a full stable ``arch`` tree, it has been argued_, is
110 the fundamental reason why the non-x86 archs have notably better QA
111 than does the x86 arch.
112
113 .. _argued: http://thread.gmane.org/gmane.linux.gentoo.devel/30369
114
115 Implementation
116 ==============
117
118 Creation of a robust x86 team is already underway. The more vital step
119 is the official change in policy, along with a sustained effort to get
120 existing x86 devs to go along with it.
121
122 Alternative Ideas
123 =================
124
125 Stuart_ has suggested the creation of a new arch keyword: "[-]maint", which
126 would exist in tandem with the normal arch keywords, thereby making the
127 package maintainer's intention explicit. Ciaranm has responded that by
128 definition a package in ``~arch`` is a candidate for ``arch``, so a package's
129 mere presence in the tree (without being in ``package.mask``) should indicate
130 the package maintainer's intention. There was a fair bit of discussion about
131 whether the idea should be a "maint" keyword, or named something else, or an
132 entirely different variable, etcetera, but the basic gist didn't change much.
133
134 Jstubbs notes that it could be a very good idea if all non-arch devs worked in
135 overlays, but that new portage (gensync) support would be needed to make it
136 truly viable. Stuart pointed out that php5 support was handled just that way.
137 One author's view is that this approach would make the "package in ``~arch``
138 means that it's a de-facto candidate for ``arch``" interpretation even more
139 valid.
140
141 Ciaranm and weeve have noted that it is occasionally necessary for arch teams
142 to override a package maintainer when it comes to stabling a package. Stuart
143 has asserted that in those cases the arch team should be willing to take on
144 the support burden for that package.
145
146 .. _Stuart: http://thread.gmane.org/gmane.linux.gentoo.devel/31060
147
148 Backwards Compatibility
149 =======================
150
151 Not really an issue here.
152
153
154 Copyright
155 =========
156
157 This document has been placed in the public domain.

  ViewVC Help
Powered by ViewVC 1.1.20