1 |
Ühel kenal päeval, P, 29.01.2017 kell 16:06, kirjutas Andreas K. |
2 |
Huettel: |
3 |
> However one valid point that came up in discussions is - whether an |
4 |
> arch |
5 |
> supports stable keywords is a per-arch setting, not a per-profile |
6 |
> setting. So |
7 |
> we can actually make things much easier (and the transition safer). |
8 |
> |
9 |
> Proposal No 2: |
10 |
> * Leave profiles.desc unmodified |
11 |
> * Introduce a new file arch.desc, which contains the "stability |
12 |
> status" of an |
13 |
> arch; |
14 |
> |
15 |
> Syntax: 2 columns, |
16 |
> # arch status |
17 |
> amd64 stable |
18 |
> mips testing |
19 |
> sh unstable |
20 |
> |
21 |
> The meaning of the keywords "stable", "testing", "unstable" is the |
22 |
> same as in |
23 |
> the previous proposal, |
24 |
|
25 |
Maybe declare from the start that any extra columns should be silently |
26 |
ignored in implementations from start, as to be able to safely add more |
27 |
columns in the future without breaking backwards compatibility. |
28 |
|
29 |
> > "Does this arch support stable keywords, and how should "arch" vs. |
30 |
> > "~arch" |
31 |
> > be treated?" |
32 |
> > - "stable": separately check consistency of ~arch and arch tree, |
33 |
> > both have |
34 |
> > to be OK. This is what repoman is doing now, and is the default if |
35 |
> > the 4th |
36 |
> > column is undefined. |
37 |
> > - "testing": treat "arch" as "~arch" when requiring consistency, do |
38 |
> > not |
39 |
> > check "arch" alone. Useful if an arch wants to prepare going |
40 |
> > stable, useful |
41 |
> > for arch teams maintaining a pseudo-stable subset for stages. |
42 |
> > repoman could |
43 |
> > have a new command line switch that temporarily upgrades from |
44 |
> > "testing" to |
45 |
> > "stable" (for arch team work). |
46 |
> > - "unstable": check "~arch" only, "arch" in an ebuild produces a |
47 |
> > fatal |
48 |
> > repoman error |
49 |
> |
50 |
> The combination of current profiles.desc and new arch.desc provides |
51 |
> the same |
52 |
> flexibility as in the previous e-mail. |
53 |
> |
54 |
> Compatibility and transition: |
55 |
> |
56 |
> 0) PMS should be amended to allow the additional file. |
57 |
> |
58 |
> 1) Compatibility: No arch.desc and new system, or arch not listed in |
59 |
> arch.desc |
60 |
> *Arches* are treated as "stable" by repoman (current behaviour), with |
61 |
> profile |
62 |
> status according to profiles.desc. |
63 |
> Gentoolkit and other tools trying to determine a list of stable |
64 |
> arches should |
65 |
> fall back to current method of scanning profiles.desc for stable |
66 |
> profiles. |
67 |
> |
68 |
> 2) Compatibility: arch.desc and old system |
69 |
> Tools ignore the unknown file (?). |
70 |
> Repoman and other tools may emit surplus errors when profiles are |
71 |
> checked on |
72 |
> arches that are "testing" (they check the consistency of the stable |
73 |
> tree |
74 |
> alone, which is not OK, since "arch" is supposed to be treated like |
75 |
> "~arch"). |
76 |
> |
77 |
> 3) On introduction of the new column, it will be set to "stable" for |
78 |
> all |
79 |
> stable arches, "testing" for all arches where "inofficial" stable |
80 |
> keywords |
81 |
> exist (sh, s390, ...), and "unstable" everywhere else. |
82 |
|
83 |
Might want a "broken" (with maybe a better name) for some of these. I |
84 |
bet the ~arch of some of these is broken too, and no-one to respond to |
85 |
keyword requests, just happens when it happens. |
86 |
arm64 and mips are in that state too until we get that fixed and could |
87 |
move to "testing" and then later "stable" in case of arm64. |
88 |
|
89 |
> 4) Arches in "testing" or "unstable" may eventually consider re- |
90 |
> introducing |
91 |
> stable *profiles* so their deptree in ~arch remains consistent. |