1 |
On Sun, 29 Jan 2017 18:20:34 +0200 |
2 |
Mart Raudsepp <leio@g.o> wrote: |
3 |
|
4 |
> Might want a "broken" (with maybe a better name) for some of these. I |
5 |
> bet the ~arch of some of these is broken too, and no-one to respond to |
6 |
> keyword requests, just happens when it happens. |
7 |
> arm64 and mips are in that state too until we get that fixed and could |
8 |
> move to "testing" and then later "stable" in case of arm64. |
9 |
|
10 |
This is one aspect I liked about my other proposal, that the behaviours |
11 |
associated with various keywords was well defined. |
12 |
|
13 |
--- |
14 |
# keywords: |
15 |
# - strict: arch-foo and ~arch-foo treated distinctly |
16 |
# - mixed: arch-foo treated as ~arch-foo |
17 |
# - any: packages that exist and have any keywords are ~arch-foo |
18 |
# |
19 |
# dependencies: |
20 |
# - enforce: referential integrity within logical keywords |
21 |
# is required |
22 |
# - warn: referential integrity within logical keywords |
23 |
# warns when its bad, and enforced with -d |
24 |
# - ignore: referential integrity is not even considered |
25 |
# without -e y and enforced with -e y |
26 |
|
27 |
# name | keywords | dependencies |
28 |
stable strict check |
29 |
|
30 |
# was strict ignore |
31 |
dev mixed warn |
32 |
exp mixed ignore |
33 |
---- |
34 |
|
35 |
In that, its instructive as to the significance of the terms. |
36 |
|
37 |
"Unstable" and "Broken" don't really say much to me. |
38 |
|
39 |
But dependencies=ignore much more adequately communicates to me |
40 |
the state that, dependency coherence is known to be problematic |
41 |
and that you should not care about dependency coherence unless |
42 |
you have some specific agenda. |
43 |
|
44 |
|
45 |
If we can mix and match some of these designs with the arch.desc file |
46 |
however, I'm all for it. |