1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
|
Summary of the Gentoo council meeting 12 June 2016
agenda: https://archives.gentoo.org/gentoo-project/message/50dbe189dd2641d5730f08944e7fa7ce
1. Roll call
=====================================
Present: blueness, dilfridge, ulm, Soap (proxying for jlec), K_F,
WilliamH, rich0
2. Discussion on mgorny's items
=====================================
https://archives.gentoo.org/gentoo-dev/message/68a870c0519fb1cb7152db38fc9d4935
Michał Gorny proposed several unrelated items in a lengthy e-mail:
a) a "file installation mask" GLEP,
b) for portage to actually adhere to USE_EXPAND semantics,
c) to abolish the sys-devel/gcc multislot use flag entirely,
d) the migration from LINGUAS USE_EXPAND to L10N
e) discussion of the gui use flag.
He later withdrew all these proposals. Items c) and d) were discussed
anyway.
Regarding c), bug 584610, blueness as toolchain member explained that
now the new implementation of the multislot use flag in sys-devel/gcc
controls a blocker. If unset, previous gcc versions are blocked and
consequently unmerged on update (leading to potential ABI breakage in
C++). If set, minor version bumps accumulate on update.
This led to a discussion why 1) the blocker exists, why 2) the slot
was refined from major version bumps (4.8, 4.9) to point releases
(4.8.1, 4.8.2, ...), and why this behaviour was not reserved to an
overlay for toolchain specialists. For none of these questions valid
technical arguments were known.
After discussion blueness proposed to e-mail vapier, request
clarifications on the changes and ask him if he wants to force
USE=multislot universally, since this seems to be the easier choice.
Regarding d), see
https://archives.gentoo.org/gentoo-dev/message/a08ea09c2c8e534fd9bc1146703c66ff
https://archives.gentoo.org/gentoo-dev/message/41e09d1ddc8b30abb9f9d21d205b7b82
ulm explained that LINGUAS should become a normal environment variable
(since it has a special meaning to gettext) and the current USE_EXPAND
functionality should be transferred to a new variable L10N. No council
decision is required; the new variable will be introduced in a news
item.
3. Open bugs with council involvement
=====================================
Dilfridge was reminded to complete the missing summaries; blueness
announced that he just finished two summaries.
No action was taken regarding games.eclass and the games team.
Bug 565566 "New ChangeLogs are in chronological order" was mentioned
but no action was taken.
Subsequent non-public meeting
=====================================
A non-public discussion of bug 592598 followed.
|
GitWeb
