1 |
On Sun, 6 Mar 2005 16:28:49 +0000 |
2 |
Ciaran McCreesh <ciaranm@g.o> wrote: |
3 |
|
4 |
> http://www.gentoo.org/proj/en/glep/glep-0029.html |
5 |
> |
6 |
> We need to decide upon a solution to the -flags problem. The proposed |
7 |
> solutions are: |
8 |
> |
9 |
> 1) Disallow all negatives. Disadvantages: we don't tend to disallow |
10 |
> things just because users could shoot themselves in the foot with |
11 |
> them. |
12 |
> |
13 |
> 2) Allow negatives, and document how to use them correctly. |
14 |
> Disadvantages: sooner or later, some developer isn't going to read the |
15 |
> docs, and will really really screw things up with a misunderstanding. |
16 |
> |
17 |
> 3) Reorder USE flags. Disadvantages: can't disable flags set by a |
18 |
> group. Breaks existing USE lines. |
19 |
> |
20 |
> 4) Have portage warn on dodgy USE flag lines. Disadvantages: tricky, |
21 |
> hard for the end user to figure out. |
22 |
> |
23 |
> 5) Use some weird tristate notation. Disadvantages: not everyone has |
24 |
> the slightest clue what set theory is. |
25 |
> |
26 |
> I'm in favour of 2) personally, but others disagree. I'd like a proper |
27 |
> discussion on this before trying to get the GLEP through. |
28 |
|
29 |
Hmm, are talking about what should be technically possible or what |
30 |
should be supported? Personally I'd like to have negatives available |
31 |
but label them as unsupported ("if you play with this it's all your |
32 |
fault"), that would also mean that predefined groups don't use them. |
33 |
|
34 |
Marius |
35 |
|
36 |
-- |
37 |
Public Key at http://www.genone.de/info/gpg-key.pub |
38 |
|
39 |
In the beginning, there was nothing. And God said, 'Let there be |
40 |
Light.' And there was still nothing, but you could see a bit better. |