1 |
On Sun, 24 Oct 2004, Ciaran McCreesh wrote: |
2 |
> On Sun, 24 Oct 2004 18:31:19 +0200 Maurice van der Pot |
3 |
> <griffon26@g.o> wrote: |
4 |
>| Whether a use flag is specifically turned off for a group or it is |
5 |
>| just not present in the group should make no difference for the |
6 |
>| meaning of GROUP. The same holds for -@GROUP. In other words, -use and |
7 |
>| -@GROUP should only be used in the definition of a group to filter |
8 |
>| flags out. |
9 |
>| |
10 |
>| Assume MYGNOME="@GNOME -gtk", then USE="... -@MYGNOME ..." should |
11 |
>| *not* be equivalent to USE="... -@GNOME gtk ...", instead it should |
12 |
>| not influence the state of gtk at all. |
13 |
> |
14 |
> Interesting idea. This then brings us back to one of the original |
15 |
> objections, which was that there should be some way to turn off flags |
16 |
> from within a group. Only way to do that would be to introduce yet |
17 |
> another modifier, which is inconsistent with the rest of portage and |
18 |
> will confuse the heck out of anyone who hasn't studied set theory... |
19 |
|
20 |
How about the following: |
21 |
|
22 |
We add a + operator, to specify settings which are fundamentally part of |
23 |
the group. Negating the group strips all negations, strips all simply |
24 |
present flags, and negates those with a +. |
25 |
|
26 |
So, we then have |
27 |
|
28 |
KDE=+kde +qt X |
29 |
GNOME=+gtk +gtk2 +gnome X |
30 |
|
31 |
USE="@KDE -@GNOME" |
32 |
|
33 |
results in |
34 |
|
35 |
USE="kde qt X -gtk -gtk2 -gnome" |
36 |
|
37 |
This could make some groups be only useful in the positive, for example: |
38 |
|
39 |
MINIMALIST=-kde -qt -X -gtk -gtk2 -gnome -this -that -theotherthing |
40 |
SERVER=@MINIMALIST hardened crypt ncurses pam readline snmp ssl tcpd |
41 |
|
42 |
Ed |
43 |
|
44 |
-- |
45 |
gentoo-dev@g.o mailing list |