1 |
Jim Ramsay wrote: |
2 |
>> > This meets the following goals: |
3 |
>> > 1) It makes it easy for "regular" users to get netscape-flash with |
4 |
>> > any additions required by any global USE flags in exactly one step: |
5 |
>> > - emerge netscape-flash |
6 |
>> So, in netscape-flash: |
7 |
>> RDEPEND=" |
8 |
>> ssl? ( foo/libflashsupport ) |
9 |
>> pulseaudio? ( foo/libflashsupport ) |
10 |
>> esd? ( foo/libflashsupport ) |
11 |
>> oss? ( foo/libflashsupport ) |
12 |
>> " |
13 |
>> and IUSE="ssl pulseaudio esd oss gnutls" in libflashsupport (which, as |
14 |
>> already said, has it's own ebuild)? |
15 |
> |
16 |
> Yes, I considered this, it is option (2) in the original post in this |
17 |
> thread. However, I do not believe this is the best solution. Consider |
18 |
> the case where: |
19 |
> - A user has 'ssl' disabled globally |
20 |
> - A user sees that netscape-flash now has 'ssl' support, so he/she |
21 |
> enables 'ssl' just for the netscape-flash ebuild. |
22 |
> - The ebuild would then install libflashsupport, but do so without |
23 |
> actually adding ssl support, which would be quite frustrating to the |
24 |
> user, and probably generate unnedded bug traffic. |
25 |
> |
26 |
Well then the ebuild for libflashsupport should pull in ssl iff the flag is |
27 |
set; other packages do similar stuff. A usr who adds ssl to package.use |
28 |
will not be surprised to see the ssl package being pulled in, so no bad |
29 |
experience for the usr, and hopefully less bug reports. |
30 |
|
31 |
I really *don't* like the option of a crippled install. It's |
32 |
counter-intuitive and is asking for trouble imo. |
33 |
|
34 |
Thanks for adding shiny stuff to flash :) |
35 |
|
36 |
|
37 |
-- |
38 |
gentoo-dev@g.o mailing list |