1 |
It'd probably need more discussion but there's two things I'd like |
2 |
to see if EAPI-8 bump this: |
3 |
|
4 |
1. remove src_prepare export in EAPI-8 |
5 |
While "some" packages need xdg_environment_reset, most don't because |
6 |
the eclass is often only inherited to handle icons/.desktop and this |
7 |
just needlessly overwrite the src_prepare of other eclasses requiring |
8 |
more careful inherit ordering (e.g. inherit xdg cmake). |
9 |
|
10 |
I'd prefer it was clear when a package need this by calling |
11 |
xdg_environment_reset directly. Unless there is a non-trivial amount |
12 |
of packages that need it (e.g. for tests) that I'm not aware of. |
13 |
|
14 |
2. no *DEPEND in EAPI-8 rather than IDEPEND |
15 |
xdg-utils is designed to not fail if these tools are missing and |
16 |
cache updates are done "all at once" catching anything missing. |
17 |
Then this cache shouldn't be needed until these tools are emerged, |
18 |
preferably as a dep of a library using this cache + do 1st generation. |
19 |
|
20 |
Many ebuilds already lack these deps by using xdg-utils directly, |
21 |
so impact should be minimal. |
22 |
|
23 |
Lack of *DEPEND is also useful for packages with optional gui support |
24 |
that don't always install icons/.desktop, without falling back to |
25 |
xdg-utils. |
26 |
|
27 |
Old DEPEND should be kept as-is not to risk breaking packages with odd |
28 |
checks that need it present at build time. I'd rather no RDEPEND switch |
29 |
either as it'll just complicate things with revbumps and tools don't |
30 |
really need to be in RDEPEND given the above. |
31 |
|
32 |
-- |
33 |
ionen |