1 |
Torsten Veller wrote: |
2 |
> |
3 |
> | 15:35 < dev-zero> prepalldocs should be kept internal and usage should be avoided |
4 |
> | 15:36 < dev-zero> reason: internal function and change of it's implementation prooves it |
5 |
> | 15:36 < dev-zero> if someone want's it's functionality he should propose a solution for a future eapi |
6 |
> |
7 |
> It is not an internal function. And it's a strange way to prove anything. |
8 |
> Around 250 ebuilds use it. So there is probably someone who actually uses this functionality. |
9 |
> |
10 |
|
11 |
And what functionality is that exactly? It's not like we would have |
12 |
mandated the removal of prepalldocs from Portage. How do you propose |
13 |
other PMs would implement prepalldocs when you have no idea which |
14 |
behavior the ebuild authors were expecting at the time they wrote their |
15 |
ebuilds? |
16 |
|
17 |
More discussion happened in the meeting before the one yesterday I think. |
18 |
|
19 |
> |
20 |
>> Maintainers should use the public APIs available to them. |
21 |
> |
22 |
> It's in ebuild.5 for years. |
23 |
> And as recruiting lead you probably know that the Developer Handbook contains it as well |
24 |
> <http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml?part=2&chap=1>. |
25 |
> |
26 |
|
27 |
If someone would have put the work into it that doc would have died ages |
28 |
ago. Any way I changed the text there to mention that new ebuilds should |
29 |
not use it. |
30 |
|
31 |
> |
32 |
> I will probably do it after i got answers to the questions above. |
33 |
> |
34 |
> Thanks. |
35 |
> |
36 |
|
37 |
I would much rather see the energy put into creating a well specified |
38 |
replacement for prepalldocs than trying to get prepalldocs to behave. |
39 |
|
40 |
Regards, |
41 |
Petteri |