1 |
Simon Stelling wrote: |
2 |
> Daniel Ostrow wrote: |
3 |
> |
4 |
>> Hrm, /me thinks you are missing something there, almost the entire |
5 |
>> tree doesn't explicitly state the mirror://gentoo SRC_URI, portage |
6 |
>> handles that automatically. That being the case portage would have |
7 |
>> change so that the automatic lookup was mirror://gentoo/${firstchar}/. |
8 |
>> So that is at least one portage change I can think of being required.... |
9 |
|
10 |
1925 ebuilds ( with a hacked up SRC_URI checking script )[1] |
11 |
URI_check.py "mirror://gentoo" |
12 |
|
13 |
> |
14 |
> |
15 |
> Huh? What does it state then? AFAIK ebuilds should ALWAYS use the |
16 |
> mirror:// URI when possible, and since this change is only affecting our |
17 |
> own mirrors, it is always possible. |
18 |
> |
19 |
>> Sure I can still see your point about needing to manually change the |
20 |
>> packages that do explicitly state mirror://gentoo in their SRC_URI, |
21 |
>> but given that you would have to do the above anyway.... |
22 |
> |
23 |
> |
24 |
> Huh?? My point was that we shouldn't have to change all those ebuilds |
25 |
> but instead just changing the mirror://gentoo-mapping. |
26 |
> |
27 |
|
28 |
See if we do it the ebuild way we can filter via EAPI. The ebuild has a |
29 |
EAPI=2 SRC_URI, but portage is only EAPI=0, then the ebuild is |
30 |
automagically filtered; as opposed to the ebuild failing miserably. |
31 |
It's getting close to the point where we can finally leverage EAPI to |
32 |
push features out faster because backwards compatability is maintained ( |
33 |
for portage ). Infra is still screwed essentially doing 2 |
34 |
implementations until such time as the old one can die. |
35 |
|
36 |
I'd prefer the mirrors not be special cased in a mapping since. URI's |
37 |
are URI's are URI's... |
38 |
|
39 |
-Alec Warner |
40 |
|
41 |
|
42 |
----------------------------------------- |
43 |
[1] dev.gentoo.org/~antarus/URI_check.py |
44 |
|
45 |
-- |
46 |
gentoo-dev@g.o mailing list |