Gentoo Archives: gentoo-dev

From: Peter Stuge <peter@×××××.se>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] Re: Banning modification of pkg-config files
Date: Mon, 12 May 2014 19:26:04
Message-Id: 20140512192555.3345.qmail@stuge.se
In Reply to: Re: [gentoo-dev] Re: Banning modification of pkg-config files by Tom Wijsman
1 Tom Wijsman wrote:
2 > besides a temporary fix downstream it should go upstream;
3
4 I think there is agreement that this is the ideal, and that the
5 discussion is about what to do when that seems out of reach.
6
7
8 > > My key point is that it isn't Gentoo's responsibility or duty to fix
9 > > problems introduced by upstreams, even if Gentoo developers are so
10 > > skilled that they would be able to.
11 > >
12 > > I think your time is better spent on things that are not broken.
13 >
14 > If we all did that, I wonder how much would still work; not as much as
15 > we have achieved now, so I like that we've made an "added value".
16
17 Don't get me wrong - I'm wholeheartedly in favor of contributing
18 fixes upstream, but not of building a support workload when upstream
19 is uncooperative.
20
21
22 > > Maybe introduce a USE flag for installing .pc:s in ${FILESDIR} ?
23 >
24 > We have recently decided to not use an USE flag for small files; so,
25 > I'm not sure if this proposal is much different from that decision.
26
27 Oh ok. Sure.
28
29
30 > You can use INSTALL_MASK for this purpose I think.
31
32 Mh but that'll also kill .pc files from upstream. The option I'd like
33 is to mask only the ones added by Gentoo. I would be fine with them
34 having gentoo- names. That makes it very clear that this interface is
35 a Gentoo-specific thing. However, that throws off the autoconf
36 pkg-config macros. :\
37
38
39 > > > (Are we seriously discussing banning something useful as pkg-config
40 > > > files?! That's retarded. Must be some joke.)
41 > >
42 > > I don't think I said to ban them. I said that I want Gentoo to stay
43 > > close to upstream by default. I also said that maintainers shouldn't
44 > > be expected to untie upstream bugknots.
45 > >
46 > > Please do not call me retarded again.
47 >
48 > That might have been meant to be about the thread as a whole.
49
50 All right, fair enough there too.
51
52
53 Thanks
54
55 //Peter

Replies