Gentoo Archives: gentoo-dev

From: Duncan <1i5t5.duncan@×××.net>
To: gentoo-dev@l.g.o
Subject: [gentoo-dev] Re: splitting build deps out from depends
Date: Sat, 02 Jul 2005 03:51:03
Message-Id: pan.2005.07.02.03.48.14.687374@cox.net
In Reply to: Re: [gentoo-dev] splitting build deps out from depends by Maurice van der Pot
1 Maurice van der Pot posted <20050701183536.GR19598@××××××××.com>,
2 excerpted below, on Fri, 01 Jul 2005 20:35:36 +0200:
3
4 > If the point is to make dependencies complete, isn't there a way to
5 > build in some support for detecting it into some tool or other?
6 >
7 > If we have a program that can create an environment and detect which
8 > programs are run within the environment (maybe sandbox can do this,
9 > maybe something with LD_PRELOAD, I'm sure we can think of something),
10 > then we can build a list of programs that are run during the build.
11 > If we have such a list, we can find out which packages are required
12 > to provide the tools in the list.
13 >
14 > Such a tool could be used to generate the correct build dependencies
15 > automatically or verify the existing bdeps.
16
17 If the other subthread touched on this, I missed it. Just because an
18 executable may be used if there, doesn't necessarily make it a depend.
19 Such a tool will document the path to configuration and compilation used
20 in the particular case of the system at the time it was run, but many of
21 those "dependencies" are one of several options (could be handled by
22 virtuals, tho such would have to be hashed out and added to a list that
23 said tool uses over time, and then we have the issue of making the tool
24 smart enough to know when the virtual is required, vs. when a specific
25 instance of that virtual is required) or are only used if present, with
26 another path entirely taken if not.
27
28 It might be possible to create a tool that could help automate /some/ of
29 this, but getting it all right even a simple majority of the time would be
30 very challenging and complex. Basically, to get it right, you'll have to
31 have a human skilled in that sort of thing parse the output and match it
32 against the actual package behavior, anyway. Thus, in ordered to have any
33 sort of consistency, the requirement would be an entire team of devs doing
34 little else but verifying this information, since many ebuild maintainers
35 would be as out of their depth as someone who's only worked on x86
36 scripting all their life would be with ppc64 bitness AND endianness
37 issues. That's a LOT of developer infrastructure we are talking about
38 creating and supporting, and a LOT of developer resources that therefore
39 won't be available for other things, when developer resource shortages
40 /already/ exist.
41
42 --
43 Duncan - List replies preferred. No HTML msgs.
44 "Every nonfree program has a lord, a master --
45 and if you use the program, he is your master." Richard Stallman in
46 http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html
47
48
49 --
50 gentoo-dev@g.o mailing list