1 |
On 26-11-2009 12:36:47 +0000, Duncan wrote: |
2 |
> > I think there's unfortunately no simple way to tell what should be in |
3 |
> > and what unfortunately has to be out. It depends a lot on the host |
4 |
> > system. I feel -- but I can't back this up with hard evidence -- that |
5 |
> > it are usually the libs that are not in *DEPEND that can only be |
6 |
> > available in the host system. Basically because they usually are part |
7 |
> > of the libc, which we assume to be installed. |
8 |
> > |
9 |
> > |
10 |
> > [1] http://dev.gentoo.org/~grobian/prefix-quiz |
11 |
> |
12 |
> Thanks again. That quiz was particularly enlightening (to this gentoo, |
13 |
> not prefix, user) on the type of stuff you think about, as a part of what |
14 |
> prefix /does/. |
15 |
|
16 |
The following bug is a classical example of something which just happens |
17 |
to work by accident, but is broken. It also illustrates what kind of |
18 |
changes people can expect to have to be made to their ebuilds to get |
19 |
them to work for Prefix: |
20 |
|
21 |
http://bugs.gentoo.org/show_bug.cgi?id=293901 |
22 |
|
23 |
Because the ebuild gives a direct pointer to a place in the filesystem, |
24 |
|
25 |
--with-apr=/usr/bin/apr-1-config |
26 |
|
27 |
the ebuild needs a change such that it points to the corresponding place |
28 |
in the offset-prefix installation, e.g. under "${EPREFIX}". |
29 |
|
30 |
--with-apr="${EPREFIX}"/usr/bin/apr-1-config |
31 |
|
32 |
When this variable is empty, there's obviously no change to the |
33 |
behaviour of the ebuild. |
34 |
|
35 |
|
36 |
-- |
37 |
Fabian Groffen |
38 |
Gentoo on a different level |