| 1 |
On Thu, 10 Aug 2006 12:26:10 -0500 |
| 2 |
Mike Doty <kingtaco@g.o> wrote: |
| 3 |
|
| 4 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 5 |
> Hash: SHA1 |
| 6 |
> |
| 7 |
> Donnie Berkholz wrote: |
| 8 |
> > Olivier Crete wrote: |
| 9 |
> >> It was chosen by brad_mssw to match the way it is done on ia64. |
| 10 |
> >> And I think we should continue to put the binary |
| 11 |
> >> app-emulation/emul-linux-x86-* in /emul/ and that lib32 should be |
| 12 |
> >> reserved for properly installed packages using portage whenever we |
| 13 |
> >> manage to get portage to support it. |
| 14 |
> > |
| 15 |
> > It makes sense that you wouldn't want these binary packages going |
| 16 |
> > into /lib32 or /usr/lib32, but /emul seems like an odd choice |
| 17 |
> > compared to something like /opt/lib32. |
| 18 |
|
| 19 |
I though exactly this when I saw SpanKY's query. Having a directory in |
| 20 |
'/' is not pretty. |
| 21 |
|
| 22 |
> IIRC, /emul predates FHS acceptance. also, while they are "binary" |
| 23 |
> packages, they arn't in the same catagory as binary-only packages. We |
| 24 |
> distribute them to assist multilib and to overcome problems that |
| 25 |
> portage wasn't really designed for. |
| 26 |
|
| 27 |
More generally we have varying approaches to pre-built packages; |
| 28 |
app-office/openoffice-bin installs to /usr for example, while |
| 29 |
mail-client/mozilla-thunderbird-bin and www-client/mozilla-firefox-bin |
| 30 |
install to /opt. |
| 31 |
|
| 32 |
In these cases, where they are installed on the same target |
| 33 |
architecture as they were built, I think it makes sense to have them |
| 34 |
install as if they were built with 'emerge -B' for installation via |
| 35 |
'emerge -K' - i.e. in /usr rather than /opt. |
| 36 |
|
| 37 |
x86-built binary packages for x86_64 are not the same, of course. One |
| 38 |
idea that springs to mind immediately is to put them in a |
| 39 |
{bin,include,lib...} hierarchy under /usr/<ctarget> (which is also |
| 40 |
where the compiler stuff for <ctarget> ends up). Conceptually at least |
| 41 |
(although no doubt problematic in practice) on x86_64 one could use a |
| 42 |
x86(_32) cross-compiler to build stuff to ROOT=/usr/${CTARGET}. Again |
| 43 |
in concept a /${CTARGET}/{bin,include,lib...} would exists for |
| 44 |
essential boot stuff, althought that's a bit academic. |
| 45 |
|
| 46 |
Just a thought for the pot ;) |
| 47 |
|
| 48 |
-- |
| 49 |
Kevin F. Quinn |
Archives