| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA256 |
| 3 |
|
| 4 |
Hi, |
| 5 |
|
| 6 |
What's the primary Idea behind multilib at all? |
| 7 |
Isn't it just a workaround to keep prebuild software |
| 8 |
from lazy/incapable/dead upstreams working (skype, ...)? |
| 9 |
|
| 10 |
Is there any other real use besides bragging about processor |
| 11 |
capabilities and compiling an stage1 boot loader? |
| 12 |
|
| 13 |
Please don't get me wrong. I honor the whole effort done to bring this |
| 14 |
magic into portage/..., but I see limits in the underlying file system. |
| 15 |
|
| 16 |
The current solution suggested by FHS-2.3 [1] is to use |
| 17 |
/lib<qual> and /usr/lib<qual>, which works for runtime emul- packages, |
| 18 |
since software either installed to /opt/{${PN},bin} or had no |
| 19 |
alternative in /bin or /usr/bin. |
| 20 |
|
| 21 |
On 01/27/2013 02:40 PM, Thomas Sachau wrote:> 2. How do you handle |
| 22 |
other abi-specific files like headers or binaries |
| 23 |
> and cases like binaries with abi-specific content? Is it possible |
| 24 |
> to preserve them for all requested abis or to preserve them for an |
| 25 |
> non-default abi? |
| 26 |
|
| 27 |
On 01/27/2013 05:08 PM, Pacho Ramos wrote:> Maybe installing headers |
| 28 |
in other place would be interesting :/ |
| 29 |
|
| 30 |
This is getting wired now, when we get an x86, x86_64 and x86_32 |
| 31 |
(srsly?) implementation of cp(1). |
| 32 |
|
| 33 |
Either we avoid collision python style with /bin/${PN}-<qual> and some |
| 34 |
link magic, to select the "best" <qual> according to moon phases. |
| 35 |
|
| 36 |
In the spirit of FHS, I thought about introducing /bin<qual> for some |
| 37 |
time, but this continues with other dirs. |
| 38 |
We would need /var/lib<qual> as well (/var/lib/munin/ has ABI specific |
| 39 |
.rrd files), |
| 40 |
/usr/include<qual>/ ... wait a minute. |
| 41 |
|
| 42 |
What about separating these ABIs on top dir and keeping the respective |
| 43 |
sub-trees clean, like /<qual>/{,usr/}{bin,lib}? |
| 44 |
|
| 45 |
/<qual>/ could be realised by one of these systems |
| 46 |
- - chroot (just like /<any-other-distro> chroots), needs root. |
| 47 |
- - Gentoo/PREFIX style |
| 48 |
- - modified runtime linker to pic correct LD_LIBRARY_PATH. |
| 49 |
|
| 50 |
These /<qual>/ can be anything, like different ABIs, |
| 51 |
different libc implementations, different keyword (stable, testing), |
| 52 |
different Distros, - as long as it runs with the current kernel. |
| 53 |
Well, thin-provisioning, qemu, *random virtualization*. |
| 54 |
|
| 55 |
One ABI, maybe the one that can run/chroot all others sould be defined |
| 56 |
as qual="", just like non-multilib systems. |
| 57 |
|
| 58 |
Replication of /<qual>/{home,usr/portage} can be avoided by symlinks |
| 59 |
or bind-mounts, or hardlinks. |
| 60 |
(Srsly, /usr/portage/"ebuilds,profiles,metadata,caches" has to be in |
| 61 |
/var/portage.) |
| 62 |
|
| 63 |
It'd be a pretty good solution for restraining mentioned (malicious) |
| 64 |
software, /skype/ for example. |
| 65 |
|
| 66 |
Some roundups have to be made for exhausive $PATH, X11 .desktop files, |
| 67 |
to enable starting other /<qual>/ |
| 68 |
|
| 69 |
Comments? |
| 70 |
|
| 71 |
[1] http://www.pathname.com/fhs/pub/fhs-2.3.html |
| 72 |
|
| 73 |
- -- |
| 74 |
Michael Weber |
| 75 |
Gentoo Developer |
| 76 |
web: https://xmw.de/ |
| 77 |
mailto: Michael Weber <xmw@g.o> |
| 78 |
-----BEGIN PGP SIGNATURE----- |
| 79 |
Version: GnuPG v2.0.19 (GNU/Linux) |
| 80 |
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ |
| 81 |
|
| 82 |
iF4EAREIAAYFAlEI20AACgkQknrdDGLu8JAGBAD+MPkmNxKSCrHcAnj5PUaxyTM1 |
| 83 |
Hhymj3cHaxBuTFHlK78A/2t5qJNyg1c0nc6FSePKXq+MXWp/RVDWMb5XCpfEh4dR |
| 84 |
=xmPN |
| 85 |
-----END PGP SIGNATURE----- |
Archives