| 1 |
Mark Knecht schreef: |
| 2 |
> Hi, |
| 3 |
> I wonder what the explanation in the emerge man page about the |
| 4 |
> --update option really means. What is meant by, and how does emerge |
| 5 |
> pick, "the best version available"? |
| 6 |
> |
| 7 |
> - Mark |
| 8 |
> |
| 9 |
|
| 10 |
For those wondering, here's the quote: |
| 11 |
|
| 12 |
--update (-u) |
| 13 |
Updates packages to the best version available, which may |
| 14 |
not always be the highest version number due to masking for testing and |
| 15 |
development. This will also update direct dependencies which may not |
| 16 |
be what you want. In general, use this option only in combination with |
| 17 |
the world or system target. |
| 18 |
|
| 19 |
I accept that this is somewhat unclear, but to explain it fully would |
| 20 |
take more space than a man page really is for. |
| 21 |
|
| 22 |
But anyway, the 'best' version available is mostly determined by the |
| 23 |
arch masks, which are determined by the developers. |
| 24 |
|
| 25 |
"Best" in this case is a somewhat poor choice of words, but at least |
| 26 |
indicates the subjectiveness of the determination. |
| 27 |
|
| 28 |
In any case, a few examples: |
| 29 |
|
| 30 |
|
| 31 |
media-video/mplayer |
| 32 |
Available versions: 1.0_pre6-r4 1.0_pre6-r5 1.0_pre6-r6 1.0_pre7 |
| 33 |
Installed: 1.0_pre7 |
| 34 |
Homepage: http://www.mplayerhq.hu/ |
| 35 |
Description: Media Player for Linux |
| 36 |
|
| 37 |
All available versions of mplayer are stable, so the 'best' version is |
| 38 |
the most recent stable. |
| 39 |
|
| 40 |
media-libs/xine-lib |
| 41 |
Available versions: ~1_rc6-r2 1_rc8-r1 1.0-r2 1.0.1-r3 ~1.0.2 |
| 42 |
~1.1.0 ~1.1.0-r1 |
| 43 |
Installed: 1.0.1-r3 |
| 44 |
Homepage: http://xine.sourceforge.net/ |
| 45 |
Description: Core libraries for Xine movie player |
| 46 |
|
| 47 |
in this case, if one was using stable arch, and had not activated |
| 48 |
unstable for this package (as I haven't), the best version is the most |
| 49 |
recent stable (1.0.1-r3), which is not the most recent version. |
| 50 |
|
| 51 |
However, if one was using ~arch, or had activated unstable for this |
| 52 |
package in /etc/portage/package.mask, then the 'best' version would be |
| 53 |
1.1.0-r1, which is the most recent revision, but not in stable Portage. |
| 54 |
|
| 55 |
media-video/ati-drivers |
| 56 |
Available versions: 8.8.25-r3 8.10.19 8.12.10 [M]8.13.3 [M]8.13.4 |
| 57 |
8.14.13 8.14.13-r1 8.14.13-r2 [M]8.14.13-r3 *8.16.20 |
| 58 |
Installed: 8.14.13-r2 |
| 59 |
Homepage: http://www.ati.com |
| 60 |
Description: Ati precompiled drivers for r350, r300, r250 |
| 61 |
and r200 chipsets |
| 62 |
|
| 63 |
Now in this case, the 'best' version is the most recent stable. The |
| 64 |
second most recent version (8.14.13-r3) is hard masked, but if I |
| 65 |
unmasked it with /etc/portage/package.unmask (and possibly also |
| 66 |
/etc/portage/package.keywords), then Portage would consider it the |
| 67 |
'best', insofar as it would attempt to install it if I upgraded the |
| 68 |
drivers. Of course, the very fact that you have to manually unmask the |
| 69 |
packages should give you pause as to whether you really want to consider |
| 70 |
this the 'best' for you. |
| 71 |
|
| 72 |
The very most recent version (8.16.20) is 'not available' -- meaning |
| 73 |
that it will never be considered the 'best' version until it returns to |
| 74 |
Portage; masking or unmasking is of no use here. I know, because I had |
| 75 |
unmasked and installed the latest drivers, which did not work well, to |
| 76 |
put it mildly, and today I synced and Portage downgraded them. I would |
| 77 |
have downgraded them manually anyway, but it was interesting to see |
| 78 |
Portage downgrade them by force despite the fact that they were still |
| 79 |
unmasked. They have so many problems that they are no longer on the |
| 80 |
Portage radar until the issues are determined to at least an extent that |
| 81 |
someone knows what to fix, and who has to fix it (Gentoo, the kernel |
| 82 |
guys if it's a kernel conflict, ATI). |
| 83 |
|
| 84 |
So, 'best' is a matter of judgement, and basically Gentoo sorts packages |
| 85 |
into categories so that you can have some context to make the judgement |
| 86 |
about what is best *for you*. If stable is best for you, then Portage |
| 87 |
will choose the stable packages (because you told it to). If unstable is |
| 88 |
best for you, then Portage will choose the unstable packages (because |
| 89 |
you told it to). If stable is generally best, but in some specific |
| 90 |
cases, unstable is best for you, then Portage will choose the stable |
| 91 |
packages except where you told it that unstable is OK. |
| 92 |
|
| 93 |
That's how it's done, mostly. |
| 94 |
|
| 95 |
HTH, |
| 96 |
Holly |
| 97 |
-- |
| 98 |
gentoo-user@g.o mailing list |
Archives