1 |
Joerg Hoh (joerg@××××××.org) wrote: |
2 |
> |
3 |
> On Tue, Aug 10, 2004 at 01:52:30PM +0200, Radoslaw Stachowiak wrote: |
4 |
> > |
5 |
> > it doesnt in proper way, because dispatch-conf does not store md5sums, |
6 |
> > but uses /var/db/pkg data for its comparision. But using it (db) after |
7 |
> > installation results in md5sum lost (only md5 sums of NEW version of |
8 |
> > files are stored). |
9 |
> > |
10 |
> > So there are two ways to do it: |
11 |
> > 1. store sums regardless of portage, and compare it after installation. |
12 |
> > 2. maybe, instead of md5sum, simple comparision if mtime < ctime than |
13 |
> > update is allowed would be fine? |
14 |
> > |
15 |
> > please comment about 2) because it would be trivial to implement it in |
16 |
> > dispatch-conf. |
17 |
> |
18 |
> ctime vs mtime would be acceptable; if it is combined with the |
19 |
> trivial-merge-approach, it's probably enough for most of the mentioned |
20 |
> issues. |
21 |
|
22 |
This still appears to have this problem, unless I missed it: What if the |
23 |
default is exactly what a user wants in one version, then the default changes |
24 |
in the next version? You assume an unmodified config file means the user |
25 |
doesn't care what's in it. |
26 |
|
27 |
But in this scenario, the file's MD5 would be unchanged from the initial |
28 |
snapshot because the file is distributed as the user desires -- yet an update |
29 |
wrongly overwrites this with a change that causes broken or unwanted results. |
30 |
|
31 |
There should be at least an option to NOT do the MD5 checking you propose. |
32 |
|
33 |
Thanks, |
34 |
Donnie |
35 |
|
36 |
-- |
37 |
Donnie Berkholz |
38 |
Gentoo Linux |
39 |
|
40 |
|
41 |
|
42 |
-- |
43 |
gentoo-dev@g.o mailing list |