[pacman-dev] [PATCH] libalpm: sync changes to disc when appropriate

[Thomas Bächler]

Dan McGee schrieb:
>> Anyway, I am not opposed to a new quick maint release, this decision
>> is up to Dan anyway. I am just saying I don't see a big need for it.
>> And in any cases, we should try to move forward for a 3.3 release :)
> 
> I never said quick with this stuff at all. Does your system crash all
> the time? Or ever? Mine sure doesn't- I'm in little rush to get this
> out.

I only wanted this patch to be applied to the maint branch, I never said 
we should release it NOW.

> Thomas points to one case where his system does, but in that case he
> is going to have a lot of broken stuff as he says as other apps also
> do it wrong as well.

This case is actually a VERY common case. You experiment with new 
applications:
1) Install new package
2) Try it
3) It crashes your machine (because you are running experimental 
software with experimental kernel interfaces)
4) You just installed the package 10 seconds before the crash, so now 
your pacman db is corrupted

My point is, running pacman directly before a system crash might be more 
common than you might think. Or the other way around: System crashes are 
more likely to happen directly after a pacman action than at any other time.

> My main point here is just to do it right- I'm
> not trying to address a specific problem, just the general concern
> that is "if I write a file and my app exits, can I be sure it actually
> got written?".

Yes, of course we should ensure that our db is in a consistent state 
when pacman exits, I am glad this has been done so fast.

> Thomas- I believe you are right, we should sync the db directory when
> we write to it as well.

Okay.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
URL: <http://www.archlinux.org/pipermail/pacman-dev/attachments/20090314/bda7d685/attachment.pgp>