I replied on IRC before I saw this, I suppose I should reply here too for the record. On Thu, 2012-12-20 at 15:23 +0100, Cyril Brulebois wrote: > Adam D. Barratt <[email protected]> (20/12/2012): > > Is there any merit to checking the result there (i.e. actually > > looking at err)? > > > > In any case, it looks like that's the style already used for the > > other options.
Right. There's a bunch of error handling type stuff which could be
improved upstream (which sadly seems dead :-(). I'll probably end up
taking a look at this stuff myself at some point, but that change
wouldn't be appropriate for wheezy now in any case.
> Please go ahead; thanks.
OK. I'll hopefully get to it this weekend, if not sooner.
> Adding the obligatory CC for
> > a d-i ack.
>
> Looking at it briefly, two points I thought I'd mention:
> - We have update-initramfs triggers so it looks like parts of the
> (updated) debdiff could go away.
update-initramfs itself takes care of this and defers to a trigger when
it can/should.
> - You have a double return, which is a bit awkward. ;-)
Yeah, it's the prevailing style upstream. Pointless but harmless.
Another thing I may end up fixing.
> Besides, I'll be happy to see such a bugfix for wheezy.
Thanks,
Ian.
--
Ian Campbell
Each of us bears his own Hell.
-- Publius Vergilius Maro (Virgil)
signature.asc
Description: This is a digitally signed message part
