[pacman-dev] [PATCH] Only notify of dependency check in removal if performed

[Xavier]

Allan McRae wrote:
> Xavier wrote:
>> Off topic : wow, I miss mutt so much, thunderbird didn't even quote the
>> patch for me to reply ... I either need mutt with thunderbird's imap and
>> multiple accounts support or thunderbird with mutt's keyboard control
>> and neat features.
>>
> Still off topic: I spent a while searching for an extension to do this
> with no luck...

:(

>> Anyway, that patch looks fine, but it would be more logical if
>> PM_TRANS_EVT_CHECKDEPS_DONE event wasn't sent in this case either.
>>
> I did not move PM_TRANS_EVT_CHECKDEPS_DONE EVENT() in remove.c because
> there is other checking for recursive dependencies done just before the
> EVENT() statement.  Thus, there could still be some sort of dependency
> checking done when using pacman -Rsd. However, now you mention it, I
> can't think of possible frontend output for the ..._DONE counterpart
> that would make sense without the ..._START so maybe it should be
> moved.  Opinions?

Ah, I was rather thinking about just adding a second check for the 
CHECKDEPS_DONE event. Attaching a patch.

>> Note that the same problems exist in libalpm/add.c. And also that the
>> _DONE events don't do anything anymore (see pacman/callback.c), so it's
>> not a big deal
> libalpm/add.c looks fine to me.  Both the ..._START and ..._DONE EVENT()
> statements are in the if(!(trans->flags&  PM_TRANS_FLAG_NODEPS)) loop -
> line 136.  Am I missing something?
>

No, my mistake, I misread the code.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-don-t-send-CHECKDEPS_DONE-event-when-NODEPS-is-set.patch
Type: text/x-patch
Size: 851 bytes
Desc: not available
URL: <http://archlinux.org/pipermail/pacman-dev/attachments/20080112/fe0326d7/attachment.bin>