On Wed, Jul 27, 2005 at 04:43:30PM +0100, Ian Lynagh wrote:
> Hi David,

Hello,

> I think there might be a problem with mail not getting through to the
> darcs devel list again. Attached is a mail I don't think made it.

Argh.  It looks like perhaps the mailman qrunner daemon had wasn't running
for some reason.  :( I've restarted it, which may help... Let me know if
there are any more problems (and I'm sending this to the list so others
will know to let me know if their mail starts getting dropped.

(I had been wondering why you'd been so silent on the lists recently...)

> On Fri, Jul 22, 2005 at 09:04:46AM -0400, David Roundy wrote:
> > 
> > Perhaps it would also be nice to make with_new_pending perform the
> > withSignalsBlocked that is always required, which would simplifiy the
> > command code a tad.
> 
> As long as you can have one withSignalsBlocked inside another (and I
> think you can, but haven't checked) I think that would be good.

Yeah, withSignalsBlocked looks to me like it's safe for nesting.

> > Let me know what you think about this (and this is a patch where it'd
> > be nice to get more than just Ian's reviews... too bad Juliusz is away
> > just now).  I think it's an nice step towards a higher-level
> > repository-modifying interface.
> 
> Looks good to me; any other opinions before I apply it?

Question stands. (for other people)

> I made a couple of tweaks as below (can't darcs send it without jumping
> through some hoops or resending your patch, unfortunately):
> 
> [Small tweaks to the with_new_pending patch
> Ian Lynagh <[EMAIL PROTECTED]>**20050727025308] 

They look good to me.

> hunk ./DarcsRepo.lhs 179
>   do let p = sift_for_pending origp
>          newname = name ++ ".new"
>      writePatch newname p
> -    mp <- read_pending newname
>      cur <- job
> hunk ./DarcsRepo.lhs 180
> +    mp <- read_pending newname
>      unless (isNothing mp) $
>        when (isNothing $ apply_to_slurpy (fromJust mp) cur) $ do
>        let buggyname = name ++ "_buggy"
...
> (hmm, from a human point of view that 179 is potentially confusing, and
> the 180 line should probably have been omitted)

Hmmmm.  Yeah, the line number is a bit confusing when context is
included... I'm not sure how to do it any better, though.
-- 
David Roundy

_______________________________________________
darcs-devel mailing list
[email protected]
http://www.abridgegame.org/cgi-bin/mailman/listinfo/darcs-devel

Reply via email to