I'll have a look at this after I finish wading through the bzr changes.
I think it's particularly important for things such as push and pull -
silent failures are bad.

--
Louis


On 12 June 2013 06:21, Kai Willadsen <[email protected]> wrote:

> On 11 June 2013 12:03, Louis des Landes <[email protected]> wrote:
> <snip>
> > On 10 June 2013 09:33, Kai Willadsen <[email protected]> wrote:
> >> Yeah, these aren't checked yet. All my API playground did was move the
> >> execution control into the VC module. It would be great to have a plan
> >> that made sense for error reporting for errors from VC modules.
> >
> > A bit of a look through, these are passed down to the scheduler, which
> > stores the error output on a consolestream object (self.consolestream on
> the
> > vcview)
> > I would have thought if we get output on stderr we should be raising a
> the
> > dialog - but this only actually happens if there's an IOError (ie
> something
> > gets screwed up with the pipes themselves, not that the stderr pipe
> contains
> > an error)
> > Should this behaviour be changed perhaps?
>
> Unfortunately we can't. Some VCs *love* to dump stuff to stderr, even
> when there wasn't an actual problem (*ahem* Git), presumably on the
> basis that even if you've redirected stdout somewhere you'll still see
> it. I think we need to do what you suggested and check the error code,
> raising some kind of notification (maybe just an infobar) if we get an
> error response.
>
> cheers,
> Kai
>
_______________________________________________
meld-list mailing list
[email protected]
https://mail.gnome.org/mailman/listinfo/meld-list

Reply via email to