[MERGE] Two little fixes/cleanups
Martin Pool
mbp at sourcefrog.net
Mon Dec 12 05:09:24 GMT 2005
On 10 Dec 2005, John Arbash Meinel <john at arbash-meinel.com> wrote:
> Wouter van Heyst wrote:
> > On Sat, Dec 10, 2005 at 04:13:16PM -0600, Michael Ellerman wrote:
>
> ...
>
> >>> bzr.dev> ./bzr cat
> >>> bzr: ERROR: exceptions.KeyError: 'changeset'
> >>> at /home/larstiq/src/bzr.dev/bzrlib/commands.py line 207
> >>> in options
> >>
> >>You don't? I hit this when one of my old plugins skewed with bzr such that the
> >>option it wanted wasn't there anymore - it took me a couple of minutes to
> >>work that out though. Having the message at least tell me it was looking for
> >>an option would have helped I think.
> >
> >
> > Oh yes, I agree with noting it is an option, I should have been more
> > explicit. The semi-traceback does more for me knowing where to look then
> > the 'Internal check failed'. Of course, by the time I had figured out
> > what your patch was supposed to do, I had been mucking around on line
> > 207 already.
> >
> > If there could be information on where the offending option comes from,
> > that would be great.
>
> I think in general tracebacks are intended to be hidden in ~/.bzr.log,
> and not shown on the screen.
> I think at present they are always logged if you user BZR_DEBUG=1,
> though I do believe that Martin intends for them to always be logged.
The behaviour is supposed to be this: tracebacks are always logged to
.bzr.log, and the file/line/function is shown for exceptions which don't
subclass BzrError. The reasoning is that is that we shouldn't show any
internal details if it's just a user or environmental error. If there
was a bzr bug then it's good to give at least the location to help.
--
Martin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20051212/2196ddca/attachment.pgp
More information about the bazaar
mailing list