[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [master 2/2] logging: be smarter logging UI module import errors.



On Fri, Oct 08, 2010 at 09:03:11AM +0200, Ales Kozumplik wrote:
> On 10/07/2010 09:11 PM, Brian C. Lane wrote:
> >On Thu, Oct 07, 2010 at 05:27:27PM +0200, Ales Kozumplik wrote:
> >>Use the logging module and log the traceback.
> >>---
> >>  pyanaconda/gui.py  |    5 ++++-
> >>  pyanaconda/text.py |    3 +++
> >>  2 files changed, 7 insertions(+), 1 deletions(-)
> >>
> >
> >Ack, although it would be nice if the traceback format was reversed,
> >like we do in tb dumps.
> 
> Hi,
> 
> I just wanted a quick fix of the obvious problem that when there's a
> substantial error in the module that won't even allow it to be
> imported we are left clueless (line number, function) and have to
> dig deeper.
> 
> The reversed traceback: well I'd need to (try to) instantiate
> meh.dump.ReverseExceptionDump (or duplicate the code in it) to mimic
> the format precisely, which I'd rather not (especially the code
> duplication). I always assumed the only reason we do the reverse
> traceback in dumps is "because we can", although Martin Sivak just
> told me it could be because in some cases the traceback can be cut
> off in the end and then we'd miss the important bits.
> 

Darn, I was hoping you had a better solution. That's pretty much the
same conclusion I came to.

> Anyway: is the reversed traceback really more readable for you (or
> anyone)? Perhaps it is even the time to move away from it and have
> it the same way Python has it naturally.

Well, now that you ask, it is a but odd the first few times you see it.
But I can live with it either way.


-- 
Brian C. Lane / Anaconda Team
Port Orchard, WA (PST8PDT)

Attachment: pgpP98LREuqiq.pgp
Description: PGP signature


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]