Re: [Twisted-Python] Tracebacks being dropped from exceptions when using inlineCallbacks
"glyph" == glyph <glyph@divmod.com> writes:
glyph> Just to drive that point home, Terry, I found an interesting error glyph> in your initial example. Your example does this: glyph> info = sys.exc_info() glyph> f = failure.Failure(*info) glyph> but sys.exc_info() is a tuple of (type, value, traceback), whereas glyph> Failure takes a tuple of (value, type, traceback). OMG! I guess I should learn to read. glyph> So, we still have some diagnosis to do on why you don't seem to be glyph> getting useful tracebacks from inlineCallbacks :). Now, not only glyph> can I not reproduce the bug, your reasoning doesn't make sense any glyph> more either. Glyph, sometimes I think you must be the most patient human ever. What a friendly and gentle correcting paragraph that is. I'm sure you'd make a great parent. (And I'm still a child, besides.) Still, I'd love to have been a fly on the wall when you realized I had the args reversed :-) I'm a big fan of expletives. Thanks a lot. I'm sure that's going to save me many hours of frustration and confusion. Terry
On 17 Jan, 05:20 pm, terry@jon.es wrote:
"glyph" == glyph <glyph@divmod.com> writes:
glyph> So, we still have some diagnosis to do on why you don't seem to be glyph> getting useful tracebacks from inlineCallbacks :). Now, not only glyph> can I not reproduce the bug, your reasoning doesn't make sense any glyph> more either.
Glyph, sometimes I think you must be the most patient human ever. What a friendly and gentle correcting paragraph that is. I'm sure you'd make a great parent. (And I'm still a child, besides.) Still, I'd love to have been a fly on the wall when you realized I had the args reversed :-) I'm a big fan of expletives.
Heh. Not at all. There were some expletives /leading up/ to that discovery, and I do often have to exercise... patience... when responding to certain internet fora, but not this time. Your error was quite subtle, and it had me fooled for at least an hour before I could figure out what was wrong. I was starting to think you'd discovered some insane bug in logging, in Failure, or maybe even Python exception handling (for some reason I couldn't get pdb.pm() to help me debug this, and I'm still looking into that).
Thanks a lot. I'm sure that's going to save me many hours of frustration and confusion.
No problem. I'm just glad to see there isn't a bug in Failure here!
participants (2)
-
glyph＠divmod.com -
Terry Jones