[Twisted-Python] test_nmea failing

Bob Ippolito bob at mastersofbranding.com
Tue Dec 17 12:26:25 EST 2002


Yuck.. floating point precision bugs.. they should have a "close 
enough" equality operator for the damn things.  I'll do some ugly hack 
to make it play nice.. but note that it's not *really* failing, that's 
perfectly acceptable output (other than the exception).

I don't have a copy of 2.1 to test against, but I'll fix it.. maybe 
change it into strings, and only check up to 10**-4 or so since that's 
about the precision you get from a NMEA message in the first place.

On Tuesday, Dec 17, 2002, at 12:05 America/New_York, Jp Calderone wrote:

> exarkun:~/projects/python/Twisted$ python2.1 admin/runtests -v -t 
> --file twisted/test/test_nmea.py
> testGPSMessages (twisted.test.test_nmea.NMEAReceiverTestCase) ... FAIL
> Expected: (-42.842648333299998, 147.30847333299999, 
> 33724.999000000003, 1)
> Results: (-42.842648333333337, 147.30847333333332, 33724.999000000003, 
> 1)
> ----------------------------------------------------------------------
> It does occassionally pass with 2.1, and it seems to always pass with 
> 2.2.





More information about the Twisted-Python mailing list