On Tuesday, October 14, 2014 11:50:45 PM UTC+2, ArnvShrma wrote:
>
> Hi Niphlod,
>
> I tried using psycopg2 but there are few reasons because of which I can 
> not use it. I am currently deploying travis CI to run Sahana Eden unit 
> tests. psycopg2 fails a few of the tests which it really should not. You 
> can see the build here [1].
>
> [1] https://travis-ci.org/arnavsharma93/eden/jobs/37883503
>
> So, any temporary fix of running pg8000?
>

web2py does not maintain pg8000 (although its developer is a contributor to 
web2py's code). Contact him about the issue on the json type.

Also, it seems that most of the errors (web2py doesn't maintain Sahana Eden 
either, and I'm not a developer or Sahana Eden) comes from lat/long 
differences from fixed values (again, probably, related from a quick 
copy/paste unittest suite than a carefully built one)

If only people could READ unittests instead of pointing out failing ones as 
actual failures....world will have less headaches.
E.g.:

26.0729016786571 != 26.072901678657075

let's confront them carefully... 

26.0729016786571
26.072901678657075

yep. psycopg2 returns the same value than pg8000, only with a lesser precision, 
estimated in roughly...wait, let's do calcs. 
26.07291 to 26.07292  equals to 1 meter. Every digit you loose in precision, 
adds up roughly to an order of magnitude... 
meaning 26.0729 to 26.0730 equals 10 meters.
sahana's tests are complaining about a difference of .... well, much more LESS 
than the dimension a grain of salt (actually, a lot less). 
If sahana is not trying to tie gps coordinates to atoms, psycopg2 is not the 
problem, sahana's tests are.

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to