PMXBOT Log file Viewer

Help | Karma | Search:

#pypa-dev logs for Friday the 30th of May, 2014

(Back to #pypa-dev overview) (Back to channel listing) (Animate logs)
[02:28:50] <agronholm> what's the situation with pip 1.5.7? there are only two open issues and the other one had a fix merged already
[02:29:01] <agronholm> the other one is a simple matter (?) of updating html5lib
[03:35:22] <Ivo> agronholm: essentially, the milestone is a necessary but not sufficient list
[03:35:47] <agronholm> any issues with 1.5.7 I'm not aware of then?
[03:40:09] <dstufft> probably that I don't remember why we were making a 1.5.7 or that we were making one so I didn't remember to see if it was time to cut one
[03:47:25] <Ivo> its not a do-this-before-sunset issue
[03:48:36] <Ivo> dstufft: do you understand html5lib's version's? https://pypi.python.org/pypi/html5lib/0.999#change-log
[03:48:48] <dstufft> 0.999 is the same as 1.0b3 afaik
[03:49:31] <Ivo> is not 0.99 the same as 1.0b3, and 0.999 newer?
[07:29:05] <agronholm> it would be really nice to get a jython fix baked in the html5 thing
[07:29:31] <agronholm> it is a bug to use lone surrogates in unicode text
[07:29:44] <agronholm> cpython may allow it but others may not, and it is so stated in the docs too
[07:30:07] <agronholm> the py4j author made the same mistake
[11:18:16] <dstufft> agronholm: yes, waiting on this https://github.com/html5lib/html5lib-python/pull/150
[17:15:12] <linovia> is it normal that pip's test suite is really slow ?
[17:15:54] <DanielHolth> yes
[17:16:32] <linovia> thanks :)
[17:19:41] <linovia> and when writing a failing test case, should it be against develop or 1.5.X ?
[17:21:15] <qwcode> 1.5.X is only for exceptional bugs at this point. most development goes to develop
[17:23:05] <linovia> ok, will do :)
[20:01:28] <linovia> that's annoying, I can't get consistent test results between my fox and travis
[20:01:53] <linovia> s/fox/tox/
[20:08:34] <dstufft> pip's tests are not that great
[20:08:37] <dstufft> what are you seeing
[20:13:06] <linovia> I'm running the test suite from another computer to see how it goes
[20:13:35] <linovia> let me grab the other one
[20:18:52] <linovia> well, I'll rerun the tests with tb=short to be able to paste something
[22:01:00] <linovia> ok, most of my issues where I hadn't bar nor mercurial installed
[22:01:48] <linovia> relaunching them. Next step is to understand why the test I added fails here and not with travis