PMXBOT Log file Viewer

Help | Karma | Search:

#pypa logs for Thursday the 29th of October, 2015

(Back to #pypa overview) (Back to channel listing) (Animate logs)
[09:42:48] <za-softcodes> hi
[09:43:32] <za-softcodes> i am having trouble installing scipy==0.14.0 in virtual env
[09:43:49] <za-softcodes> ImportError: liblapack.so.3: cannot open shared object file: No such file or directory
[09:46:40] <za-softcodes> but if i install it on the system then it installs perfectally
[10:25:22] <ionelmc> za-softcodes: pastebin some install logs, is the virtualenv freshly created?
[10:35:27] <za-softcodes> http://pastebin.com/DiACW03q
[10:40:50] <ionelmc> za-softcodes: looks like you have a broken numpy in your virtualenv
[10:41:27] <za-softcodes> hmm
[10:41:42] <za-softcodes> let me reinstall it
[10:42:51] <za-softcodes> while uninstalling i got OSError: [Errno 13] Permission denied: '/edx/app/edxapp/venvs/edxapp/bin/f2py'
[10:43:58] <za-softcodes> ionelmc: ok sucessfully uninstalled by adding sudo
[10:44:41] <za-softcodes> ionelmc: ops! it removed the system one
[10:44:59] <ionelmc> :)
[10:45:01] <ionelmc> yes
[10:45:16] <za-softcodes> how can i uninstall the virtual env one
[10:45:16] <ionelmc> do not use virtualenv activation, it's full of pitfalls
[10:45:26] <ionelmc> just manually remove the files
[10:45:35] <za-softcodes> http://pastebin.com/nqzYr050
[10:45:42] <za-softcodes> oh
[10:46:03] <ionelmc> looks like you installed some stuff as root
[10:46:12] <ionelmc> good luck fixing it
[10:46:23] <ionelmc> just recreate your virtualenv from scratch ?
[10:46:24] <za-softcodes> Thanks
[10:46:36] <za-softcodes> that would not be possible
[10:46:47] <za-softcodes> its edx venv
[11:09:30] <nedbat> za-softcodes: come to #edx-code, let's see if we can untangle it.
[11:16:44] <za-softcodes> ionelmc: Thanks alot it was due to broken numpy
[11:17:19] <za-softcodes> nedbat: lets see what edx-code has init for me
[11:29:03] <ionelmc> nedbat: what's this edx-code thing?
[14:13:45] <nedbat> ionelmc: it's an irc channel
[14:14:00] <nedbat> which you seem to have already figured out...
[14:15:19] <ionelmc> nedbat: seems silly to ask such question, maybe i wanted to find out what edx is? ;)
[15:54:08] <dstufft> wsanchez: btw is there any insight into what versions of OSX a fix might be released for? Would it make it into El Cap? Older versions? Or no idea?
[16:07:41] <linovia> Question, are the trove classifiers for Python version checked by pip against the current Python version ? like can't install a py3 app on python2 ?
[16:09:30] <xafer> linovia: no
[16:11:16] <linovia> xafer: thanks
[17:42:17] <wsanchez> dstufft: No can comment.
[17:42:55] <dstufft> wsanchez: Okay
[17:44:21] <dstufft> wsanchez: Well, if you're involved at all, I guess I'd just request that it at least goes into El Cap and not just whatever comes after that, and if it can go further back than that too that would be great, but I understand you can't comment on where it's planning on being applied :)
[17:54:11] <wsanchez> I'm involved insofar as I file the bug and am poking them again.
[17:54:36] <wsanchez> The odds of prior-to-El Cap are basically zero.
[17:54:59] <wsanchez> The odds for El Cap are, I think, low.
[19:03:34] <alanoe> hi. I'm in a Linux with uses python 2.6 (RHEL 6), but I need to be able to execute pylint latest version, which runs on python 2.7. I tried to install python 2.7 from source and create a virtualenv using it, but I get a gcc error when trying to install some python packages via pip
[19:04:08] <alanoe> I'm trying to use virtualenv for a wrong purpose?
[19:05:28] <alanoe> I am guessing the error occurs because python-devel-2.7 is not installed in my OS
[19:11:37] <Wooble> alanoe: if you install python built from source that should include the headers you need to build stuff.
[19:16:40] <alanoe> Wooble, makes sense. it is another problem then.
[19:38:42] <dstufft> wsanchez: https://github.com/pypa/pip/issues/3165#issuecomment-152295686 ??
[19:41:39] <alanoe> snack Python package installation fails due to a missing cblas C header, I tried to install a cblas rpm package for RHEL 6, but there are other dependencies... =/
[19:44:59] <alanoe> not sure how I should proceed here
[19:46:06] <alanoe> some Python packages depend on C/C header files, and those files depend on others. I am required to install the entire chain of dependencies manually (as my distro rpm package manager wont solve them)?
[19:49:25] <alanoe> I'm talking about virtualenv btw. context: "I'm in a Linux with uses python 2.6 (RHEL 6), but I need to be able to execute pylint latest version, which runs on python 2.7. I tried to install python 2.7 from source and create a virtualenv using it, but I get a gcc error when trying to install some python packages via pip"
[20:24:51] <dangit> I’m installing a pacakge with pip via github (with -e git+ssh://….) but it doesn’t look like pip is installing the dependencies of that package. It just sees that it needs the dependencies and fails because they don’t already exist.
[20:25:00] <dangit> What is the right way to get those deps installed?
[20:29:09] <wsanchez> dstufft: (a) Let's say that hearing from wsanchez != hearing from Apple.
[20:29:46] <wsanchez> (b) Radar 20471644 is regarding rootless and pip wanting to modify files in /System.
[20:30:01] <dstufft> wsanchez: heh, I just said that because I didn't want to call you out explicitly :)
[20:30:04] <dstufft> I can edit it
[20:30:06] <wsanchez> I think the solution to 20471644 is for pip not to do that, because it shouldn't.
[20:30:33] <dstufft> I have no idea what the radar bug numbers are, if that number is pip trying to fuck with /System then I agree
[20:30:42] <wsanchez> OK
[20:31:37] <dstufft> wsanchez: "I've gotten some updates, and it appears there might movement on the sys.path problem in Apple Python. " better?
[20:34:30] <dstufft> wsanchez: meh, just deleted it instead
[20:34:32] <dstufft> it wasn't important
[20:41:08] <wsanchez> :-)
[20:43:44] <wsanchez> OK I'm claiming that pip will fix the rootless issue but that is blocked on the sys.path insanity
[20:45:03] <dstufft> wsanchez: I will totally commit to fixing that