PMXBOT Log file Viewer

Help | Karma | Search:

#pil logs for Friday the 21st of June, 2013

(Back to #pil overview) (Back to channel listing) (Animate logs)
[14:33:35] <Arfrever> aclark: I noticed some Python-3-specific bugs in Pillow.
[14:35:25] <Arfrever> aclark: I think that PIL/_compat.py could be created with some helper functions to avoid duplication of them in multiple other files.
[14:37:48] <Arfrever> aclark: Do you like this location?
[14:49:47] <aclark> Arfrever: sure, although run it by wiredfool_ too, thanks!
[14:50:29] <Arfrever> aclark: How many days do I have before next release?
[14:52:44] <aclark> Arfrever: I'm loosely planning to cut 2.1.0 at the end of June, so about 9 days
[18:11:22] <backz> Cheers
[18:13:12] <backz> I'm looking for samples of PIL.Image called by C, inside a extension. Where can I find it?
[18:23:57] <wiredfool_> backz: I'm not aware of code that does that, but I'd bet that it works like any other python called from c
[18:24:54] <backz> ok, I'm calling it like any other python object
[18:25:15] <backz> but I still can create a Imaging object without tell Python
[18:26:10] <wiredfool_> not sure I'm following you there
[18:28:51] <backz> np
[18:30:06] <wiredfool_> got some code that I can look at?
[18:34:50] <backz> no, thanks, it's messy for while :) I'm solving it by call PIL.Image as a PyObject* :)
[18:44:28] <backz> I'm creating a tuple to call the 'frombuffer' method
[18:44:51] <backz> http://pastebin.com/e4WkzkLi but now I'm fighting with GIL :)
[18:50:12] <wiredfool_> The Py_BuildValue call looks wrong to me
[18:51:24] <wiredfool_> since you're going to at least need the initial string, I think
[18:52:10] <backz> yes, I forget the values
[18:52:57] <wiredfool_> So, the ensure is blocking?
[18:54:09] <backz> no, this is the first problem
[18:54:18] <backz> I'm getting Fatal Python error: PyMUTEX_LOCK(gil_mutex) failed
[18:55:34] <wiredfool_> well, you're definitely out of my experience range with python and threading
[18:57:18] <backz> this is not a sleep problem, why the function algorithm_wait_get_frame waits for a pthread's cond
[18:57:26] <backz> np! thanks for your time, anyway
[18:58:09] <wiredfool_> but my gut is that this is way more fundamental than PIL
[23:32:40] <backz> Cheers, I'm getting ValueError: buffer is not large enough, http://paste.pound-python.org/show/34167/