Jason Tishler wrote:
As my README indicates, Cygwin Python has some pthreads-related issues.
Given that you don't observe this behavior on other platforms, it seems
like this is a Cygwin specific problem. So, I don't think raising an
upstream bug report will help.
I would like to help with this problem, but my expectation is it will
take many hours or days to debug. I've been in the Cygwin and Python
pthreads-related code before and it is not a fun place to be. :,)
Sorry, but the reality is I don't have the time to help with a problem
of this difficulty.
Understandable. I wish it were otherwise, but no worries. I actually
worked around the problem in my real app by avoiding threads entirely --
at least in this context -- on unix (incl. cygwin), since
select.select() can be used with regular file descriptors on those
platforms. I only really needed threads+subprocess.Popen.stdout in this
context on win32, because win32 select.select() works only with sockets.
It's funny. I dimly recall running into something almost the same --
maybe exactly the same -- about two years ago. At that time I just
completely punted and dropped cygwin from the list of supported
platforms...I like my current workaround better that /that/. <g>
FWIW, I did try a few recent thread-related patches from python SVN --
some from the release25-maint branch and some I backported from trunk
(like this: http://bugs.python.org/issue874900). No joy there,
unfortunately.
--
Chuck
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
