1 Jul 2011 01:11
[issue8716] test_tk/test_tkk_guionly fails on OS X if run from buildbot slave daemon -- crashes Python
Ned Deily <report <at> bugs.python.org>
2011-06-30 23:11:47 GMT
2011-06-30 23:11:47 GMT
Ned Deily <nad <at> acm.org> added the comment: Victor, I don't understand what your patch is trying to accomplish. The problem is not that Tkinter isn't built; the problem is simply at execution time. Yes, you'll see exactly the same behavior if you are logged in via ssh and the usename you are running under is not logged in as the main GUI user. The solution is to not run Tkinter stuff in that situation. ---------- _______________________________________ Python tracker <report <at> bugs.python.org> <http://bugs.python.org/issue8716> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/python-python-bugs-list%40m.gmane.org
> Error checking can't just be probabilistic. Perhaps there's a bug in the
> file-like object; or perhaps it is a non-blocking IO object and read()
> will return None at times.
You're right, so I've raised a TypeError if PyBytes_Check fails in r_string.
> Well, it wouldn't fail any slower if you didn't do it, since you need to
> call read() very soon anyway (presumably as part of the same call to
> marshal.load()). Failing "fast" doesn't seem to bring anything here. My
> vote is for removing the complication.
Actually I misremembered the complete reason for the call - it was there to
additionally check that the passed object has a read method. I also realised -
duh - that I can read zero bytes and still get an empty bytes object back, so
I've done that, and it does look cleaner. I've also reorganised the marshal_load
function a little so it flows better.
RSS Feed