On Tue, Feb 28, 2012 at 9:50 PM, Schmurfy <email@example.com>
If the bug I reported is confirmed by someone else It should be considered a blocker too, being able to crash any zeromq server is nothing less than critical for me.
Yep, I've confirmed that and attached a testcase that uses the raw API rather than czmq. It kind of makes sense, they're using the same pipe, and one is getting terminated, but we need to clean up the other socket. On my side it would need a lot more reading to work out what the proper fix is, so if someone more familiar can look that would be great! I'll try and have a look at the weekend if not.