00:36 <spacekookie> mathju: http://imgur.com/a/ehNGmj4 got the android app to link against the Rust code again \o/

00:51 <noracodes> Yay!

00:52 <noracodes> TCP stuff update is that I spent all day packing and going down kernel bug rabbit holes to no avail. Tomorrow I'll step back and work from the protocol in, pull up Wireshark or something

11:52 <spacekookie> noracodes: hope the move goes well!

11:52 <noracodes> Thanks!

11:52 <spacekookie> I'm also wondering to what extent maybe throwing out some of the socket.rs code and replacing it with a slightly different model (the one I outlined in the cryptpad) might just make the problem go away? x)

11:53 <spacekookie> Like, as long as we can deal with send and receive failures somehow...

11:53 <noracodes> That's definitely possible tbh, this feels like we're hitting a weird edge case and nothing core to the model

11:56 <spacekookie> Yea, quite possible

11:56 <spacekookie> It's also bizzare that it doesn't happen when both tcp listeners are running in the same process O.o

18:35 <spacekookie> noracodes: kinda curious if you can reproduce the issue with a tiny TCP server that does essentially what socket.rs doos but less overkill

18:35 <noracodes> Will try asap, currently can't build bc no internet :( but the contractor is coming within this hour so I will get back to you

19:03 <spacekookie> Cool, hope the internet hookup goes smoothly!

20:49 <spacekookie> mathju: my computer just died. I'm debugging what's going on and will post once it's bock up