[Time] Name | Message |
[00:51] cremes
|
sustrik: go ahead and link the ruby-ffi page; i have updated it
|
[05:33] sustrik
|
cremes: liked from the frontpage
|
[05:33] sustrik
|
linked*
|
[13:10] cremes
|
i need a helping hand getting the local_lat/remote_lat sample C programs running
|
[13:10] cremes
|
http://pastie.org/953645
|
[13:11] cremes
|
it looks to me like everything is there; i'm a bit out of practice with diagnosing C compilation issues
|
[13:11] cremes
|
btw, there were no errors during the make cycle
|
[13:13] sustrik
|
perf tests are made with the library
|
[13:13] sustrik
|
you don't need to build them by hand
|
[13:14] cremes
|
i didn't build them by hand; i ran make && make install then cd perf and tried to execute them
|
[13:14] cremes
|
i built & installed the library; now i want to run the C perf tests
|
[13:20] cremes
|
when i run ldd on the binaries, it finds all libs except for libzmq.so.0
|
[13:20] cremes
|
do i need to create an env variable for it to find that library?
|
[13:25] mato
|
cremes: you need to run 'ldconfig' as root
|
[13:25] cremes
|
adding /usr/local/lib to my /etc/ld.so.conf.d/ld.so.conf file and rebuilding the linker cache solved it
|
[13:25] xla
|
did u ran ldconfig -v ?
|
[13:25] cremes
|
yep, just figured it out
|
[13:26] cremes
|
heh
|
[13:28] cremes
|
wow, the ruby ffi bindings appear to only incur about 15 usec of overhead versus a straight C implementation
|
[13:28] cremes
|
not too shabby
|
[13:34] cremes
|
hmmm... it's actually better than that...
|
[13:54] cremes
|
lots of jitter when doing timing tests with local_lat/remote_lat; ~5usec jitter between runs
|
[14:14] sustrik
|
yup, the latency ditributions tend to have fat tails
|
[14:14] sustrik
|
shrug
|
[14:17] cremes
|
np
|
[14:49] mato
|
sustrik: I've updated OpenPGM in Git; my limited testing seems to show that streams hang even when publishing way below ZMQ_RATE
|
[14:49] sustrik
|
mato: hm
|
[14:49] mato
|
sustrik: perhaps someone else will test it in anger and complain, I've kind of given up on the current state of affairs :-)
|
[14:50] sustrik
|
yup, just post a note on the mailing list
|
[14:50] sustrik
|
to make Steven aware of it
|
[14:50] mato
|
I just did
|
[14:50] sustrik
|
great
|
[14:51] sustrik
|
ah, i've meant report the problem you've seen
|
[14:51] sustrik
|
otherwise Steven will have no idea that there's still a problem there
|
[14:51] mato
|
then I'd have to write some sort of test program :-)
|
[14:52] sustrik
|
well, it's up to you
|
[14:52] mato
|
since the only thing I'm using at the moment is the camera example...
|
[14:52] mato
|
I'll see if I can hack something together that won't take longer than half an hour...
|
[15:22] mato
|
sustrik: hmm, it seems to be sporadic, i can't reproduce it any more
|
[15:22] mato
|
sustrik: maybe the switch just needed to warm up or something :-)
|
[15:22] mato
|
sustrik: i'll leave it for now and see if anyone else reports problems
|
[15:23] sustrik
|
mato: ack
|
[21:26] mike8901
|
Anyone know why https://gist.github.com/2218f508e914c162a8e2 would generate the exception "Error, no such device" ?
|
[21:26] mike8901
|
(on the bind)
|
[21:27] mike8901
|
exception: No such device, to be specific
|
[21:40] mikko
|
try 127.0.0.1 instead of localhost
|
[21:44] mike8901
|
same issue with 127.0.0.1
|