| [Time] Name | Message |
|
[00:00] cremes
|
locks: i think there is a --path-to-libzmq option you need to use
|
|
[00:00] cremes
|
alternately, use the ffi bindings for ruby
|
|
[00:00] locks
|
I've tried pointing to /usr/local/lib which is where zmq should be, but it won't find it still
|
|
[00:00] locks
|
yeah, the ffi bindings installed ok
|
|
[18:43] shanewholloway
|
I had a devil of a time building an OSX Universal binary for ZeroMQ from git sources last night -- does anyone have a guide? Or a installable .pkg? ;)
|
|
[19:24] sustrik
|
try asking on the mailing list, there are more people there
|
|
[19:38] shanewholloway
|
thanks sustrik
|
|
[21:29] jras
|
Hi! For securing 0mq, I've seen vpn tunnels mentioned or payload encryption. Is anyone using other options? Perhaps openssl or similar?
|
|
[22:19] sd88g93
|
Hello. another problem here: i have a situation where various threads are publishing messages for a subscriber thread in another process. This works fine for the number of threads i have running, i can see threads 0, 1, 2, 3 and 4 each sending off a published message. However, when it cycles around again to the starting thread, at 0, suddenly, the subscriber thread in the other process wont recieve any messages. The published messa
|
|
[22:19] sd88g93
|
ges are being published through one global socket that the threads have access to through a mutex. any ideas ?
|
|
[22:24] mikko
|
jras: what do you mean?
|
|
[22:24] mikko
|
jras: i guess you could use stunnel
|
|
[22:34] sd88g93
|
here's the code: http://pastebin.com/DC3a4qcW
|
|
[22:35] sd88g93
|
there's "publisher code" at the top and "subscriber code" about half way down
|
|
[22:44] jras
|
mikko: Maybe stunnel might work. I'll think about it.
|
|
[22:44] jras
|
mikko: I was thinking along the lines of an ssl: or tls: transport (even though the faq notes it isn't include). Since I have adhoc connections I cannot ensure a VPN's are installed and setup. So I must bundle any securing mechanism.
|
|
[22:46] jras
|
mikko: I am open to being to the idea that payload level encryption is acceptable. I was curious what other people were doing.
|
|
[22:48] jras
|
mikko: s/to being// Typing too fast for my own good. :)
|
|
[22:57] mikko
|
jras: there is no built in encryption
|
|
[22:57] mikko
|
jras: it's a very tricky problem to solve and comes up often
|
|
[22:58] mikko
|
there are some tricky questions like if you use a forwarder device should that device be trusted?
|
