| [Time] Name | Message |
|
[14:50] btatnall
|
Hi, I'm trying to get this example to work: http://gist.github.com/500642. It's using the clrzmq binding from github. I'm not sure if I fully understand how to have a context have multiple subscribers. Anybody willing to look at it with me?
|
|
[14:51] btatnall
|
The test works about half the time.
|
|
[14:54] CIA-19
|
zeromq2: 03Martin Lucina 07master * r544b36d 10/ (src/xreq.cpp src/xreq.hpp): (log message trimmed)
|
|
[14:54] CIA-19
|
zeromq2: XREQ: Correct behaviour on hitting ZMQ_HWM
|
|
[14:54] CIA-19
|
zeromq2: This reverts part of commit 84e0c7991a9b316ed571533abc628cc1175750a3 to get
|
|
[14:54] CIA-19
|
zeromq2: correct ZMQ_HWM semantics with XREQ sockets:
|
|
[14:54] CIA-19
|
zeromq2: When sending a message to an XREQ socket, the underlying pipe is selected in
|
|
[14:54] CIA-19
|
zeromq2: a round-robin fashion. If an underlying pipe is full it is skipped. If there
|
|
[14:54] CIA-19
|
zeromq2: are no underlying pipes, or all underlying pipes are full then zmq_send()
|
|
[15:05] drbobbeaty
|
I'd be willing to help, but I'm not using it in that way - XREQ... I'm using the reliable multicast, and it's working really well with multiple zmq::socket_t and connections per socket.
|
|
[15:09] btatnall
|
Are you using multiple SUB sockets with one context?
|
|
[15:14] drbobbeaty
|
I was, but then I switched to the single zmq::socket_t with multiple connect() calls.
|
|
[15:15] drbobbeaty
|
I had a scheme where I had multiple sockets - one for each of the UDP multicast addresses I was listening to.
|
|
[15:15] drbobbeaty
|
But it seemed so much easier to have one socket and multiple connect() calls. So I switched over.
|
