Min,

Why is the null-message delimiter needed?

Brian

On Fri, Sep 3, 2010 at 9:47 AM, MinRK <[email protected]> wrote:
> Yes, I believe REQ should work with a null-message delimiter, if that fits
> your send pattern:
> xrep.send_multipart(['a', '', 'the message'])
>
> On Fri, Sep 3, 2010 at 02:55, Naveen Chawla <[email protected]> wrote:
>>
>> This is perfect for my system, and clever, thank you.
>>
>> (btw can the target be REQ? I understand that messaging REQ from XREP
>> requires a "delimiter" but isn't this just an empty message?)
>>
>> On 3 September 2010 01:44, MinRK <[email protected]> wrote:
>> > Hello,
>> > You can do this with XREP and PAIRs (or XREQs). Just set the IDENTITY of
>> > the
>> > receivers with setsockopt, and send the same thing as the first part
>> > each
>> > message from the XREP socket.
>> > For instance:
>> > (this is for Python bindings, but similar in C/C++)
>> > addr = 'tcp://127.0.0.1:12345'
>> >
>> > x = ctx.socket(XREP)
>> > x.bind('tcp://127.0.0.1:12345')
>> > a = ctx.socket(PAIR)
>> > a.setsockopt(IDENTITY, 'a')
>> > a.connect(addr)
>> > b  = ctx.socket(PAIR)
>> > b.setsockopt(IDENTITY, 'b')
>> > b.connect(addr)
>> > x.send_multipart(['a', 'hi a', 'from x'])
>> > x.send_multipart(['b', 'hi b')
>> > a.recv_multipart()
>> > [ 'hi a', 'from x' ]
>> > b.recv()
>> > 'hi b'
>> >
>> > You could also have symmetric p2p messaging through a central server if
>> > you
>> > use a modified version of the Queue device with a single XREP socket
>> > that
>> > swaps the first two parts of a multipart message (to get identity order
>> > correct).
>> > -MinRK
>> > On Thu, Sep 2, 2010 at 13:07, Naveen Chawla <[email protected]>
>> > wrote:
>> >>
>> >> I want to be able to message certain peers by identity.
>> >>
>> >> Can I set the identity using setsocketopt, and then use that identity
>> >> number from the originator-side as part of a multi-part message, to
>> >> target that socket?
>> >>
>> >> What do I need to put in that message to do this?
>> >>
>> >> N
>> >> _______________________________________________
>> >> zeromq-dev mailing list
>> >> [email protected]
>> >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>> >
>> > _______________________________________________
>> > zeromq-dev mailing list
>> > [email protected]
>> > http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>> >
>> >
>> _______________________________________________
>> zeromq-dev mailing list
>> [email protected]
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>
> _______________________________________________
> zeromq-dev mailing list
> [email protected]
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
>
>



-- 
Brian E. Granger, Ph.D.
Assistant Professor of Physics
Cal Poly State University, San Luis Obispo
[email protected]
[email protected]
_______________________________________________
zeromq-dev mailing list
[email protected]
http://lists.zeromq.org/mailman/listinfo/zeromq-dev

Reply via email to