Re: [sc-users] Re: OSC trouble with proce55ing and SC Server

["Rohan Drape" <rd@xxxxxxxxxxxxxxx>]

On Fri, January 6, 2006 10:24 am, JostM said:
> Is SC server refusing to send a reponse to processor, or is
> processor not getting it for some reason.

I think you can discount scsynth not sending the reply, it is *very*
reliable.

As I understand:

  Only one socket can bind to a given UDP port.

  Each UDP packet includes addresses for the receiver _&_ the sender.

  If scsynth needs to reply to an incoming packet it sends the reply
  to the address of the sender stored in the incoming packet.

The scsynth model efficiently handles both multiple concurrent clients
and clients that use different addresses over time.  [The alternative
would be to require that each incoming packet explicitly include the
address the reply should be sent to.]

However, for environments that provide send and receive mechanisms on
different sockets this model presents a problem, since the sockets by
definition must be bound to different addresses.

I believe this is an issue with some patchers (pd?, max?), where the
send and receive "objects" are isolated.

Regards,
Rohan