Experimental Physics and
Industrial Control System

Bruno Coudoin <[email protected]> · Wed, 04 Mar 2009 22:54:35 +0100

Le mercredi 04 mars 2009 à 14:56 -0600, Mark Rivers a écrit :
> I don't know about the CAS server, but in an IOC I post the array
>  data first and then the size.  The client waits for a monitor on
>  the size, and then does a ca_get on the array.  Obviously this is
>  not foolproof, depending on the rates of new data and latencies but
>  in practice it works well for me.

Hum, this works but its not really straightforward.

So it seems that I am not alone with this issue. Is there a plan to
enhance the protocol to make it easy.

For example, could we use the parameter 1 or 2 of the message header for
storing the payload size. Better, why do we have to send the max
possible array size in the first place and not just the actual payload
size.

Bruno.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System