Incorrect statement about ACK resending in CoAP (Esko WGLC)

[EskoDijk]

Section 9: " For example, the processing delay of CAs could be less then 2s, but in this case the EST-coaps server should be sending a CoAP ACK every 2s while processing."

This seems incorrect; the EST server will send only a single ACK to a received CON message with request. The only case where an ACK is repeated is when the server receives the same CON message again (which was a resend, done by client in case the server ACK wasn't received).

Also this sentence is unclear to me - what is it trying to say? If the CA delay is e.g. 1.5 seconds, or well below 2 seconds, the server can do one of below actions: 1) ACK the received request immediately ; then request to CA; then send the response back to client in a separate response (non-piggybacked). 2) don't ACK the received request ; but request to CA; then send the response back to client in a piggybacked CoAP response (including the ACK).

If the CA delay is ~2 seconds or more, it has to do action 1) of above list.

[csosto-pk]

Removed

For example, the processing delay of CAs could be less then 2s, but in this case the EST-coaps server should be sending a CoAP ACK every 2s while processing.

No point trying to explain when the server would be resending ACKs if they are lost in the medium and the client resends the request.