[rtcweb] Notes from Data discussion today

[Cullen Jennings <[e-mail]>]

Data Presentation

Issue #1 

- Will say MUST use NDATA and ignore intermediate states till we have that
- Will keep the signaling of receiver seize limits

Issue #2
- add general statement on congestion control and specific for TCP windows based congestion control 


Issue #3
- say nothing about this in our specification and leave it to the SCTP specifications
- will propose the path max retrain = association max retrans in TSVWG
- no text involving this in rtcweb specs 

Issue #4 
- will remove negation and discussion of this. It is a point for the WG developing  SCTP.
- Ted to send text on how to craft such a statement 


Issue #5 
- part of the issue here is there is a protocol on top of SCTP 
- change to text say “reset channel”


Issue #6
-  will use the the "IANA registry for WebSocket Subprotocol names "


Issue #7 
- skipped due to time 


Issue #8
- The transports documents should make the DCEP  be MTI 
- The data channel draft will not require implementation of DCEP


Issue #9 
- this is an example of something you can do with tool set we provide and may or may not be a bad idea 
- do nothing for now and be sad later