[bfcpbis] comments on draft-pascual-bfcpbis-bfcp-websocket-00

["Charles Eckel (eckelcu)" <[e-mail]>]

[as an individual]

I read the draft and have a few comments/questions I wanted to share.

Section 1, Intro
As mentioned during the BFCPBIS session last week, it would be helpful to
clarify form the start that this draft is concerned with transporting
BFCP/TCP only, not BFCP/UDP.
Also, where it reads, "a new reliable and message boundary transport for
BFCP², I think you meant to write ³Š boundary preserving transport Š².

Section 2.1 Definitions
Definitions are provided for BFCP WebSocket Client/Server. BFCP itself
defines a BFCP client and BFCP server. My assumption is that these are
independent, meaning a BFCP WebSocket Client could be a BFCP client and/or
a BFCP server. It would be helpful to clarify this in the draft.

Section 4.1
The sample GET and 101 response contain the following to identify the
WebSocket payload:

     Sec-WebSocket-Protocol: bfcp


With rfc4582bis we added the definition of a BFCP version. Have you
considered if you need to do anything to address this within WebSockets?

Cheers,
Charles