[rtcweb] Positioning draft-ietf-rtcweb-ip-address as a standards-track document

[Harald Alvestrand <[e-mail]>]

I've read through draft-ietf-rtcweb-ip-address again, with an eye to
seeing how it would read as a standards-track document.

As such, it seems to me to provide rather tentative language in a few
places, which could be strengthened by use of the CAPITALIZED WORDS.
I think this is a Good Thing; if this is standards-track, with
imperative, normative language, an implementation can claim support for
RFC XXXX and expect the statement to be understood.
(Remember - there is no protocol police. Nothing forces people to
support RFC XXXX.)

Changes needed:

- Add the RFC 2119 incantation, including Stefan's addition ("lowercase
words mean what they usually do")

- Replace all occurences of "We recommend that..." with "an
implementation MUST".

- In section 3, end of first paragraph: "   Specifically, WebRTC
should:" -> "Implementations of this specification will:"

- In section 4: "We recommend Mode 1 ... " -> "Mode 1 and mode 2 MUST be
supported. Mode 1 SHOULD be the default when the user has not granted
access to camera / microphone. Mode 2 SHOULD be the default when the
user has granted access to camera / microphone. The implementation MAY
provide means of letting the user or administrator select between modes.
These means MUST NOT be placed under the control of WebRTC applications."

I think this would make the spec a more useful tool for specifying what
an application does.



-- 
Surveillance is pervasive. Go Dark.