Re: [MMUSIC] Changes for draft-ietf-mmusic-trickle-ice-sip-14 - ABNF

[Thomas Stach <[e-mail]>]

Christer,



On 2018-05-09 22:57, Christer Holmberg wrote:
>
> Hi,
>
>     Hi Thomas,
>
>     I hope Paul can jump in on this one too, since it’s ABNF related :)
>
>             However, I think there are some issue with the ABNF (not related to Adam¹s
>
>             issue).
>
>             For example, the syntax says:
>
>                ice-pwd-attribute      = %s"a" "=" ice-pwd-att
>
>             Then the draft says:
>
>                with ice-pwd-attribute from [I-D.ietf-mmusic-ice-sip-sdp],
>
>         which says
>
>         ice-pwd-att           = "ice-pwd:" password
>
>         i.e. doesn't introduce a second a=
>
>     First, I think the structure in draft-ice-sip-sdp is wrong. It shall be something like:
>
>            Attribute Name: ice-pwd
>
>            Attribute Value: ice-pwd-att
>
>            Usage Level: media
>
>            Charset Dependent: No
>
>            Mux Category: TBD
>
>          The Augmented BNF syntax [RFC5234] for the attribute is:
>
>            ice-pwd-att = password
>
>            password = blah blah...
>
>     Now, whether we should use that syntax also in trickle-ice I don't know. I would like to hear Paul's opinion.
>
> You are mixing up the ABNF syntax and the IANA registration template.
>
> The ABNF syntax in section 5.4 of draft-ietf-mmusic-ice-sip-sdp-16 and 
> also section 15.4 of RFC 5245
> both define the ABNF grammar as
>
>    ice-pwd-att           = "ice-pwd" ":" password
> That is the “old” way of defining attributes. My example above is 
> according to the “new” way, that has been used lately (and, no, I 
> didn’t mix up the ABNF syntax and the IANA reg).
> Another example is in section 6 of draft-bundle (with the 
> corresponding IANA reg in section 16.3).
> Regards,
> Christer
>
That's a bad example since a=bundle-only doesn't have a value.
I'd rather prefer ( just personal preference ) the style of 
https://tools.ietf.org/html/draft-ietf-mmusic-rid-14, where the 
registration in section 11 refers to the grammar in section 10.

Nevertheless, the only thing I care about is that we don't confuse the 
reader by having two conflicting definitions of the token ice-pwd-att in 
RFC5245 and draft-ietf-mmusic-ice-sip-sdp. If you want to do changes in 
the grammar, use a distinct token.


Regards
Thomas