Re: [VCARDDAV] ietf78 Maastricht meeting notes

[Florian Zeitz <[e-mail]>]

> based on Cyrus notes and mine, here is meeting notes. Please send diffs 
> to me. Thanks Cyrus.
> 
> Marc.
> 
Thank you for writing this up.
I have some questions/remarks about some of the decisions and myself to blame for not being present. Feel free to scold me if you feel I should better be quiet.

> ====================
> 
> IETF Vcarddav working group
> IETF 78 meeting, Maastricht, NL
> July 28th 2010, 10h30-11h30
> 
> 2. vcardrev wglc comments (Perreault, Resnick)
>     draft-ietf-vcarddav-vcardrev
> 
> Remaining issues:
> - PID and LANGUAGE parameter:
>           Proposal - put the "language grouping" inside the LANGUAGE 
> parameter value rather than overload the PID. Make cardinality 
> independent of PID. Concensus in the room to implement the proposal.
I don't really understand this. IMHO this is not overloading the PID parameter. It is meant to uniquely identify a property, different language versions are logically the same property therefore should have the same PID. I didn't really catch a compelling reason to change this and it seems to me it only complicates parsing. It would be great if someone could clarify this decision for me.

> 3. vcardxml wglc comments (Perreault)
>     draft-ietf-vcarddav-vcardxml
> 
> WGLC done. Issues fixed.
> Recent suggestion to map LANGUAGE parameter to xml:lang - no consensus 
> to use xml:lang.
So, what I grasped is that the consensus was "nice to have for XML folks, but not worth the special casing".
I suggested it, so of course I'm really biased, but personally the nice to have outweighs the special casing for me. It shouldn't be a lot of hassle to add an extra attribute while adding the LANGUAGE parameter and it would allow for constistency in environments where xml:lang is being used (e.g. I'm thinking of embedding vCards in XMPP where the stream and body tags usually have xml:lang specified).

--
Florian Zeitz