Re: Media Fragments Working Group: Agenda 17 March, Telecon 1000 UTC

Dear Philip,

> I notice there's no action for fixing the ABNF after I removed the
> mediasegment definition and all of the *segment definitions (those with
> '=', ignoring the encoding of the name and value).

Yes there is somehow ... it is part of ACTION-152 given to Yves, I have 
updated the title which was indeed incomplete.

> I'm not sure what the
> fix is, but would really like to see this issue settled in a way that
> everyone is happy with.

Me too :-) That's why I suggested to make it the main topic of 
tomorrow's telecon. However, I just read the email from Yves sending 
regrets and telling us it will not be ready for tomorrow.

> The problem: It's currently completely undefined at the URI level if
> something is a valid media fragment or not. I think it should be
> undefined at this level, but that we should add the constraints on the
> name-value level. Other think differently, but we need to do it in some
> way.

Yes it will be defined. The idea is that we have the constraints we can 
express in the ABNF grammar and additional constraints that cannot be 
represented (or could be represented in ABNF but due to verbosity, we 
just provide a paragraph explaining these constraints). As an example, 
one could write the 13 cases of allowed combinations of dimensions in 
the URI, or just add a paragraph explaining what are the valid combinations.

> I should be able to attend if this will be discussed (but I'd rather
> discuss it per mail, as usual).

Try to be on IRC at least during tomorrow's call. I'm not sure how much 
of that we would be able to discuss given the numerous regrets already 
sent ...
Best regards.


RaphaŽl Troncy
EURECOM, Multimedia Communications Department
2229, route des CrÍtes, 06560 Sophia Antipolis, France.
e-mail: &
Tel: +33 (0)4 - 9300 8242
Fax: +33 (0)4 - 9000 8200

Received on Tuesday, 16 March 2010 17:24:28 UTC