Normative/Informative status of the spec

Dear all,

During the discussion on a bug [1] and associated pull request [2], the 
question of the status of our spec was raised. Bob mentioned a previous 
discussion with the HTML WG chairs [3]. The assumption in that thread 
was that our group wants "to publish [a spec] along the same lines as 
the "Media Source Extensions Byte Stream Format Registry" was published 
and referenced from the MSE specification.".

If we want Web applications to be able to use in-band tracks in browsers 
interoperably, according to our spec, we need to be able to check 
conformance to our spec. For that, we need to have normative statements 
in our spec. Currently, the spec is in my opinion too soft about that. 
In my view, if an implementation decides to support both our spec and a 
particular media resource format (say MP4), then it shall expose tracks 
according to our spec.

This does not seem to me contradictory to the discussion with the HTML 
WG chairs because if you look at the ISOBMFF byte stream format for MSE 
[4], it does indeed use normative statements such as:
"The user agent must support setting the offset from media ..."
"These boxes must be accepted and ignored by the user agent ..."

So, my recommendation would be to rephrase our spec to be clearer as to 
what UA shall/should/should not/may ... do using normative statements. 
What's the opinion of the group here ?

Cyril

[1] https://www.w3.org/Bugs/Public/show_bug.cgi?id=26923
[2] https://github.com/w3c/HTMLSourcingInbandTracks/pull/32
[3] http://lists.w3.org/Archives/Public/public-html-admin/2014Jun/0050.html
[4] 
https://dvcs.w3.org/hg/html-media/raw-file/default/media-source/isobmff-byte-stream-format.html

-- 
Cyril Concolato
Multimedia Group / Telecom ParisTech
http://concolato.wp.mines-telecom.fr/
@cconcolato

Received on Thursday, 30 October 2014 17:20:05 UTC