Re: [Slim] #40 Syntax Extensibility

[Gunnar Hellström <[e-mail]>]

Den 2017-06-09 kl. 23:54, skrev Randall Gellens:
> At 5:27 PM +0000 6/8/17, Natasha Rooney wrote:
>
>>  Hi all,
>>
>>
>> <https://trac.ietf.org/trac/slim/ticket/40>https://trac.ietf.org/trac/slim/ticket/40 
>>
>>
>>  Ticket #40 discusses enhancing the syntax for the benefit of 
>> extensibility. Please review this and respond (author and all WG 
>> attendees). Please see the original text from Paul Kyzivat:
>
> I proposed a more general version of the ABNF earlier today; however, 
> Adam's comment that a space-delimited list of values is the easiest to 
> parse and most consistent with other SDP attributes makes me reluctant 
> to switch to a more complex syntax absent assurances that it wouldn't 
> add extra complexity.
>
We also do not know that it is exactly new parameters on the language 
tags that will be the desired extensions.
It could just as well be alternatives to the asterisk, valid for the 
whole media description. Or other features.

My efforts to make a requirements overview show some features that could 
be solved by new language tag parameters:
1) in 3.10.2 :    P.  Information to offering party about both common 
capabilities and
    selected languages in negotiation result

2) In 3.8.5:    P.  Coarse preference Hi/Lo per language and direction 
in different
    modalities

3) In 3.5  X. Combination of simultaneous languages in different media 
and same
    direction.

I hope that 2 and 3 can be solved either by the current ideas of using 
the asterisk and the "t" extension, or the new SDP grouping semantics.

That would leave us with item 1 as a feature that could use new language 
tag parameters. Is it worth the efforts? Do others see other extension 
needs than the ones identified through my Requirements collection under 
subject "[Slim] Requirements for the real-time case!.

/Gunnar


Gunnar


-- 
-----------------------------------------
Gunnar Hellström
Omnitor
[e-mail]
[phone]