Re: [Slim] Moving forward on draft-ietf-slim-negotiating-human-language

[Gunnar Hellström <[e-mail]>]

Den 2017-11-20 kl. 21:04, skrev Paul Kyzivat:
> On 11/20/17 2:10 PM, Bernard Aboba wrote:
>> Paul said:
>>
>> "ISTM the real problem is with language tags in video media. These 
>> could indicate that the lip motions of people in the video reflect 
>> speakers of the tagged language. Or they could indicate written text 
>> in the specified language is embedded in the video. (Could be closed 
>> caption text or just signage.) Or (in the case of signed language 
>> tags) it could indicate use of sign language in the video.
>>
>> But in the end, if this is declarative about what is being sent then 
>> it isn't clear whether it is important. If it is an indication of 
>> what is being requested, then it is more important."
>>
>> [BA] Yes, that is the core of the problem.  As has been noted 
>> earlier, the modality isn't indicated explicitly. I'm not sure 
>> whether we have enough experience to know whether this represents an 
>> important deficit.  But we could indicate that the problem 
>> potentially exists and that further work might be needed.
>
> Of these, I think we can declare the lip sync "modality" to be a 
> non-issue. There is already a way to signal lip sync in SDP. If that 
> is used, then the language tag on the audio can be inferred to apply 
> as well to the linked video. That wouldn't cover a case where you can 
> see moving lips but there is no corresponding audio, but that doesn't 
> seem like a very important case to address.
>
> I'd be happy to just leave it marked as for future work.
<GH>Yes, further work, or application pre-agreement.

Gunnar
>
>     Thanks,
>     Paul
>
>> On Mon, Nov 20, 2017 at 10:53 AM, Paul Kyzivat 
>> <[e-mail] <mailto:[e-mail]>> wrote:
>>
>>     On 11/20/17 1:41 PM, Bernard Aboba wrote:
>>
>>         [BA]  This is where the ground gets less solid - we don't really
>>         have a general mechanism for distinguishing spoken and written
>>         modality among non-signed languages. Perhaps we should just say
>>         "language tags in audio media indicate spoken modality and
>>         language tags in text media indicate written modality".
>>
>>
>>     ISTM the real problem is with language tags in video media. These
>>     could indicate that the lip motions of people in the video reflect
>>     speakers of the tagged language. Or they could indicate written text
>>     in the specified language is embedded in the video. (Could be closed
>>     caption text or just signage.) Or (in the case of signed language
>>     tags) it could indicate use of sign language in the video.
>>
>>     But in the end, if this is declarative about what is being sent then
>>     it isn't clear whether it is important. If it is an indication of
>>     what is being requested, then it is more important.
>>
>>              Thanks,
>>              Paul
>>
>>
>>     _______________________________________________
>>     SLIM mailing list
>>     [e-mail] <mailto:[e-mail]>
>>     https://www.ietf.org/mailman/listinfo/slim
>>     <https://www.ietf.org/mailman/listinfo/slim>
>>
>>
>
> _______________________________________________
> SLIM mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/slim

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