Re: preparation of working draft as outlined by Raphael

Dear Raphaël,

On Oct 22, 2008, at 11:49 AM, Raphaël Troncy wrote:

> Dear Christian,
>
>>  for mpeg21 in http://www.w3.org/2008/WebVideo/Fragments/wiki/Existing_Technologies_Survey 
>>  you might use the following references:
>> - http://www.chiariglione.org/mpeg/technologies/mp21-fid/ 
>> index.htm.htm
>> - http://standards.iso.org/ittf/PubliclyAvailableStandards/MPEG-21_schema_files/fid/ 
>>  These are the two assets which are publicly available for ISO/IEC  
>> 21000-17.
>
> Thanks a lot for these pointers.
>
> During our face to face meeting the last 2 days, we wondered how  
> often the MPEG-21, Part 17 is practically used. More precisely:
>  - Are you aware of any client-side application that can dereference  
> what occurs after the '#' in the URI as it is defined by the Part  
> 17. For example, how many applications can parse http://www.example.com/myfile.mp4#mp(/ 
> CD/track/~time(%27npt%27,%270%27,%2760%27)) ? or http://www.example.com/myfile.mp4#mp( 
> ~region(rect(20,20,40,40)))?
>  - Do you have any ideas of how many applications make use of this  
> URI scheme? Are we likely to find such URIs on the Web?
No. I'm not aware of any industry adoptions of this particular  
standard because I simply didn't follow up this but I informed the  
main editor and contributor.

However, I think I can reply to some of your questions below...

>
>  - Is the defined scheme only applicable over the http protocol, or  
> can it be applied on top of other protocols?
FID is not bound to http and, thus, can be used with any other protocol.

>
>  - It seems to us that the Part 17 of the MPEG-21 does not specify  
> any particular communication between the UA and the server, which  
> means that  a UA always request the full resource to the server,  
> that the server sends back the whole resource (a complete movie or  
> audio file) and then the UA plays the fragment. Is it the intended  
> case? Has the working group, that has specified the Part 17,  
> consider to specify communications between UA and servers that will  
> have as consequence of not serving the complete resource?
No communications are specified afaik. Only the pointer schemes are  
specified, i.e., ffp(), offset(), mp(), and mask() including the  
resolution mechanism for each scheme. Furthermore, a schema for  
"logical models" is specified to enable the addressing of fragments of  
media content, i.e., an abstraction of the media content that is  
meaningful to the user. In this way one could describe the structure  
of a DVD, CD, or MPEG-4 files. The latter is normatively specified in  
Part 17 of MPEG-21. However, this schema for logical model  
representation could be also used to define a logical model for any  
other media to be used in the context of fragment identification.

>
>
> Will you be able to reply to the technical questions me might have  
> regarding the URI schemes defined? Or could you point us towards the  
> editors of this part?
Yes, I can try, of course but I have informed the main editor as  
already said above. You may find the editor at http://www.itscj.ipsj.or.jp/sc29/29w42911.htm#MPEG-21 
  (scroll down to 21000-17) ...

Best regards,
  -Christian

>
> Cheers.
>
>  Raphaël
>
> -- 
> Raphaël Troncy
> CWI (Centre for Mathematics and Computer Science),
> Kruislaan 413, 1098 SJ Amsterdam, The Netherlands
> e-mail: raphael.troncy@cwi.nl & raphael.troncy@gmail.com
> Tel: +31 (0)20 - 592 4093
> Fax: +31 (0)20 - 592 4312
> Web: http://www.cwi.nl/~troncy/
>