Re: AW: Properties for media fragment?

Hi all,

Ok, let's drop the video-book comparison, which indeed is not perfect.

My problem is not about time vs. interval, but rather about the non
trivial propagation of properties on sub-intervals. More formally:

let I and I' be 2 intervals, such that I' is included in I
assume that property P has value V on I,
what can we tell about property P on  I'?

my answer would be: in the general case, nothing. I suspect it strongly
depends on the semantics of property P, and may be what you need its
value for. The example I have in mind is the "right to quote" (not sure
about the translation): in France (probably elsewhere as well), you are
allowed to quote a part of a protected work without requiring the
author's consent if the part is short enough (roughly). So the copyright
on I (or at least some of its implications) may not apply on I'.

This is just an example, it might not be perfect either, but this
captures my doubts about asking questions like the one proposed by David.

I would rather stick to asking questions about and identified resource R
(whole or fragment), *but* having a mean to warn the user that several
values may apply on several part (also identified resources) of R. This
is the work of the user / the application to figure out which ones of
these values are relevant to them.

  pa


Bailer, Werner a écrit :
> I agree that in most cases a time interval would last longer than one frame, but there could be cases when time points are useful.
> 
> Concerning the copyright example: I think there are cases (e.g. news content created from different sources) where the copyright might be different for different segments. Even in the book example, the book might be a collection of articles for which the rights situation is different.
> 
> Best regards,
> Werner
> 
> ________________________________________
> Von: public-media-annotation-request@w3.org [public-media-annotation-request@w3.org] im Auftrag von Raphaël Troncy [Raphael.Troncy@cwi.nl]
> Gesendet: Mittwoch, 06. Mai 2009 08:48
> An: Pierre-Antoine Champin
> Cc: David Singer; public-media-annotation@w3.org
> Betreff: Re: Properties for media fragment?
> 
>> I'm not sure that I agree with the idea of having a value for property X
>> *at each instant*...
> 
> Hum, is it a question of time vs interval? If yes, then I agree, the
> notion of point is probably an overkill, but you might have intervals
> that are extremely short in duration ...
> In the media fragments WG, we cannot address a single time point, but it
> will always be an interval, of potentially of one frame duration.
> 
>> Back to your copyright example: for me, the copyright is on a whole
>> media entity, spanning from T to T', even if embeded in a bigger one. It
>> does not feel right to state that at time T", the copyright is C, just
>> like it would not feel right to state that the copyright of a book (or
>> its author, or its title) is the copyright (author, title, resp.) for
>> each and every word of the book.
> 
> The analogy of words in a book vs a time point of a temporal resource is
> tricky! I'm not sure the words are the right unit to consider in a book,
>   and I'm pretty sure you don't want to discrete a temporal resource.
> 
>    Raphaël
> 
> --
> Raphaël Troncy
> CWI (Centre for Mathematics and Computer Science),
> Science Park 123, 1098 XG 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/
> 
>