- From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
- Date: Fri, 13 Feb 2009 08:35:08 +1100
- To: Conrad Parker <conrad@metadecks.org>
- Cc: Raphaël Troncy <Raphael.Troncy@cwi.nl>, Media Fragment <public-media-fragment@w3.org>
On Fri, Feb 13, 2009 at 12:52 AM, Conrad Parker <conrad@metadecks.org> wrote: > > 2009/2/12 Raphaël Troncy <Raphael.Troncy@cwi.nl>: >> Dear Conrad, >> >>> This rationale seems to be limited to subviews of the original >>> resources, eg. an excerpt of video; in that situation it makes sense. >>> >>> There was an earlier discussion about addressing a single frame of a >>> video as an image, ie. where the returned data would be formatted as >>> valid jpeg or png. In that situation, I think the mime-type of the >>> returned data should be image. >>> (Apologies if that is outside the scope of ISSUE-2). >> >> This is perfectly in-scope of this ISSUE. However, it seems to me that the >> group consensus is that "addressing a single frame of a video as an image" >> will create a *new* resource. It is therefore *NOT* a fragment. It might be >> possible to create such a resource using a '?' followed by the same syntax >> of the media fragment URI. It might be possible to use the link header >> provided by http to provide a (typed) link towards the video resource from >> which the image comes from. The mime type of this new resource would >> certainly be image/jpeg for example. >> >> The summary is, returning an image frame from a video is not a fragment of >> this video. > > Ok that's much clearer. Perhaps that clarification should be added to > the summary, as that case is (apparently, nearly) in scope of the > issue :-) Agreed - in particular since the whole discussion about ISSUE-2 was about extracting keyframes from videos, it is important to explain this particular case. Silvia.
Received on Thursday, 12 February 2009 21:35:43 UTC