On Thursday, October 16, 2003, at 02:29 PM, Andrew Savory wrote:
On Thu, 16 Oct 2003, Stefano Mazzocchi wrote:
Andrew knows a lot about SMIL, any suggestion there?
We shouldn't need to specify audio separately if it's in the video. It should just work.
Of course, it'd be lovely if we had audio and video tracks encoded
separately, so people that just wanted to watch slides and listen would be
able to.
I have a working version for linux! I can't get a connection to apache.org
to upload it though *sob* (serious network-fu here).
Jeremy, can you hack your files thus:
- Edit index.html and add a text link for us linux users to <a href="steven.ram"/>
- Create "steven.ram", containing one line: http://www.apache.org/~jeremy/steven.smil
- Add to steven.smil the following namespaces: xmlns="http://www.w3.org/2001/SMIL20/Language"; xmlns:rn="http://features.real.com/2001/SMIL20/Extensions";
- Remove the "id" attribute from the video element
That should improve things ...
This is setup now.
<http://www.apache.org/~jeremy/gt2003/>
Thanks for your suggestions.
Steven, can you give Jeremy your presentation or take the snapshots yourself? that would help the overall quality feeling.
If they were exported as HTML, iirc powerpoint creates images of each page? We could then reference each page image from the smil.
I have them now as PDF.
Note that we might be able to do automatic redirection to the closest mirror, just like geoip-based URL redirection. shouldn't be much different from the mirror.cgi we already have
Yup, should be ridiculously easy.
How?
Now we just need volunteers to do text transcriptions of all of the audio,
so we can generate subtitles. And to listen to the video and give the
transcriptions timings ...
And/or the collaborative notes. In different languages :)
regards Jeremy
