- From: Philipp Hoschka <ph@w3.org>
- Date: Wed, 16 Sep 1998 17:30:38 +0200
- To: www-smil@w3.org
Just had a discussion with a colleage about the difficulty of making sure that SMIL presentations play back on a wide range of SMIL players, rather than only on one particular implementation. He proposed to use helper applications for SMIL players, similar to what made integrating new media formats into HTML easy in the early days (and in some cases still today). Here's how this would work (coming from a Unix background): - To add a helper to for a media type that is not supported by a particular player, the user adds the MIME type and the application to call to a MIME configuration file (this should probably be specific to the SMIL player rather than the general MIME configuration file used by the browser and the mail reader) - If the player cannot play back a particular media type, it sends it to the helper instead. This provides a good solution at least for audio - if you can't play RealAudio, open up a RealAudio helper. It is a bit more difficult for visual media types, since the SMIL file typically also contains layout information. A possibility to resolve this is to do the layout as described, and put some information like "see helper for media object" into the space reserved for the layout - similar to the broken images/unloaded images in html. This would make sure that people have access to the full information on all SMIL players, even though the presentation may not look as "nice". Installation of helpers and thus which media types can be played back is under the control of the user. I think there is enough evidence from early days in HTML that this is a good way to bootstrap usage. Thoughts ? -Philipp Hoschka, W3C ---------------------------------------------------------------------- Philipp Hoschka | http://www.w3.org/people/hoschka | | World Wide Web Consortium | MIT-LCS ph@w3.org | 545, Technology Square Tel:(+1) 617.258.0604 | Cambridge, MA 02139 Fax:(+1) 617.258.5999 | USA ----------------------------------------------------------------------
Received on Wednesday, 16 September 1998 11:30:14 UTC