On 7/13/07, Steve Jolly <[EMAIL PROTECTED]> wrote:
Only if you think of a location URL as a key, not a value. It's certainly convenient to be able to do so, and it can work well in a web-centric world under circumstances where you can uniquely identify content by its location. In general, URIs and URLs are different beasts though, and for good reasons.
Certainly, I imagine that you'll want to be able to distinguish between a string that identifies a TV program which a set top box can expect to record off-air, and an arbitrary URL like http://google.com . I want to be able to tell my box (which isn't on the net) to record Doctor Who, but I don't want to be able to tell it to try to record google. So, if you're going to use a URL as the primary key for identifying a TV program, then I guess you want to have a definite schema for what constitutes a valid TV program URL, as opposed to any arbitrary web address. You could do that by making it a rule that all valid TV program URLs begin "http://www.bbc.co.uk/tvprogram/"; - but that might annoy other broadcasters. It would also mean that the "http://www.bbc.co.uk/tvprogram"; but of the URL was redundant, and you'd likely end up with only the absolutely necessary parts of the URL being broadcast in the on-air EPG (to save that tiny bit of valuable on-air bandwidth). At this point, the identifier you actually use starts to look quite like a CRID anyway... ;) G - Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial list archive: http://www.mail-archive.com/[email protected]/
