[mythtv-users] tv guide line-up free

[Robert Johnston]

On 9/6/07, Jeff Wormsley <daworm at comcast.net> wrote:
> Michael T. Dean wrote:
> > Besides, you don't need to redistribute it.  Just let each person run
> > the program that gets the data in the first place.  That program simply
> > has to be able to output an XMLTV file, or, with XMLTV baseline
> > compliance could even be used directly by Myth 0.20.2 or above.
>
> I think the point being that one of the problems with scrapers is that
> they cause a heavy load on a website, leading them to frequently change
> their coding to stop scraping.  By just having one or two users hit the
> website and then distribute that data to everyone else, you eliminate
> the load on the website and distribute it amongst all the other users.
> Hopefully, this keeps the website from changing so often, and gives you
> a stable source of data.

I can't speak for other scrapers, but I've been working with DrLava on
the MSN scraper, and we're actually rather bandwidth friendly, mostly
because of the way MSN have created their AJAX site.

You can make a call to one webservice, and the MSN site will return
you a single "Page" with a full listings grid (Of basic information)
with every channel available for the entire listings period you
request. It's then a (simple) XSLT to transform that into an
XMLTV-compatible format.

If you want detailed program listings, you can call another webservice
(Per show) which will give you detailed results of every showing of
that title for a 14-day period. So if you take that and cache it (as
we are doing) it makes for a very friendly "scraper".
-- 
Robert "Anaerin" Johnston