[mythtv] where to configure datadirect?
Tako Schotanus
quintesse at palacio-cristal.com
Fri May 7 04:36:53 EDT 2004
Bruce Markey wrote:
> Tako Schotanus wrote:
> ...
>
>> Even after reading al these posts about DataDirect and the link to
>> the original anouncement I still don't really understand why
>> DataDirect was made and who this is for. This is no comlaint or
>> anything, I just don't understand and am curious. Anybody?
>
>
> What Ivan quoted...
>
> What it boils down to is that Zap2it correctly sees that it
> is inefficient to have thousands of people each scraping hundreds
> of web pages each night. There is more load on their servers and
> no benefit to advertisers because no one is seeing any ads.
>
> Their solution is to bundle all the data in an XML format that
> users can grab in one 'hit' if they register. Zap2it benefits
> because there is less unnecessary traffic to their web servers
> and they at least can gather demographic info from the registrations
> and surveys. Users benefit because they get the data in seconds
> rather than minutes, feel less like cheaters because they are
> registered users rather than scraping web pages, and the data
> is much more complete.
>
> So the data flow changes something like this:
>
> 0.14 0.15
>
> TMS DB TMS DB
> v v
> Zap2it DB Zap2it DB
> v |
> Zap2it web pages |
> v |
> XMLTV tv_grab_na |
> v v
> mythfilldatabase mythfilldatabase
> v v
> mythconverg DB mythconverg DB
>
> So, now there is a lot of noise because XMLTV is trying to comply
> with zap2it's wishes that users convert. Therefore they are trying
> to force XMLTV users to use the tv_grab_dd for XMLTV reformatted
> data.
>
> Funny thing is, we don't need XMLTV at all anymore if we can grab
> the complete, unfiltered, unabridged, unreformatted data ourselves.
> David Shay was interested in this and Isaac encouraged him to go
> ahead and write a direct internal grabber.
>
> As a result, there is no dependency on Zap2it web pages, XMLTV,
> it's pre-requisites, it's versions, it's arbitrary changes, it's
> neglected data fields, etc.
>
> The complete DataDirect data may make some new features possible
> and old ones more efficient but we need to be careful that every-
> thing still works just fine for standard grabbers as well as the
> new data.
>
> -- bjm
Aha, so this is similar to services provided here in The Netherlands by
the likes of http://xmltv.assies.info and http://wolf.xs4all.nl/xmltv
but officially supported by the original provider of the data. Nice.
Here the system worked more or less the same (scraping webpages) until
some decided to publish the resulting XML somewhere for others to grab.
Of course this lessens the load on the original server and makes things
a lot faster but partly also because lots of people scraping the website
would be noticed and I somehow doubt that scraping webpages is legal
here, but seeing as there is no official distributer of TV guide data
here (at least none that mere mortals can get a subscription to) we're
stuck with the current system.
But anyway, as long as you don't live in the States or Canada you don't
have to worry about DataDirect and XMLTV will remain an equivalent
option besides DataDirect for the rest of the world. Is that correct?
Cheers and thanks for the explanation,
-Tako
More information about the mythtv-dev
mailing list