[mythtv-users] Australians: can you help me gather some data?

[Xander]

>Is this Freeview EPG via DVB exactly the same as this:
>http://www.freeview.com.au/tvguide/

Thats the one!

>... or are there some differences? Does the version available over DVB
>offer more up-to-date schedule changes or something?
Indeed it does offer about as up to date as EIT gets, with better (and actually consistent) formatting

>These are kind of enormous TODOs, just FYI. For example, if you want to
>blend EIT data and Freeview data, you need to handle the fact that two
>sources don't always call a show the exact same name. (In fact, that's
>an issue anyway: what if a datasource contains a typo one week? Can you
>correct it, or does everyone miss their show?) 
My c++ version already does this (but as is evidenced by this thread that's not always a good idea for the general user, part of the reason of a perl port); and titles are mostly a non-issue, as both sources provide a crid (there are exceptions, like Network 10 only providing the crid for now/next last I checked)

>Scraping the net for
>augmented data: that's okay until the data source changes format and
>breaks your scraper, which happens from time to time. Also you have to
>deal with multiple shows sharing the same name and typos again. 
Again, c++ version does this. Although it was not clear I meant thetvdb.com and themoviedb.com, both of which have a nice api (though I run a mirror of thetvdb to make it even easier, plan is to further develop this into something that can handle Australian show quirks too)

>Config file stuff: Almost all reported Shepherd problems are related to
>configuration and integration with MythTV, because everyone's setup is
>different in subtle ways.
I'll explain this in more detail off-list (at least off here), but I need more data than Shepherd provides (currently; I do my own channel scan so it's not tied to Myth). All I was going to have in the way of config initially was changing the XMLTVIDs that are output, for those who already have them set up.

While I'm not completely averse to the idea of combining forces (since I'm only starting out again in Perl that would likely be a huge benefit, though perhaps slowing down my learning), I think there are a few philosophical/technological differences in approach. Last time I tried feeding Shepherd with this data I don't think it did anything else to it!

I'm quite happy to discus these developer oriented points on a more appropriate medium, you (and other Shepherd devs) can possibly/probably come up with things I've not thought of; and explain Shepherd better than what I've figured out alone.