Re: [TV] Hi

["Edward Wildgoose" <Ed@xxxxxxxxxxxxxx>]

> > a) Is there an archive of this list?
>
> Not at the moment - given this is only the third mail (yours was the
> second!) I was waiting to see what the take-up/demand for it would be.
> I'll probably try and sort out an archive soon, though.

Wahoo, I get second AND fourth position!

I think that there is HUGE demand for this data, but perhaps no one knows
about it!!  Writing or helping write an XMLTV provider would probably get
you lots of attention (not to mention huge bandwidth bills).

My use for the data is for the EPG in mythtv (www.mythtv.org).  Currently
this uses a script called mythfilldatabase, which in turn relies on xmltv to
fill the SQL database.  I can write a direct way of filling the database for
my own needs, however, others will be using the various free epg type
programs with the xmltv output and so they would care about that.

Basically I don't know how you make money out of this service, and also why
you would be interested in making it easy for others to grab the data if
they only want to view it using their own offline apps, so I don't know what
kind of suggestions you are open to?

Personally, if you can provide the raw, high quality data for the same price
as mydigiguide, then count me in as a future subscriber!

> > b) I would be interested to know whether I could use the XML data from
the
> > tv2 site as a replacement from that from xmltv.  The main change would
be
> > whether it would be possible to get a few extra fields added to the XML
> > feed, for example "genre" (ie movie, documentary, etc).  Also, usage of
the
> > RFC channel names would be nice.
>
> I had a quick look at xmltv when deciding on the XML format for the
> bleb data, but given the amount of traffic received it was rather
> complex. It would certainly be possible to write an interface between my
> XML and the XMLTV DTD - I'd be willing to put this (say as a CGI script)
> on bleb, so a URL of the form:
>
>     http://www.bleb.org/cgi-bin/xmltv.cgi?c=bbc1+bbc2&d=0+1+2
>
> ...returned XMLTV data for BBC 1 and BBC2 for today, tomorrow and the
> day after.

Wow, really cool (I think).  I wonder if the value isn't in being able to
run the xmltv script on the local machine to get the data though?  If you
already have a URL generating the data though, then this will be a very thin
wrapper basically supporting the various command line functions  (I haven't
looked too closely to even see what they are...)

The point of that is *probably* these free tools are built to run the xmltv
scripts to get the data, but I haven't looked to see if this is true.  Just
thinking off the cuff here.

I think the main thing which I am missing from myth that the US users get is
colour coding of films, serials, etc.  I think this is supported on the
xmltv syntax? and your website also seems to do some of this kind of rather
cool stuff.  Could we get it in the XML file though???  (Pretty please...)

Bearing in mind that there could be a script retrieving the file, you could
also consider having a switch which compresses the data before transmitting,
it can then be expanded quite easily by the script.  I suggest zlib so that
we have windows compatibility, but bzip will be very much more efficient for
XML I think...  Either way this should save you a ton of bandwidth (by the
way I didn't check, but have you got mod_gzip installed on your server?
This is a great way to speedup download times.  Give me a shout if you need
a hand getting it configured).


> As for the data on /tv being more complete, in what way?

Oops, I misread something.  I could have sworn that the old XML file had
loads more fields in...  I wonder what I was looking at then..  Sorry

Thanks for creating this.  It looks really good!