User Tale

XMLTV User Adventure Story

My "personal research" application of XMLTV seems to be pretty far from the
usual "on-screen guide for home media" usage. So I thought I might describe
how my project evolved to suit my family's needs.

First thing to know is that my wife and I are of the generation where "TV
Guide" magazine was an absolute essential. It was an important ritual to
page through the magazine, dog-earing pages with interesting content (plus
see if I could do the crossword puzzle in under two minutes). But "TV
Guide" went downhill after it introduced grids, and the DVR revolution
introduced different needs for our viewing.

Our listing needs come down to: Here's a list of shows or people or things
I'm interested in; make me a report of what's on that matches. We lose the
serendipity effect ("Oh, I didn't know about that but it sounds cool.
<dog-ear>"), but our other media sources help fill that gap.

I'll skip the part where I got a Zap2It Labs subscription and wrote a Perl
program to do our searching ... in plain text ... or the second program
which was pure XML/XSLT ... or the years in the wilderness when I couldn't
get the tools to work at all and so we had nothing. I'll also gloss over my
idiocy in failing to preserve the XSLT-based version of the tool. The
XMLTV.exe Windows package put us back on the map.

As far as retrieval goes, a setup download collects seven days (as
individual files). Then a daily update gets fresh data for today and
tomorrow, new data is added for the seventh day, and the other days' files
get renamed to their appropriate future-ness. Concatenate the seven and
sort. (The tv_cat manual says it's a mostly useless feature. Not so!
Utterly vital! Thank you for making it!)

Now for the user work. This is the slowest part of the process, as my wife
and I have lots of things to look for and each thing calls for a tv_grep.
With each tv_grep, I throw an XSLT transform to make the output HTML of
content and appearance that pleases me, and then append that HTML to the
report. Repeat etc. until all that user's data is done. Do the next user.

I'll probably horrify the power-users by noting the repeated passes mean it
takes about 35-45 minutes to complete everything. But it's running as a
scheduled event in the background, so I never see or notice it. (Yes, it's
running at around the time recommended by Schedules Direct, plus it's only
doing three days at a shot.) And it does just what we need, especially when
it comes to new episodes of our favorite shows or changed schedules for
live events (within the limitations of the listing services' resources).
Every person in the household can have their own watch list. And we don't
have to fight over or take turns with a "TV Guide."

We now return you to your regularly scheduled service.

Bob P.
Sparks, NV

------------------------------------------------------------------------------

_______________________________________________
xmltv-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/xmltv-users