[mythtv-users] Import and Export of Program Listing Data

[Kevin Kuphal]

On 9/6/07, Robert Current <robert.current at gmail.com> wrote:
>
> > I fail to see any
> > benefit in sharing data between users since the
> > likelyhood of any two users sharing the same dataset is small.
>
> Presumably, more than two people would use it.
>
> And two users having the exact same lineup is totally un-nessessary.
>
> This was addressed with xmltvid numbers.  A cable TV user might pull
> local lineups from someone else in town that provided OTA-EIT data,
> and pull national lineups from some guy that had an xmltv file with
> complete data for a satellite listing.
>
> Yes, that's not a pretty automated system to pull from a single
> central server.  But that doesn't make it useless.


It does, however, make it complex.  You're proposing developing or asking
for developers to assist in a project to coordinate all these
micro-schedules across peers rather than saying "Download XMLTV grabber XYZ
and run it to import data".   While  it could be as simple as "Download
XMLTV micro-schedule grabber XYZ and run it to import data from a multitude
of sources", it seems an incredible undertaking in contrast with the tried
and true "download it all myself from the source" methods.

And as far has your examples, the cable TV and OTA-EIT information are not
likely a full overlap.  If nothing else, OTA-EIT is more likely a subset of
the full cable TV lineup leaving the cable TV guy still with missing
pieces.  I wonder what the likelyhood is with this micro-method of any one
person actually getting a full lineup for their area from any number of
peers considering the likely pool of users.  On the other hand, development
of a full grabber (scraping or no) gives everyone the same data (complete or
not) since there is no discrepancy between what I was able to get vs what
you were able to get other than based on the data coming from the source.
And if the source is incomplete, then it probably isn't worth grabbing in
the first place...

> The complexities surrounding the
> > distribution methods seem to outweigh
> > any perceived benefit.  A majority of
> > users will be well served simply getting
> > their data directly from the
> > original source, not trying to gather
> > it from peers.
>
> For you, that might be true.  So, there's no reason for you to consider
> this.
>
> But not everyone on the planet has a single, complete, original source
> that is providing them data freely.


True, but then also, it is highly unlikely that Person A, B, and C will each
find a 1/3 portion of a complete schedule for each of them and then somehow
coordinate to share those pieces with each other.  It would see far more
likely that Person D (Developer) wrote a grabber that collected the total
schedule and Persons A, B, and C and anyone else just ran it to download
their entire schedule.  What if Person C moves and no longer provides that
critical piece of data to A and B because they don't need it anymore?  If B
was getting it direct, I don't see the situations were A and B couldn't also
get it themselves...

Kevin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mythtv.org/pipermail/mythtv-users/attachments/20070906/c4350814/attachment.htm