[mythtv-users] Anyone else have guide data problems? (SchedulesDirect only)

[f-myth-users at media.mit.edu]

    > Date: Fri, 29 Feb 2008 02:48:34 -0500 (EST)
    > From: f-myth-users at media.mit.edu

    > Yes, I believe it does.

Actually, let me be slightly more precise here.  Tivo downloads don't
happen at the same time each day for each unit; they tend to rotate
around the schedule.  Each download clearly contains data for n days
out, -including- the current day, since I've seen updates actively
change schedules only a few hours out.  Presumably the updates are
batch-created in some fashion for a given lineup and all units that
grab the update will get the same one---even if, by the time they grab
it, some of the scheduling data applies to -things that have already
aired- by then.  This is perfectly plausible if the unit is grabbing
the current day's listing---unless all units grab that day's listing
the moment before it becomes valid, some of the listing data will
necessarily be retrospective.

In this case, it looks like DSC made up their minds -just- in time to
get it into an update that went out roughly the same day as the
episode aired, and my tivo (and maybe yours) wound up grabbing and
processing the relevant update only -after- the episode had actually
aired.  It then applied the update, which presumably consists (among
other things) of a set of mappings from tivo-specific showing
identifiers to human-readable names, and I'll bet that there's only
a single occurrence of each string anywhere in that dataset (and
everything in the tivo then points at those strings) to save RAM and
so forth.  So when the update got applied, shazam---that showing
identifier now pointed to the updated string, regardless of whether
the identifier was being used to name things in the upcoming schedule
or things already in the recording database.