[mythtv] [Draft] Duplicate docs

Bruce Markey bjm at lvcm.com
Sun Feb 8 16:34:34 EST 2004

As I've mentioned before, all the threads on the subject of
duplicate detection are full of mis-conceptions. I've always
said that if we ever had user configurable duplicate detection,
there would need to be user documention and that I would be
willing to take on the task.

So, here it is =). Comments, questions, clarifications,

--  bjm

-------------- next part --------------
	Duplicate Checking

The MythTV scheduler can check the descriptive information for a show
to try to determine if it is a duplicate of a previously recorded or
previously scheduled episode. How to best determine if there are
duplicate episodes depends largely on how the information is presented
in the program listings. Different titles may have different formats
for their information. Therefore, MythTV allows you to choose from
several different methods to determine if showings are duplicate
episodes. The "Duplicate Check:" method can be set from the advanced
options page.


This is the safest method. The scheduler will consider all showings to
be unique without checking for duplicate subtitles or descriptions.
For this and all other methods, a showing will only be recorded if it
matches the record type (timeslot, weekslot, channel or all) and if it
'wins' the time slot in conflict resolution when there are other shows
scheduled at the same time.

Sub & Desc

This is the default method. The scheduler will consider two showing to
be duplicate if the subtitles descriptions match. It is unlikely that
you will miss a unique episode if this method is used.

Even if Empty

The scheduler will consider two showing to be duplicates if both
subtitles match and both descriptions are blank or if both
descriptions match and both subtitles are blank. This may work for
some titles, however, it is subject to the same problems as both the
"Subtitle" and "Description" methods. In most cases it may be more
accurate to choose the "Subtitle" or "Description" methods to avoid
missing episodes that you would have wanted to record.


The scheduler will consider two showing to be duplicates if both
subtitles match with no consideration of the description. This is
useful for shows that always have subtitles but don't always have
descriptions. This could cause you to miss an episode in a case where
a two part episode has the same subtitle but "Part 1" and "Part 2"
appear in the description.

This is not the best choice for sporting event either. Often games
will list the teams in the subtitle and have no description giving the
appearance that subtitle matching would make sense. However, if you
were to record "MLB Baseball" with the subtitle "Boston Red Sox vs
Chicago Cubs" and use the "Subtitle" method, you would never again see
these teams play in the World Series.


The scheduler will consider two showing to be duplicates if both
descriptions match with no consideration of the subtitle. This is the
most likely to cause you to miss episodes that you would have liked to
record. The reason is that often "generic episodes" are used as place
holders when the station does not declare which episode will be
shown. For example, "Seinfeld" may have no subtitle and a description
like "Jerry and his friends face life in New York". The episode shown
may be "The Soup Nazi" or "The Contest" but this cannot be determined
from the information in the listings. If you use the "Description"
method, a generic episode will only be recorded once. After that, you
would miss all the unique episodes that have this generic description.

Duplicate Location

The scheduler can check for duplicates against it's list of saved
recordings, it's list of all episodes that were previously recorded or
both. "All Places" is the default and you would rarely need to change
this. "Previous Recs" will always give the same result as "All Places"
unless the list of previous recordings had been manipulated in some
way. "Current Recs" will only check to consider an episode to be a
duplicate if there is a matching episode in your list of saved
recordings. After that episode has been deleted then the scheduler is
free to re-record the same episode again.

Choosing a Method

"None" is the surest way to not miss an episode but in most cases
would record duplicate episodes unnecessarily. "Sub & Desc" is the
most accurate method to skip showings that are truly duplicates. The
other methods make it more likely (not less likely) that you will miss
unique episodes.  You should use the default "Sub & Desc" unless there
is a clear pattern in the listings that make it obvious that one of
the other methods would be a better choice.

Initially you should click the "See a list of all up-coming
episodes/playtimes" button on the advanced options page to see the
current listings for a show. The fact that the current listings match
a certain pattern does not guarantee that future episodes will also
match. Therefore, you should check periodically until you are sure
that the method you've chosen is correct.

On the "Fix Scheduling Conflicts" page, showings that are considered
to be duplicates are marked with a "P" for previously recorded or "O"
if another showing of the episode is scheduled to record. If you have
chosen another method for a title, you should check these to be sure
that they are duplicate showings. If not, you have two choices; if it
is an anomaly that you don't expect to happen again, you can use
overrides to "record it anyway" or you can go to the advanced options
page to chose a safer "Duplicate Check:" method.

More information about the mythtv-dev mailing list