[mythtv-users] Inconsistent treatment of starttime/endtime vs runtime

f-myth-users at media.mit.edu f-myth-users at media.mit.edu
Wed Apr 22 05:03:46 UTC 2009


    > Date: Tue, 21 Apr 2009 18:00:25 -0700
    > From: Rob Smith <mythtv-users at kormoc.homelinux.org>

    > You have the problem of TV 'cuts' of Dos Boot having a duration of 2
    > hours but a listed time of 8 hours. It's a extreme case, but it
    > demonstrates why this isn't correct.

Holy crap.

(Good lord, what's left?  Or is this extreme case one where it's being
aired as a miniseries?)

    > Does this make sense on why this isn't a valid solution for the
    > majority of people?

Yes, assuming that the runTime delivered in the XML really says things
like that.  Thank you.

I'm left wondering what to do in my particular case, though.  For the
sources I have, runTime (or, rather, noticing that it's too big) is
the perfect solution for the boneheaded scheduling data coming in from
certain broadcasters.  Anyone else who records from those channels is
going to get burned in the same way, unless they hugely pad.  (We're
talking up to half an hour in some cases.)  The reason we're not
hearing screams is that it's relatively rare---rare enough that I'll
bet it gets written off as bad clock skew or a single off-the-wall
schedule and not investigated.  (Especially since you have to look at
the raw XML returned by SD---if you don't save that, there's nothing
left in Myth's tables to let you know that the XML kinda-sorta knew
better.)

Given that I'm going to fix this for myself somehow (having been
pissed off by it several times---obviously enough to spend the time
tracking this down and seeing how mfdb handles it), I'm tempted to
at least post the mdfb portion of the patch to trac anyway---after
all, -someone- might be able to benefit.  (I'm speaking here of
something that just stores the info in a new column so I can use
it later in Mythweb to alert me that boneheadedness has been seen;
applying this patch in any Myth will add and update the column but
will have no other effect unless somebody wants to write code to
use it---or even use it in a power search rule or something.)

I'm guessing that at least having the column (a) lets people use
it on channels where it makes sense, and (b) might allow collecting
data mroe easily on what's going on/which channels to enable it on/etc.

I realize that 90% of the patches in trac languish for months in the
best of cases, and perhaps that this will never be applied because
nobody can figure out how to make it actually useful (assuming that
just -warning- users is deemed bad because for many channels you'd get
that warning for everything---crying wolf isn't so great, either).
But if it'll just get instantly closed as not-a-bug, I won't even
bother doing that, 'cause why waste someone's time?  I'll just patch
it for myself (and unfortunately will have to keep walking the patch
forward, making it impossible to trivially use prepackaged versions of
Myth), and people who -do- record from Sundance, TCM, etc will get
zapped by it occasionally.

So I dunno what to do about that.

Thank you again for taking the time to help nail this down.
(And I wish the technical part of the discussion about this
hadn't happened off-list; I would have seen it if it had.)


More information about the mythtv-users mailing list