[vdr] Re: VDR developer version 1.3.9

[a.share@t-online.de (Andreas Share)]

> Andreas Share wrote:
> >
> > Hi Klaus,
> >
> > do you think about (if you have some time and the OSD stuff is
finished;))
> > the (little buggy) timer event handling?
> > Eg:
> >
> > May 31 22:00:30 vdr vdr[3226]: timer 3 (59 2059-2150 'First Wave - Die
> > Prophezeiung') set to event Die 01.06.2004 20:15-21:00 'Mutant X'
> > May 31 22:00:30 vdr vdr[3226]: timer 4 (59 2010-2058 'Mutant X') set to
> > event Die 01.06.2004 19:25-20:15 'Stargate Kommando SG-1'
> > May 31 22:00:30 vdr vdr[3226]: timer 5 (59 1835-1930 'Andromeda') set to
> > event Die 01.06.2004 18:40-19:25 'Andromeda'
>
> When this happens, is there at all an event that would match better?
> For instance, in the 'timer 3' case, was there an event at 21:00-21:45 or
so?
> If the channel only provides present/following events, this might happen.
>
> Since I assume this happens regularly for you, could you perhaps add
> some debug output to 'int cTimer::Matches(const cEvent *Event)' and
> also maybe 'void cTimers::SetEvents(void)' to see what exactly happens
> and why it selects the wrong event?
>
> Klaus

Hi Klaus,

because vdr set in in this log the events for the following day i assume
there are more than present/following events. And you could see at timer 5
the event handling will work, if the timer "encapsulate" the whole event.
In case of timer 3 was the right event also present, i see this because vdr
restart the recording caused by an PID change at the end of this episode
from "First Wave". In this case i get an additional short recording with the
right event (Episode description).

I will later insert some debug output and mail you my results.

Greetings

Andreas Share