[vdr] Re: Broken Channel-ID in timers.conf prevents starting

[Klaus Schmidinger <Klaus.Schmidinger@cadsoft.de>]

Roland Rosenfeld wrote:
> 
> Hi!
> 
> I noticed a very strange phenomenon on vdr 1.2.5:
> 
> With the following entry in timers.conf vdr 1.2.5 doesn't start:
> 
> 1034649601:S19.2E-1-1073-28227:21:1258:1340:60:99:Sag' die Wahrheit (3/24):Ratespiel mit Spaß und Schwindel * Mit: Michael Antwerpes
> 
> The problem seems to be the channel id S19.2E-1-1073-28227, because I
> see the following in the syslog:
> 
> Sep 20 18:54:53 vdr vdr[15391]: ERROR: channel S19.2E-1-1073-28227 not defined
> Sep 20 18:54:53 vdr vdr[15391]: ERROR: error in /etc/vdr/timers.conf, line 12
> 
> which causes vdr 1.2.5 to stop exit on startup.
> 
> vdr 1.2.2 also wrote these two lines into the log file, but ignored
> the problem and continued to start, while 1.2.5 simply exists.
> 
> After removing the above entry from timers.conf, vdr 1.2.5 starts up
> without problems.  But if I create another timer on this channel (NDR
> FS SH), the problem arrives again, independently from the time,
> description or whether the timer is entered by hand or via vdradmin
> (autotimer).
> 
> But I still cannot see what's wrong with the channel id
> S19.2E-1-1073-28227. It is used in epg.data without problems and
> except the above mentioned timer problem, I can use this channel
> without problem:
> 
> $ grep S19.2E-1-1073-28227 /video/epg.data
> C S19.2E-1-1073-28227 NDR FS SH
> 
> $ grep 28227 /etc/vdr/channels.conf
> NDR FS SH:12110:h:S19.2E:27500:2401:2402:2404:0:28227:1:1073:16387
> 
> I use vdr 1.2.5 with the patches from Andreas Kool including the
> autopid patch and the scanner plugin.
> 
> Any idea, what's wrong with this channel id?
> 
> And is there a way to make vdr 1.2.5 more fault tolerant in this
> situation (like vdr 1.2.2 was)?

Well, I don't think that allowing VDR to start even if it looses
timers can be called "fault tolerant". It used to do so before version
1.2.3, but after some heavy complaints from Rainer Zocholl I changed
that (and I also believe that not starting at all if the timers.conf
can't be processed correctly is the right way, since then you notice
immediately that something is wrong - not after some days when you wonder
why your important recordings didn't work).

The problems here were introduced by tools that use the NID and TID values
inconsistently. VDR itself has never officially released the use of these
parameters. They will be used as soon as VDR starts processing PID data
itself.

Klaus


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe vdr" as subject.