On Sun, 10 Apr 2005 00:39:29 +0300 Rantanen Teemu teemu.rantanen@tekla.com wrote:

Timo Laitinen wrote:

...

I've tried this now, but not very successfully. The live-tv has still a gap, and the syslog gap is there too. In the recording, there is no gap

Well, yes. The PMT_SCAN_TIMEOUT=1 wasn't expected to remove the gap, only make sure the gap happens as soon as pids are changed (at the program start, not after the start)

Sorry, I was a bit unclear there, the syslog gap means that there is still a large difference between event change and pid change.

Now I could verify what happened with a familiar recording I have every week.

Apr 9 22:05:24 vdr vdr[7835]: timer 2 (1 2129-2211 'Uutisvuoto') set to event La 09.04.2005 22:05-22:26 'Frasier' Apr 9 22:05:25 vdr vdr[7843]: changing pids of channel 1 from 512+128:650=fin:2321 to 512+128:650=eng:2321

The time between event change and pid change is only ~1 second, as it used to be ~5-7 seconds. It is now only ~2 seconds after the program has started (when the program really starts, after the introduction).

Hmm... I didn't record the program, but my logs say, that

Apr 9 22:05:19 localhost vdr[4722]: channel 1 (YLE TV1) event 22:05 'Frasier' status 4 Apr 9 22:05:25 localhost vdr[4722]: channel 1: ALang changed bud don't care

but I can't confirm if the possible gap would have been inside the program or not here, as the modification prevents pid change. The latter line is a (slightly typo'ed) log message I added along with the modification, so it corresponds to the "changing" log message. The PMT_SCAN_TIMEOUT is set to 1. Without the modification and the timeout=1, I seem to still get differences of ~5 seconds, so the "wrinkle" appears well inside the program (tested it again today).

Timo