PDA

View Full Version : Bug Report: Split recording, possible cause?


FarNorth
04-30-04, 12:51 PM
Last night, my wife and I were watching the second episode of 'Prime Suspect 6/Masterpiece Theater' which I had recorded last Sunday. The first episode had recorded fine, 123 minutes and we watched that first. Then, we went to the second episode. There were two recordings listed, one for 56 minutes and another for 63. I wasn't sure if this was a one-hour episode recorded twice or the first and second halves of the same episode. We watched the first and, sure enough, it cut off half way through the episode. Switched to the other and there was most of the second half. Most of it but we were missing just a bit in the transition. Now, I was nervous about missing this program since my wife and I are huge fans and have the other 5 on DVD, so I had also recorded this same program on TiVo. I checked that recording with the 2 split recordings and found that I had missed exactly 3:20 seconds of programming. In my opinion, this is not enough time to allow for a reboot.

So what happened? Sunday night is a busy night for my 921. I record 'The Sopranos,' 'Deadwood,' 'Alias,' 'Arrested Development,' and 'Cold Case' every week and have to juggle a bunch of time slots to get it all in. Then, throw in a wild card like a movie and there is a lot of starting and stopping. And we often watch the beginning of a recording while it is still recording the end. Why would it stop a 2-hour recording in the middle and pick it up? My guess is the software decided to 'verify'recordings and ran into a conflict due to overlapping timers and time buffers. Instead of just letting the recording roll, it switched tuners. I had Prime Suspect on Tuner 1, say. The 921 needs to record the 'The Sopranos' so it starts on Tuner 2. Okay, now what about 'Alias?' Okay, stick it on Tuner 1. But 'PS' is there. Well, let's put it back on Tuner 2 as soon as we can, as soon as 'Sopranos' ends....

I believe the 921 allowed me to set some timers that had short conflicts due to overlapping time buffers and then resolved them the best it could. And it is really not that big a deal; at least it resumed recording as quickly as it could.

Cheezmo
04-30-04, 01:02 PM
I just timed a reboot...

3 minutes, 20 seconds. Sounds familiar.

If they spent as much time debugging the code as you did coming up with that scenario, they'd be working perfectly by now.

More like "So what happened? Sunday night is a busy night for my 921. I record 'The Sopranos,' 'Deadwood,' 'Alias,' 'Arrested Development,' and 'Cold Case' every week and have to juggle a bunch of time slots to get it all in. Then, throw in a wild card like a movie and there is a lot of starting and stopping. And we often watch the beginning of a recording while it is still recording the end."...

so it crashed.

SimpleSimon
05-01-04, 01:45 AM
At least the 3:20 reboot time is much faster than before.

As for FarNorth's hypothesis, while well-stated, it gives WAY too much credit to whomever programmed the tuner selection code. My gut says the dang thing can't even figure out that overlapping padded shows that are on the same channel doesn't require two tuners. I'll try to run some tests to (dis)prove it.

Update: Yeah, sure enough, I'm right - 921 has no clue it's already got a tuner where it needs it.