- September 1, 2008
- 21,544
- 8,236
- Home Country
-
New Zealand
I'm a bit confused.I agree it is possible but if the program had crashed at 2:55 how did it restart and crash again later?
You said:
The timestamp on both those error messages is the same time almost as when debug mode created the log zip (MediaPortalLogs_31_05_13__06_55), I am guessing they aren't related to the crash earlier in the morning?
I'm confused about how you knew there was a crash earlier in the morning and how you're grabbing these logs. So let's straighten this out.
Have you enabled the watchdog in MP config?
http://wiki.team-mediaportal.com/1_...aPortal_Configuration/11_General#General_page
...and then you're leaving TV playing overnight and finding zip files on your desktop in the morning?
OR... are you using the debug mode link in the MediaPortal folder in the Windows start menu to run the watchdog tool and collect all available logs after you observe a crash?
If the watchdog is enabled in MP config (ie. first option) then that explains how MP was restarted - the watchdog is configured to do that. It also explains why the previously noted times are so close. The system event log entry was created by Windows when the OS detected the crash; the watchdog created the zip file after the crash, and then restarted MP.
Okay.MP disappears showing the desktop and a windows error message saying Mediaportal has stopped working.
And you're not interacting with MP or the PC at the time right? TV is playing nicely one minute... then BAM...crash?
Not at all.It is like your system is not stable. Like something in the playback chain - ie. TsReader, codecs or renderer - gets upset decoding and displaying the content of the DVB-T streams.
If any one of those things were causing this surely the logs would point that out?
For a start, some of the components involved like codecs are 3rd party. In other words they don't log into the MP log files (if they log at all).
Also, crashes are usually caused by exceptional/unexpected conditions that are not handled. When a crash happens during some particular interaction with MP then we have some context to help us reproduce and find the cause. When a crash just happens and you weren't actively interacting with MP... as appears to be the case here... then it can be quite a bit more difficult to find the source of the problem, let alone fix it.