[ViX_Misc] After upgrading to 5.0.016 AutoTimer fails

[tapkor]

The title pretty much says it all. Below is the screenshot of the error message:



Downgrading to 5.0.014 fixes the problem.

 

[bellejt]

working fine on duo 2

 

[goRt]

As op or getting clash warnings where none exist, no disabled timers set and no timers set where clashes do not exist.
Restore to .14 results in normal function.

Sent from my SM-G950F using Tapatalk

 

[jukkal2]

The same happened to my Ultimo. The error message didn't even disappear with th exit key.
Flashed back to 014, waiting for 017.

 

[abu baniaz]

Autotimer is a shared plugin between the images. There have been changes to it.

Unless people are going to post debug logs and step by step instructions on how to reproduce the problem/s, it is going to be difficult to report the matter to the person who has tried to make other improvements.

 

[abu baniaz]

The commit log is here
https://github.com/oe-alliance/enigma2-plugins/commits/2.3/autotimer

although we use another branch
https://github.com/oe-alliance/enigma2-plugins/tree/services-removed

 

[adm]

Autotimer is a shared plugin between the images. There have been changes to it.

Unless people are going to post debug logs and step by step instructions on how to reproduce the problem/s, it is going to be difficult to report the matter to the person who has tried to make other improvements.

I'm getting the same on an Extrend 1OK (build 5.0.016)

No debug or crash logs are being produced when the error occurs (and yes debug logs are enabled in the logs setting menu)

How to make it happen....
Upgrade to 5.0.016
Restore settings
Wait 4 minutes and the error message appears
Hit exit and the error message disappears
wait 10 minutes and the error message appears again

Edit 2
On my box the error message can be triggered just by setting up a new autotimer. I suspect that it occurs when trying to match/find programs in the EPG

 

[ccs]

I'm not seeing this on my et10k - I couch flashed 14 to 16 and restored settings and plugins.

Mine waits 3 minutes and then runs every 25 minutes.

Only Freeview dvb-t2 tuners.

 

[adm]

I'm not seeing this on my et10k - I couch flashed 14 to 16 and restored settings and plugins.

Only Freeview dvb-t2 tuners.

I couch flashed 13 to 16 and restored settings and plugins and have the problems.
Mine automatically polls at 3 minutes and thereafter 30 minutes
Most of my autotimers are on the satellite tuners (UK Freesat channels) although I have a few on T2 tuners (UK Freeview) channels.

I'm just about to flash back to 13 as the error message keeps appearing [edit] with a period of a lot less than 30 minutes (more like 10 minutes)

 

[abu baniaz]

Edit 2
On my box the error message can be triggered just by setting up a new autotimer. I suspect that it occurs when trying to match/find programs in the EPG

This is indeed one of the new checks.

I have tried to recreate your steps, (Was on Dev, flashed 016 Release and restored settings/plugins) but mine is not returning that error. Tried setting up new Autotimer, that was fine too. Perhaps different EPG searches.

Would you mind trying this file?
https://github.com/oe-alliance/enig...2c758e8953/autotimer/src/AutoTimerResource.py

 

[abu baniaz]

BTW, debug logs would be created constantly. A restart is required after enabling the option.



Attached is my debug log. I added "Fajr Prayer- Sehri 1st" as the autotimer in the test above.

EDIT:
Added using Blue button in Graphical EPG.

 

[tapkor]

Attached is the debug log just after the error had happened:

View attachment Enigma2_debug_2017-05-31_08-30-44.log

And yes, this happens in 5.0.016 every time AutoTimer searches or tries to search EPG.

 

[Huevos]

Recheck using the old drivers.

 

[birdman]

The only change I can see in the changes that affects extdesc is this in AutoTimer.py (old line 704, new line 732):

-                                       extdesc = event and event.getExtendedDescription() or ''
-                                       timer.extdesc = extdesc
+                                       if event:
+                                               timer.extdesc = event.getExtendedDescription() or ''
+                                       else:
+                                               remove.append(timer)

so whereas previously the timer always had a extdesc attribute, now it only has one if there is an event.

 

[adm]

Would you mind trying this file?
https://github.com/oe-alliance/enig...2c758e8953/autotimer/src/AutoTimerResource.py

I only have a compiled Python file of that name on my box - I don't have a compiler.

log file attached

 

[abu baniaz]

As mentioned by birdman, there is this in your log

Error: [Failure instance: Traceback: <type 'exceptions.AttributeError'>: 'RecordTimerEntry' object has no attribute 'extdesc'

Unless another-birdman has done it already, I'll leave a comment for the committer.


As a side note, the way I test .py files is as follows:
Rename the original.pyo file with a .bak
Transfer .py file and restart
A new.pyo file will be created
I delete the.py file straight away to avoid issues with updates etc

If new file works, well and good. If there is a crash/problem, we can delete the new .pyo and revert the rename so that we use teh original file. A restart is required.

 

[birdman]

I only have a compiled Python file of that name on my box - I don't have a compiler.

You do. It's called python. If you put the *.py file into the directory where the *.pyo file is then when python next tries to load the file (which will be when enigma2 restarts) it will see it has a source file whose modification time-stamp is later then the compiled file it has and so will compile a new file from the new source.

 

[birdman]

As mentioned by birdman, there is this in your log

Error: [Failure instance: Traceback: <type 'exceptions.AttributeError'>: 'RecordTimerEntry' object has no attribute 'extdesc'

Unless another-birdman has done it already, I'll leave a comment for the committer.

I haven't, as we're apparently not supposed to be using that branch anyway.

FWIW, I have received notification of your comment on github. But it's:

• against a commit that is nothing to do with it and
• I can't actually find any sign of it when I follow the link to github (which means I can't add a comment to it).

 

[birdman]

As mentioned by birdman, there is this in your log

Error: [Failure instance: Traceback: <type  'exceptions.AttributeError'>: 'RecordTimerEntry' object has no  attribute 'extdesc'

Unless another-birdman has done it already, I'll leave a comment for the committer.

I haven't, as we're apparently not supposed to be using that branch anyway.

FWIW, I have received notification of your comment on github. But it's:

• against a commit that is nothing to do with it and
• I can't actually find any sign of it when I follow the link to github (which means I can't add a comment to it).

OK. Found it. It is actually against the correct commit - just that github seems to have picked a random one from of a set of commits to link the mailing to.

 

[abu baniaz]

Since my initial comment in the other thread, it has come to light that the 2.3 branch get merged to the services removed branch periodically.