[TV] Unable to record/watch consecutive shows on the same channel (4 Viewers)

[jaymode]

So it looks like Whurlston's patch may no longer function properly; there were some changes to the code since RC1 to fix freeing sub channels. I don't currently have time to investigate , but maybe someone else does?

 

[whurlston]

I'll try to look into it again but it may not be until the weekend. Since I'm not actually an MP user, I have other things higher on the priority list, sorry.

 

[Andrew H]

So it looks like Whurlston's patch may no longer function properly; there were some changes to the code since RC1 to fix freeing sub channels. I don't currently have time to investigate , but maybe someone else does?

Are you saying this because your testing has also flushed this out, or simply because of the update I posted on my last post? Are you still using the patch with RC1 or are you now on RC2 as well?

whurlston - Any help would surely be appreciated... I can't understand why more ATSC users aren't complaining although it sounds like it's a Win 7 issue and folks on XP are not experiencing this; moreover, it seems a good many of the developers are international and do not own ATSC tuners! Thanks for your help on this... so far this week I'm back using the default MP 1.1.0.0 RC2 download (no patch to the TVLibrary.dll) and using the "--DebugOptions with Reset Graph" and the default Generic Network Provider (not the ATSC change on the Tuner card) and it seems to be doing well! I've thrown a few "triple" scheduling overlaps and the scheduler keeps looking for a tuner every few seconds and will actually record once one becomes available. So it seems using the Reset Graph option is key in finding a true code fix.

bnsaflsujt - How are you coming along? You posted using the initial patch under RC1 - how are you finding RC2 or have you upgraded?

elliottmc - Mark, any progress on getting the team to address this as an official item in the bug-tracker? And, YES, I have found the "fix" has restored my ATSC tuners' ability to tune the same channel (same exact major with minor subchannel) back-to-back... thanks for pointing me in the right direction!

I have been a proponent of MediaPortal and been using the program for several years now - you developers do one fantastic job pulling together the code and fixes for so many countries, broadcast formats, hardware cards, Operating Systems... THANKS!

 

[whurlston]

So it seems using the Reset Graph option is key in finding a true code fix.

It was definitely one of the key parts of my previous patch. My patch forced that option for ATSC cards. If just changing that setting fixes the issue for you, it should be easy to patch it. I should have a new patch by tomorrow for you.

 

[jaymode]

So it looks like Whurlston's patch may no longer function properly; there were some changes to the code since RC1 to fix freeing sub channels. I don't currently have time to investigate , but maybe someone else does?

Are you saying this because your testing has also flushed this out, or simply because of the update I posted on my last post? Are you still using the patch with RC1 or are you now on RC2 as well?

I have not had the chance to test this, but I was a little concerned when applying whurlston's patch and noticing the differences in the tuning code and that combined with your results led to that statement/thought.

 

[whurlston]

I just tested RC2 and the changes to the subchannel handling seems to have fixed one of the issues. My patch for TvCardAtsc.cs is no longer needed but the graph still does not get disposed properly. Setting the "Reset Graph on stop" debug option does solve the other issue as Andrew mentioned or you can use the TvCardDvbBase.cs patch to force it. I don't really see the need to apply a patch simply for that since it is an option that can be set. But if you wanted to do so, here is the patch to force the option for ATSC cards with updated line numbers for SVN 25546:

Index: Implementations/DVB/Graphs/TvCardDvbBase.cs
===================================================================
--- Implementations/DVB/Graphs/TvCardDvbBase.cs	(revision 25546)
+++ Implementations/DVB/Graphs/TvCardDvbBase.cs	(working copy)
@@ -692,7 +692,7 @@
         }
         _conditionalAccess.OnStopGraph();
         // *** this should be removed when solution for graph start problem exists
[COLOR="Red"]-        if (DebugSettings.ResetGraph)[/COLOR]
[COLOR="DarkGreen"]+        if (DebugSettings.ResetGraph || _cardType == CardType.Atsc)[/COLOR]
           Decompose();
         else
           _graphState = GraphState.Created;

Unfortunately, I don't really have the time at the moment for debugging why the graph is not disposed properly but I'll try to tackle it whenever I have a few minutes to spare here and there.

 

[Andrew H]

elliottmc - Mark, any progress on getting the team to address this as an official item in the bug-tracker? And, YES, I have found the "fix" has restored my ATSC tuners' ability to tune the same channel (same exact major with minor subchannel) back-to-back... thanks for pointing me in the right direction!

elliottmc - any progress on getting a bugfix assigned? Seems this issue was also being "worked/discussed" over on this [fixed] Scheduled Recordings Did not record thread without generating a true bugfix either -- see the last post by riksmith where he believed a test.dll patch had fixed it. (And, it did for RC1 but nothing's been incorporated into the RC2/RC3 code yet.) Thanks!

 

[riksmith]

Seems this issue was also being "worked/discussed" over on this [fixed] Scheduled Recordings Did not record thread without generating a true bugfix either -- see the last post by riksmith where he believed a test.dll patch had fixed it. (And, it did for RC1 but nothing's been incorporated into the RC2/RC3 code yet.) Thanks!

That sounds a bit strange, if it was fixed in RC1 the code still is in RC2. Only possibility then is something else broke it again.

 

[riksmith]

So i have been reading up on this topic, but in the last two pages i get completely lost. Who is having problems now, and what are they?

So please tho the following:

1. If you have a problem, post ALL logs containing one failing scenario and describe exactly what you did. (Andrew H: your file on the previous page is that big that i didn't even spot an issue) SO: Clean all logs before you start.
2. Please test with clean RC2, we can't fix issues which are part of a patch that is not in official code, and also whurlston seems to suggest that his patch isn't even necessary.

 

[Andrew H]

Let me clarify and THANKS for all your help! The test.dll patch was a TVLibrary.dll (in the form of a .zip attachment) that was made available after RC1 was published. Folks evaluated test versions of TVLibrary.dll from this and the other thread and found the changes got ATSC tuners working... then on 4/14/10 RC2 was published. Since no one generated a bugfix the 'testing' done by folks on both of these threads has no way to get incorporated into the next RC, namely RC3.

I can see why it's difficult to pull these patches together - we post the issues on many different sub-groups and sometimes arrive at the same conclusion. I'm tring to speak out for ATSC users and asking for a bugfix so this can be fixed without having to use the ResetGraph Debug option.