Marko M?kel? wrote: > On Fri, Nov 04, 2005 at 03:32:50PM +0100, Klaus Schmidinger wrote: > >>Anssi Hannula wrote: >> >>>Marko M?kel? wrote: >>> >>> >>>>I'm watching Finnish DVB-T streams with VDR. Often, the EPG information >>>>for a program contains part of the description of the previous or >>>>following >>>>program. EPG bugfix level is set to 2. Should it be configured >>>>differently >>>>for Finnish transponders? Has anyone tried to contact the transponder >>>>people? >>> >>> >>>I set it to 0 and those problems were gone. >> >>This would mean that VDR's efforts to "fix" the EPG data would >>actually "damage" it?! > > > I tried setting the bugfix level from 2 to 0 and requesting an EPG update, > but it didn't fix anything. When I browsed the EPG, I saw the same program You should remove the epg.data file before (re)starting VDR in such a case, so that it freshly fetches all EPG records. > entry twice. When I looked at the details in the nearby entries, I saw > the long description of one program for both that program and the following > one. > > >>Maybe you could investigate a little more into this to find >>out what actually does the damage? >> >>Which transponder/channel is this data on? > > > I don't know how the EPG works, but I think the master DVB-T transponder > here is the one that carries the state-owned YLE channels. There are two > other transponders; I practically never use the third one. I have encountered > bugs in the EPG entries of all channels I've browsed. > > What exactly would you like to see to troubleshoot this problem? Is the > epg.data file an exact dump of the EPG data, or does it leave out something The epg.data file contains what VDR stores, after any "bugfixes". > already? Is it possible to get a raw dump of the EPG data? You could add some code for that in the eit.c file. If you run VDR with EPG bugfix level 0 (and an empty epg.data) then you should see the EPG data in the menus pretty much as it comes in. If you see a difference between bugfix level 0 and any higher level, then it's most likely VDR that modifies the data in a way you don't like (see my reply to Rolf's posting in this thread). In that case we need to refine VDR's EPG bugfix code. Klaus
