> From: Jouni Karvo [mailto:jouni.karvo@xxxxxx] > Kartsa writes: > > From: "Pasi Juppo" <pasi.juppo@xxxxxx> > > > > > > It would be a good debug tool if either VDR or vdradmin > stores the EPG > > > data (history data not just the new data) for later > usage. Of course the > > > file would be large but that would be a good tool when > reporting to the > > > broadcast companies if their EPG is not correct. > > > > > I think it should be VDR that would create the history > file. Its VDR that > > creates the file and removes entrys from it, isn't it. It > could also be some > > kind of setting for how old data would be stored and > possibly for which > > channels. The problem I think would be how to present the > history data > > because of possible changes in the EPG data itself. There > could be an error > > in the data for many days and in some point it would be > corrected and > > How about using a cvs repository? The problem would be of course then > _when_ to make a check-in. Perhaps a diff with the current epg.data > and the latest check-out would be made and if it differs, a new > check-in would be made... > > EPG data repository would be for debugging anyway, so I do not really > know about user interfacing. > Well the user interface was just a thought and no means necessary property. Propably a simple text file could be enough where would be all occurrences of EPG data. If it would change both old and new version would be stored possibly with a timestamp of when the data was seen. And as Jouni wrote it would be for debugging anyway and therefore a simple textfile could be enough. If diff would be used would the history data be only for changes and no other. If so one could not find those errors where EPG data would be error free and the reason for problem lies elsewhere. \\Kartsa
