On Saturday 12 November 2005 12:16 pm, Daniel Kristjansson wrote: > On Sat, 2005-11-12 at 11:52 -0500, Isaac Richards wrote: > > On Saturday 12 November 2005 11:49 am, Daniel Kristjansson wrote: > > > I wonder how hard would it be to add 'seconds' to the starttime? > > > > A second or two would work just as well as a minute, I'd just be worried > > about a race when creating the recordings. > > Right now we call ProgramInfo::StartedRecording() after we've > created the Ringbuffer so the threat of a race is real. But if > we called this just before creating the RingBuffer, we could > insure that the chanid w/starttime is unique in StartedRecording() > and then generate the RingBuffer filename with the possibly > modified starttime.
Allright, sounds reasonable. Uh.. make it so? =) > If the recording never starts filling up the file for any reason > (such as no-signal in the DTV case) we could delete the recorded > table entry and the file, when the recording finishes. It could > even be handled with the auto-expire of short recordings code... Good place for it. Chris said he'd be probably doing the short recordings auto-expire this weekend. On this topic, the last major things on my todo list before re-enabling live-tv: - software .nuv support (NVR.cpp switching files) - dirty exit of live-tv (kill the frontend) - toggle record - dealing with missing program data. - deleting needs to modify the tvchain table as appropriate. and possibly: - getting rid of the recorder->PauseRecorder() call in TV::PauseLiveTV? It seems extraneous now that it's not pausing most of the NVP threads anymore. There's a bunch of other stuff, of course, but everything's mostly minor enough to address after it's re-enabled. Isaac
_______________________________________________ mythtv-dev mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
