Mythtv

Today was spent primarily on cleanup work.  Most of the time was creating the lircrc file for MythTV to use, which maps the remote control button codes to commands within an application.  Luckily for me I had a backup of the lircrc files that I had been using before the upgrade.  The only problem is that all of the button names changed with the update of LIRC. So instead of "Play" I had to change the config file to act on "KEY_PLAY".  They standardized all of the names on me.  So most of my time was spent figuring out what the proper name was for all of the buttons on my remote and mapping those to the correct commands in MythTV.  

Once that was done I had all of the functionality that I had in my previous install. I figured, however, that as long as I was doing this I might as well go for gold.  There were two scripts that I had found and tweaked early in my MythTV days that I was never able to get working on my previous installation of MythTV.  The first script cycled the MythTV frontend application via a button press - so if it hung I could kill the app and if it crashed and died I could restart it.  The second script suspended and woke up the computer via a button push.  I was able to wake up the frontend via a button push but was not able to get suspend to work prop ferly; every time I wanted to watch TV I would have to wait for a minute or two while the frontend booted.  Annoying, but luckily my wife was very tolerant on this problem.  As luck would have it, the frontend cycling script only needed a couple of tweaks to get it running.  The suspend script ended up using a completely new power management interface to get suspend to function properly as a user and not root, but luckily someone had already figured out how to do it and it was a non-trivial matter of Googling to find the solution.

Well, it turns out that I'm a moron.  Apparently everything was pretty much working for me already.  I had all of the pieces, I just didn't put them together correctly.  After days of work I finally have LIRC recognizing my remote inputs properly.  Basically what happened is that I configured the startup script for LIRC to load the correct driver for my IR receiver with the correct device and the correct parameters.  I had the correct config file being used to interpret the key presses correctly.  I had it all.  The problem was that I was doing all of my testing by launching the LIRC daemon from the command line instead of via the startup script that I had configured so none of my tweaks were being taken [facepalm].  Needless to say, after verifying this by running the command line with all of the correct parameters and then again verifying by rebooting the system and having it launch automatically, the remote control is now being properly identified and controlled via LIRC.  I can now launch irw and see all of my key presses as they happen.

At this point there seem to be only a few tweaks left to make before the system is ready for use again.  First, I need to setup the MythTV frontend application to automatically launch at login.  Next, I need to make MythTV always appear on top of the screen; it is currently being covered with the top and bottom panels of the window manager.  Finally, I need to configure the LIRC configuration file to interpret the button presses and pass them to the various applications, like MythTV.  Once I've got that in place, I think that all of the major components will be in place and I can put the unit back into service.  Yay.

I've done a bit more research into this remote control thing and found a couple of commands to use for debugging.  First is ir-keyboard.  If I run ir-keyboard with no options it will show me what device is found, what protocols it is capable of and which one( it is using.  In my case the output looks like:

Found /sys/class/rc/rc0/ (/dev/input/event5) with:

        Driver imon, table rc-imon-mce

        Supported protocols: RC-6

        Enabled protocols: RC-6

        Repeat delay = 500 ms, repeat period = 33 ms

In addition, I can run ir-keyboard -t and it will put the IR receiver into test mode.  Every time I press a button on the remote it puts up the codes and key names that it decodes.  According to this, everything it being seen properly and should be working.  I guess it's just an issue with the handoff to LIRC that's the problem.  I'm a step closer but still confused...

Again, I can say that the hardware is working fine and it's a software config issue that's holding me back.  Now to figure out what that is...

Ok, after a good night sleep I have cooled down enough to be productive again.  I decided to reinstall Fedora 15, this time using KDE as the desktop instead of Gnome.  I'm sure that if I had dug into it far enough I could have figured out how to resolve the windowing issues that I was facing.  Maybe there was a desktop effect that I could have turned off somewhere or something but I didn't feel like figuring out that problem when all I want to do is get the MythTV frontend up and running.  So I decided to kill that installation and load up Fedora 15 with KDE as the desktop environment this time.  The rest of the install process was identical as far ask I can tell.  After I got the graphics drivers and MythTV installed I found that I did not have that screen corruption issue.

I did still face the second problem that I reported in my last entry - MythTV playback was corrupted.  This problem turned out to not be as big of a problem as I originally thought.  To fix it I went into the setup options for TV Settings -> Playback and changed the profile to High Quality.  Once I completed that, playback proceeded as normal.  I'm sure that down the road I may tweak the playback profiles some more to get a bit more out of it, but at least it works now.  That's as far as I'm going for now.  I'm still a little annoyed that I had to go through that in the first place.  

Needless to say, WAF is not at its highest right now.  Hopefully there will not be many more setbacks in this process.  Need to get that remote figured out...