Many of you are familiar with MythTV
and know that it's a mature PVR project that's been around for 4+ years. They have a very strong active community and I'd encourge any Neuros hackers to work closely with them.
MythTV is a desktop Linux program built mostly in C++ and anticipates a relatively high end PC, so the code itself is pretty different from the OSD code, but there are lots of ares that we should be coordinating with them. A few obvious ones include anticipating usage of the OSD as a nice front end for MythTV (where the OSD could be used to view files from a mythTV on a TV in a different room or record from a different set-top box in a different room or both). I think it's also important that we learn from the MythTV crowd about all the high level stuff that they have spent the last four years working on. From integrating an EPG to controling the IR blaster, etc. There are a lot of things you learn in 4 years of hacking, let's try to make sure we "stand on the shoulders of giants" so to speak and don't reinvent the wheel.
I recently spoke to Isaac, who's the benevolent despot over there (who by coincidence I met when building the original Neuros). Our code is GPL so it's compatable and just as important, he's ok with us using it and us working together. If any of you are a part of the MythTV group, we'd love to hear from you as well.

Comments
Any progress on this?
The Neuros was brought to my attention because it was reviewed in the latest Linux Format Magazine and they said that potentially for Google Summer of Code, someone might be working on a MythTV frontend integration. That'd be my main area of interest because my wife has a strong antipathy towards computers in the bedroom, so I need a nice, silent embedded device that can play our MythTV recordings.
Has anyone looked at what the guys over at Gnome did for the Totem MythTV plugin?
I bought one to tinker with,
I bought one to tinker with, and more or less try the mythtv front end, but I dont have a backend yet... Just gona probibly get a tuner card and make my poor server work a bit more, yay Mythbusters, I get to watch your show when I want, not when my cable companys DVR decides to work! ;)
Mythtv frontend
It would be relatively painless to use mythweb (maybe slightly customised) to control a mythtv backend for managing and scheduling recordings. Mythweb can also link to the url of a file on a shared drive. Anything else you want the backend to do could be scripted and as mythtv now accepts commands via telnet then there's no major need to load up a full client on the neuros machine. All you need is a web browser on the neuros machine.
The only spaner in the works I can see is does neuros support the mpg2 streams that are output by digital tuners?
f
feedback
I need to setup RSS to feed me these comments, I'm just not getting them in a timely basis. Sorry about that. I would love to continue this on the google mailing list. Please post there if you can. Otherwise, I can post these there myself if that's ok.
MythTV ideas
I've been working with Myth for a while - mostly improving the Mac OS X support. I purchased an OSD last week for the express purpose of trying to get as much of Myth running on it as possible.
A UPnP client should be doable though it's 'not really myth'. Myth-0.20 supports UPnP clients fine (bugs not withstanding) so if the support is added to the current Neuros apps that should work.
Getting the Myth front end to work in it's current guise will mean getting Qt to work on top of Nano-X and also getting the SQL client libraries to run. Neither is 'rocket science' but I am a little concerned about the amount of memory required. Myth's videoout classes will also need to be changed to use the hardware codec packages on OSD .
Myth Backend would require mostly the same stuff (though less Qt - no UI, but the libraries would still be needed).
The other thing to bear in mind is that Myth is fairly 'agnostic' when it comes to compression formats, and although MPEG-4 is supported it's not the only format. In particular if you use a firewire connection to a settop box (or if you're using DVB) you'll be getting high bit rate MPEG-2 data. The standard Myth frontend implements most formats using software codecs (hence the high end PC requirements). It's also worth remembering that the OSD doesn't have anything better than composite output so supporting hi-def content at 720p (or 1080i) is really not worthwhile.
It's early days for me - I just got an old Linux box 'dusted off' so that I can build the BSP etc. etc. That seemed to go OK overnight - so on to the next steps.
hi def upnp
Hello.
How hard would it be to have a upnp frontend even without the myth ui (better with it, just saying for starters) that would allow you to stream and play mythtv, music and video over a network easily. Hopefully including hidef over 100mb ethernet.
just plug it in, it finds your backend, and "just works", with a remote and component out and coax digital sounds out.
This would be very appealing rather than having a big pc frontend, which can mean too much money, effort, noise etc.
I know a number of people on myth and any number who want to be. The backend is pretty easy to build, it's mysql, httpd, setting up the tuner.. yum install is half of that.
The frontend on the other hand involves lirc, alsa, mixing digital and analog sound, video drivers, sometimes xvmc... it's can be a big hassle. And, once you have a good backend, you want to have 2 or three frontends for the home theater, bedroom, etc. (which i do). if you could just buy a box that says MythCertified i think there would be a market for that? just my opinion.
onerous to have to buy and configure new custom pcs for that. Doing scheduling in MythWeb on your laptop would be more than good enough a lot of the time.
current media players on the market seem to be pretty weak, jittery video, doesn't find the backend, mpeg works but mp4 doesn't work, etc.
Post new comment