Slashdot Mirror
Linux Looms Large in DVRs, PVRs
An anonymous reader writes "According to an article at LinuxDevices there's a new fanless digital entertainment center reference design based on Linux and the MythTV open source DVR (digital video recorder) software. The 'Royal Linux Media Center' runs ESG's Royal Linux OS on a Transmeta development board based on its Efficeon chip. Linux has been increasingly popular in DVRs and PVRs, with examples including TiVo (of course), HP's recently unveiled Linux media hub, i3's Mood box, Interact-TV's Telly, Siemens' Speedstream, VWB's MediaReady 4000, Amino's AmiNet500, Sharp's Galileo, Dream-Multimedia-Tv's Dreambox,
NEC's AX10, and Sony's CoCoon, to name a few."
.. which mythtv doesn't then have? because that's what you're implying by the way you're puffing them.
looking at the features, mythtv looks like it does more, a LOT more.
including stuff like picture in picture, multi card support - and get this, transparent multi machine support: "Distributed architecture allowing multiple recording machines and multiple playback machines on the same network, completely transparent to the user.", rss, mpeg4, mpeg2 decoders/encoders and a whole lot of other stuff.
maybe mythtv gets mentioned more often because it does more and is prettier? anyhow, if you say that one thing is better why not back it up with features the other doesn't have
world was created 5 seconds before this post as it is.
MythTV currently relies on libavcodec on the backend to do video compression/decompression. The libavcodec library implements the various MPEG compression algorithms, which are *very* vigerously protected by the LA MPEG patent pool group.
Any commercial implementation of a DVR using MythTV would be at extreme risk of prosecution by the LA MPEG group for unauthorized usage of the MPEG patents.
It would be very nice to see MythTV transitioned to use the Theora (www.theora.org) video codec, as this is a patent-free video compression / decompression library.
But why not, say, hardware Theora encoding/decoding?
Because hardware that encodes/decodes Theora does not exist to the best of my knowledge. that, and MPEG 4 (and its varients) is widely supported by many systems/devices now (it's the video equivalent of mp3).
Why not software encoding/decoding, if it was just as fast?
Okay. I'm not fundamentally opposed to this, especially on the decoding side. A hardware encoder gives you the opportunity to use a much lower power (ie, no fans needed, lower power consumption) general purpose processor. This also generally brings the cost of the hardware down (which any embedded systems engineer like myself is obsessed with).
8) Upgradable
I thought harddrive upgrading was implicit in 6, but might as well make it explicit. In fact, sell it to me without a harddrive, just an image of the firmware on a CD. Further, since the firmware is entirely open, you can boot whatever you wish.
9) No reliance on proprietary/Windows stuff.
Absolutely.
And btw, how do you get the content of subscriptions, without the subscription?
Easy: you buy a subscription, but not from the hardware manufacturer. Instead of trying to make the money back on loss-leading hardware, the hardware people are out of the picture now. I can buy a subscription at a super-low rate from anyone who will sell it to me (competition), scrape it from a website, type it in myself. And when I stop paying my subscription, my device doesn't stop working.