Bluecherry Releases GPL'd MPEG-4 Driver 45
azop writes "Today Bluecherry released a GPL'd driver for its multiple-input MPEG-4 hardware compression cards. The driver supports audio and video capture from 4-, 8-, and 16-channel single-card encoders using the Video4Linux and ALSA APIs. More information about the driver and its features can be found on Bluecherry's development blog and on Ben Collins' personal blog. Bluecherry is the first Linux software company to release a complete driver based on Linux kernel APIs (Video4Linux and ALSA) for multiple-input hardware-compressed MPEG-4 capture cards under the GPL. The cards are designed for security applications (digital video recording), but other applications could potentially make use of the compressed streams and Video4Linux API integration. An H.264 version is 'in the works.'"
Re:Patent pools! (Score:5, Informative)
First? No. (Score:2, Informative)
It wasn't a "multiple input" device, but many years ago Plextor released a GPL driver for their go7007 based video capture devices, which captured directly to MPEG-4. Unfortunately few people bought them, so Plextor stopped working on the driver and it has since disappeared from the kernel, even though it was in the staging branch for a while.
Re:First? No. (Score:1, Informative)
The go7007 driver is still in staging in the v4l-dvb tree. It still needs quite a bit of work to get out of staging however.
Good company (Score:5, Informative)
Re:First? No. (Score:3, Informative)
I don't think the Hauppauge 1212 (the HD-PVR) is a comparable product. The 1212 takes in a single analog video up to component 1080i and produces a x264 stream from it. This card has multiple inputs (4,8, or 16) that are D1 (720x480) max. Utilizing 16 inputs it does not support 30 frames/sec that NTSC video uses (7.5 fps max @ 16 inputs). This is aimed at the digital security market. The 1212 is aimed at the HTPC market.
Cheers,
the_crowbar
Re:Patent pools! (Score:5, Informative)
Patents forbid the use of any patented device, not just the sale or manufacture. There's fortunately something called the Exhaustion Doctrine, that says as soon as a patent holder has sold a device in an unrestricted manner on the open market, their patent rights have been exhausted. So, if you buy a MPEG-4 encoder card without any restrictions on that purchase, all patent rights to it are supposed to be done. However, you'll have to go to court to prove that.
Additionally, it's not very surefire of a defense at all, due to the fact that the courts are at best a crapshoot, and at worst a popularity contest. LG licensed some patents to Intel, with the condition that Intel inform all buyers of Intel chips that the LG patents are only licensed on the condition that you only use the chip with Intel hardware. In defiance of this notification, a computer reseller violated patent law by selling systems built with Intel chips, but containing non-Intel brand peripherals. Though the computer retailer lost, it eventually reached the Supreme Court, where it was ruled unanimously against LG. This is contrary to all previous lower court rulings, that held that just a note on the box constitutes a "restriction" and so the patents remain in full force. The Supreme Court said that since the reseller didn't need a patent, the note from Intel that they don't get one doesn't really matter, and doesn't constitute an implicit contract.
Re:Kernel update breaking drivers? (Score:3, Informative)
>> Bluecherry is the first Linux software company to release a complete driver based on Linux kernel APIs (Video4Linux and ALSA)
>
> Are these some sort of stable APIs, or are they the driver APIs that are randomly changed every few kernel versions to break binary compatibility?
The only people that seem to complain about that are sandbagging companies that need to distract from the fact that the community is doing better.
V4L dates back to the original bt848 carts.
Re:First? No. (Score:1, Informative)
an "interesting" product won't necessarily fade away for lack of interest
That's not necessarily true, unfortunately. While I was 100% happy with my Plextor m402u (which used the go7007 driver) on an older kernel, I eventually had to upgrade my mythtv backend and the newer "staging" kernel driver is a P.O.S. So now I have switched to the Hauppauge HDPVR in order to get back to having actively maintained well written drivers. In other words, even though the driver exists and my hardware still works perfectly, the lack of maintenance on the driver means the device itself has become useless under Linux.
Re:Kernel update breaking drivers? (Score:3, Informative)
The point of the story is that they release their driver source code, under the GPL.
So you dont care about binary compatibility, you just recompile the driver for every version of the kernel you need....
And you can fix the source code incompatibilities yourself if they ever happen and they dont keep up.