Legal status of Rockbox and MP3

[Gnostic]

...

[Llorean]

Additionally, patent holders declined to enforce license fees on open source decoders, allowing many free MP3 decoders to develop. Furthermore, while attempts have been made to discourage distribution of encoder binaries, Thomson has stated that individuals using free MP3 encoders are not required to pay fees. Thus while patent fees have been an issue for companies attempting to use MP3, they have not meaningfully impacted users, allowing the format to grow in popularity.

[saratoga]

Furthermore, licenses for hardware devices are typically per device, so in theory Thompson and Co have already collected their fee when you pay for the hardware.

Though I'm very glad they're accepting of open source development too since they could certainly make our lives more difficult if they wanted to.

[Llorean]

I'm not sure if that applies to the software codec players. It really depends on whether they were licensed as MP3 playing devices, or devices that run MP3 playing software.

[saratoga]

I believe that any device thats not running on a PC has to be licensed as hardware, which has a different pricing structure.  I think the software platform thing doesn't really matter since all MP3 decoders are "software"; the MP3 spec is impractical to implement as a state machine or hardware pipeline because of the many different features in it (stereo modes, block sizes, frame sizes, etc).  I believe the determining factor is how open the platform is (PC verses DAP/DVD player/whatever.  

This is from memory though, its been ages since I looked at the license.

[Llorean]

I couldn't ever get what was defined as a "hardware" device and what was "software" from the license. My personal assumption had always been if there was a dedicated decoder (like the MAS in the Archos targets) it'd be hardware, but if the MP3s were decoded on a multipurpose microprocessor, like they are on the iRivers and so on, that it'd probably be "software". This was more from my personal definitions of Hardware and Software than any evidence elsewhere.

[saratoga]

I couldn't ever get what was defined as a "hardware" device and what was "software" from the license. My personal assumption had always been if there was a dedicated decoder (like the MAS in the Archos targets) it'd be hardware, but if the MP3s were decoded on a multipurpose microprocessor, like they are on the iRivers and so on, that it'd probably be "software". This was more from my personal definitions of Hardware and Software than any evidence elsewhere.

AFAIK the MAS is a general purpose microprocessor (or at least DSP) though.  It's called a hardware platform by Rockbox because we never figured out how to program it.  However, if we had purchased the compiler, we could load any codec we wanted to on there (well provided we could fit it in the very limited memory space available).  From what I gather, its a lot like the DSP built into the Ipod Video in the sense that it's programmable but only if you fork out the money for the compiler and kit.

AFAIK a true hardware decoder (IE a device that decoded without the use of a software decoder inside) has not been made for MP3.  Though if one exists I'd love to hear about it as these things are very interstesting to me!