Re: Solving the linux-2.6 firmware issue
On 01/09/07 11:50, dann frazier wrote:
> On Mon, Jan 08, 2007 at 07:49:19PM -0800, Jeff Carr wrote:
>> I've not seen conclusive evidence that the keyspan firmware file is
>> not the best effort of freeness.
> "This firmware may not be modified and may only be used with
> Keyspan hardware."
> That cannot be considered a best effort of freeness.
Well, that's a nice sound byte, but I think it's not so easy.
I hope you would agree that it's the best effort of the kernel
developers! They are certainly working as hard as possible to support
Lets focus on keyspan. For one, it looks like Keyspan might have been
(and might still be) trying to make progress. These firmware files are
old. As the years went on, it seems like at some point someone (from
within keyspan) figured out how to describe some of firmware
functionality as code. So it looks like there was some forward
progress. It's still likely that the raw hex values are generated by
some hardware guys and there isn't any understanding of what they
really mean -- or what they mean makes no sense or can't be described
in C. This happens all the time. The hardware guys could have quit,
been fired or died and the knowledge lost.
What doesn't make sense to me is to throw out stuff like this because
we don't have the code. Lots of other hardware like this has this same
non-free stuff on chips inside of it that we can't see the binary. I
don't see any way around problems like this. Do you have any
suggestions here? The problems are very complicated to solve. What is
best for the free software movement going forward in your opinion?