Re: please release sarge instead of removing binary firmware
>>>>> "Evan" == Evan Prodromou <firstname.lastname@example.org> writes:
>>>>> "AW" == Allan Wind <email@example.com> writes:
AW> If the above sounds about right, then perhaps someone
AW> representing Debian could contact the relevant vendors to make
AW> them aware of the problem and request a resolution that is
AW> acceptable to Debian?
Evan> What resolution _is_ acceptable to Debian, anyways? Assembly
Evan> language code for some weird chip's instruction set? Do we
Evan> even have assemblers for that code if we get it? Are we
Evan> going to have to write them? Are all kernel images going to
Evan> have to build-depend on them?
Evan> Why are we deciding that the source code released by the
Evan> copyright holder -- usually hex-encoded binary in C -- isn't
Evan> the preferred source format? Can't we just get some kind of
Evan> assurance from the copyright holder that, yes, that gigantic
Evan> C array is their preferred source version?
If such assurance is true, then yes. I suspect there is probably at
least one chip out there where people really do just do bit twiddling
on binary images to change the firmware. ANd if such a binary blob is
placed under the GPL by its copyright holder and then placed in the
Linux kernel, then that code is free and GPL compatible.
But there are a lot more chips where that is not the preferred form of
modification. And it doesn't matter so much what the copyright holder
says as what they actually do when they want to change their firmware.