On Wed, May 05, 2010 at 11:32:47AM -0400, Kent Watsen wrote:
quoted text
> There is a discussion on the osol-discuss mailing list this morning where
> it's pointed out that OpenBSD source tree has a blob in it:
> 
> http://osdir.com/ml/opensolaris-discuss/2010-05/msg00095.html
> 
> The location of the blob in the tree is here:
> 
> http://www.openbsd.org/cgi-bin/cvsweb/src/sys/dev/microcode/afb/microcode.h?rev=1.1;co...

From the linked message:

"They're not happy about it, but a firmware blob isn't really any
different if it's loaded from disk where the OS can see it vs. baked
into a PROM on the card where the OS is unaware of it."

There it is in a nutshell. Do you only allow hardware with fixed
microcode, or do you allow the kernel to load it from a blob? What are
the practical and ideological differences?

Practical: microcode can be updated. Otherwise it's the same as burned
in microcode.

Ideological: non-free microcode is not free, whether it's burned in or
uploaded by the OS kernel. Take your gNewSense no-blob Linux distro and
put it on any computer you can find and you'll be running with non-free
microcode, starting with the CPU.

I don't pretend to know all the issues involved here, but I think what
I've said above is accurate as far as it goes. You can't turn on a
mainstream computer that is free of closed microcode: such computers
just don't exist.

-- 
Darrin Chandler            |  Phoenix BSD User Group  |  MetaBUG
dwchandler@stilyagin.com   |  http://phxbug.org/      |  http://metabug.org/
http://www.stilyagin.com/  |  Daemons in the Desert   |  Global BUG Federation