Wireless: the new barrier for desktop Linux

[Darrin Chandler]

On Wed, Oct 10, 2007 at 01:39:58AM -0400, fouldragon at aol.com wrote:
> Now, linux is going to falter as long as wireless is still a mess.

> In the time I spent fighting, I could probably have drilled a hole 
> through my floor, fed some Cat 5 through down to the router, and had a 
> supported wired network.

> I got wireless working once, in 2003, on my little PII-266 Thinkpad, 
> but that was in the days of open 802.11b networks and only a handful of 
> chipsets.

It's not easy at all, unfortunately. I'm sure there are some people who
are thinking, "gee, wifi has always worked for me on Linux," but that's
the luck of the draw having well supported hardware.

This isn't because the Linux developers are doing a bad job writing
code, it's because wireless chipset vendors are very tight with their
specifications. Some people (Greg Kroah-Hartman) think it's cool to sign
NDAs to get this info, but that means that only that one developer knows
what's going on. Doesn't sound like the open source / free software
model to me.

The path to real support is to get the chip vendors to release
documentation without NDA. OpenBSD has been working on this for years
and has had some success, with some vendors. OpenBSD has also reverse
engineered support for some chipsets. In both cases the code is
available, and well documented. There was recently a licensing issue
with Linux using OpenBSD code for Atheros chipsets, but hopefully that's
all been resolved and the Linux wireless guys can take what they need
from OpenBSD. Even so, that'll only fill in *some* pieces of the puzzle.

Developers are out there pushing vendors for open specs. As consumers we
can help by only buying cards where the vendors regard *us* as their
customers, instead of thinking their customer is MS, Apple, or whoever.
I want my hardware to work in *my* choice of operating systems, and the
only way for that to work is open specs.

FWIW, open specs from chip makers USED TO be the norm, and was that way
for a looong time. It's only in the last several years that they've
started keeping specs secret. I'll leave it to you to imagine why that
change occurred, and who's interests are served by it, and how
individual developers signing NDAs plays into that scheme, and how using
ndis_wrapper as a consumer helps the vendors justify it...

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