Kernel 2.6.2 Changes

[AthlonRob]

On Thu, 2004-02-05 at 14:37, Bradley Chapman wrote:
> > Of course, CVS isn't quite perfect.  I now have a wifi0 that's just
> > funky and a wlan0 that works as normal and no working WEP support, but
> > some of that may be my specific card and some of it may be my
> > configuration.
> 
> I get a strange wifi0 interface under 2.6.0, 2.6.1 and 2.6.2. I think it's
> a weird bug in the CVS code.

I have a feeling there's a reason behind it, I'm just not sure what it
is.  Maybe they're planning on changing the name from wlan0 to wifi0?  I
like the idea of the interface being named wifi better than it being
named wlan, I think.
 
> > I'm using hostap as a regular driver as the orinoco_cs and wlan-ng
> > drivers wouldn't either work quite right with my card.
> 
> So am I - I kept getting interrupt storms with the orinoco_cs drivers (for
> which development is somewhat calcified, BTW). And the linux-wlan-ng drivers
> are too hard and too complex to set up (I fought with them for months before
> going with orinoco_cs, and now hostap).

I never had any interrupt issues with the orinoco_cs drivers (although I
did with my Atmel card and the SF drivers - but that is another
story)... the drivers really worked fine, except they never could locate
a base station.  They recognized the card, the module loaded, wireless
extensions, etc.  It even seemed to show signal strengths, it just never
actually associated itself with the base station.  So... I moved on to
linux-wlan-ng.  Why the HELL would they refuse to use Wireless
Extensions, which are *SO FREAKING EASY*?  After sitting on their
mailing list for a while, I think I've come to the conclusion the
wlan-ng dev(s) just got pissed off about wireless extensions becoming
more popular than them, so they say piss off to anybody who reasons
wireless extension support in their drivers might be a good thing.

</rant>

Rob