Legatus wrote:
Rick Funderburg wrote:
Tracy R Reed wrote:
Sounds like this could make wireless on Linux much easier to use and
make Linux compatible with much more wireless hardware. I sure hope
so because it seems we have all had our fights with wireless on Linux
over the past few years.
http://www.linuxdevices.com/news/NS1977847793.html
Sounds interesting, but it is hard to see from the article exactly
what their wifi stack is. It seems to be a driver framework that
includes software implementations of many wifi features.
-- Rick
From the reading I did, I think it is a framework with the protocol
drivers built in. Now the only thing that has to be written for a card,
will be the hardware drivers, and the API they need to be written for is
clear.
Which is pretty much the current problem. Some (many?) of the WiFi
chipset makers either don't make access to their proprietary hardware
drivers easy, improperly implement existing community standards, or just
make up their own "standards". Right now there are so many unknowns
about various chipsets that building a workable framework which "just
works" in Linux is a herculean task. In fact, it's really not much
better in Windows. For some things (WPA/WPA2), Linux operability is better.
And few chipset makers - Intel probably being the most cooperative and
the most prolific in terms of contributions - provide much support to
FOSS coders.
The only way I see this as helping is that it may provide a single API
and/or capability list that all chipset makers can use as a reference.
This will certainly make it easier for Linux programmers, but only if
the chipset makers cooperate. But I wonder if when WiFi chipset makers
can't all agree today, why should I expect them to cooperate tomorrow
just because someone else is baiting the hook?
--
Best Regards,
~DJA.
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
