On Mon, 2009-05-18 at 18:40 +0100, Stuart Ward wrote: > Isn't this just mesh networking without the auto configuration bits.
You mean the Windows thing? I don't think so... could be wrong, but when I first read about this, I thought it actually was radio timeslicing. Dan > -- Stuart Ward M +44 7782325143 > > > > 2009/5/18 Dan Williams <[email protected]>: > > On Mon, 2009-05-18 at 11:44 -0400, Darren Albers wrote: > >> On Mon, May 18, 2009 at 11:00 AM, John W. Linville > >> <[email protected]> wrote: > >> > On Mon, May 18, 2009 at 10:57:05AM -0400, Tom Sutherland wrote: > >> >> Ran across this article today: > >> >> > >> >> http://www.engadget.com/2009/05/18/microsofts-virtual-wifi-will-make-windows-7-wireless-adapters-d/ > >> >> > >> >> > >> >> >From the article... > >> >> "The tech lets one piece of WiFi hardware be represented in Windows as > >> >> two separate adapters, meaning you can connect to two hotspots > >> >> simultaneously if you like, or turn your virtual device into an access > >> >> point that others can connect to." > >> >> > >> >> In the long term, would something like that be relevant to > >> >> networkmanager? > >> > > >> > You have to make it work in the kernel first... > >> > > >> > John > >> > -- > >> > John W. Linville Someday the world will need a hero, and > >> > you > >> > [email protected] might be all we have. Be ready. > >> > >> Don't some wireless drivers support this already with multiple > >> subinterfaces that correspond to a VLAN? > > > > They dont' really correspond to a VLAN, they appear as separate > > "wiphy" (aka wireless phy) devices which can be configured separately. > > > > Note that by doing this "splitting", you're effectively time-sharing the > > adapter, and your transfer rate will go down by more than half, since > > the adapter has to be on one channel for 1/2 the time, and the other > > channel for the other 1/2 the time. It's actually more than 50% because > > there's channel switch time involved, and probably null-packet exchanges > > when jumping to a channel to notify the AP that you're coming back from > > sleep. > > > > Yes, it can work, and yes, some mac80211 drivers support this type of > > functionality, but something that *also* should be done is figuring out > > how to tell users that little Elves didn't magically replace the single > > radio in their wifi card with two separate radios, and about the > > inherent limitations thereof. > > > > You could of course do a STA + AP with the same card on the same > > channel, where for instance the STA was connected to an existing AP, and > > you were connection-sharing the traffic out over the your own > > just-created AP to friends or something. But of course there you'll > > also be increasing contention of the medium (since you're effectively > > broadcasting the same traffic on the same channel twice, once from > > friend -> you, then you -> real AP). There is no silver bullet here of > > course, unless you really do have two wifi cards in your machine. > > > > Dan > > > > > > _______________________________________________ > > NetworkManager-list mailing list > > [email protected] > > http://mail.gnome.org/mailman/listinfo/networkmanager-list > > > _______________________________________________ > NetworkManager-list mailing list > [email protected] > http://mail.gnome.org/mailman/listinfo/networkmanager-list _______________________________________________ NetworkManager-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/networkmanager-list
