Mikkel L. Ellertson wrote: >>> This is the explanation put forward by the binary-only wireless network >>> companies, that regulatory requirements mean there can be no OSS >>> wireless drivers because it would facilitate using the PLLs on the chip >>> outside of the allowed frequencies for the countries it is sold in. >> >> Hmm. Actually, this makes sense. It would be a lot cheaper this way >> than actually to ship different hardware. And the regulations make >> it clear that it cannot be something someone could easily figure out >> and change as a user. >> > It would make since if they were supplying different drivers > depending on where they are selling the device. But what I have > seen is that the installer asks what country (drop-down menu) > and this determines what channels the driver uses. So all you have > to do is pick a different country... So a closed-source driver > does not insure compliance... Yes it's nonsense anyway since binaries are only obfuscated code, not code that can't be turned back to source, as we see in the reversed Broadcom driver project(s). http://bcm43xx.berlios.de/ In addition the registers are sitting there ready to be twiddled by any kernel code anyway. The only real way to stop the oscillators going to "bad" frequencies is to disallow them in hardware and this isn't what we see. And some people like Atmel created OSS drivers for their Wireless chipsets making the whole excuse hollow! http://atmelwlandriver.sourceforge.net/news.html -Andy
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature