Re: Bug#391686: ITP: ipw3945-daemon -- Binary userspace regulatory daemon for Intel PRO/Wireless 3945ABG cards
On 10/09/06 12:11:28AM +0100, Stephen Gran wrote:
> This one time, at band camp, Jim Crilly said:
> > Most people are willing to deal with firmware since they don't run on the
> > host CPU but only on the card that they're controlling. Intel's daemon
> > isn't as bad as the Atheros HAL or nVidia's blob, but it's still non-free
> > and non-necessary. OpenBSD developers wrote a new driver that doesn't require
> > the regulatory daemon in ~2 months after the initial Intel release. Given that
> > and the fact that there's already the Intel GPL'd driver available wouldn't
> > it make sense for someone to 'fix' the GPL'd driver to not require the daemon?
> My understanding of the problem is that this 'fix' means roughly 'remove
> half the functionaliy'. I'm not saying it's the wrong answer, just
> pointing out that it's not quite that straight forward.
I don't know about half, the only extra functionality in the daemon
mentioned in the KernelTrap article is automatic tuning of the radio based
on temperature and signal strength. So while there would be reduced
functionality it would probably go unnoticed by the majority of users and
could be reimplemented in the GPL'd driver or as an optional free daemon
if someone wanted to do that.