On Fri, 28 Aug 2009 08:48:47 +0100
Alan Maguire <Alan.Maguire at Sun.COM> wrote:
> Michael Hunter wrote:
> >
> >
> >>>> ncu_phys.c
> >>>> 1759: Indeed! Need to resolve the XXX somehow...
> >>>>
> >>> If I figure it out I'll change it but I really think Alan needs to give
> >>> us some insight.
> >>>
> >> Okay...but this needs to be dealt with before we integrate.
> >>
> >
> > Agreed.
> >
> From experience, WiFi is very flaky wrt link
> state events. Often when we connect, just as
> the connect occurs, we get a link state down
> event, which of course throws a spanner in the
> works. Instead of using link state events, we monitor
> the link state using a wifi monitor thread when
> connected. We have to monitor signal strength
> anyway, so it's not much hassle to additionally
> monitor connection status.
Okay, I moved the check for WiFi to the top of the function and updated
the comment.
mph
>
> Alan
