On Wednesday 25 April 2007 19:20, Antonio-M. Corbi Bellot wrote: > Jack Malmostoso <[EMAIL PROTECTED]> writes: > > On Wed, 25 Apr 2007 16:30:27 +0200, Antonio-M. Corbi Bellot wrote: > >> As other people in this list I had problems (well I couldn't obtain a > >> dhcp ip from my router) with 2.6.20 (2.6.18 works flawlessly in this > >> aspect) > > > > One difference I see between .18 and .20 is in the signal strength. > > With .18 all networks are at 100% of signal, while with .20 they have > > different values.
.18.* is reporting much more accurate values than .2*.* kernels. I KNOW I got 100% on local wifi and I also know the other values I usually get on other networks. I to noticed a drop over 50% of my strenght... I can't even get a reading on my router from my bathroom with the latter. > > Didn't notice this but I remember having read something about it, I > mean it was a known 'bug' and some patch corrected it's behaviour > making bcm43xx report the exact signal level instead of always 100%. 17+ kernels has not always shown 100%. It is very close to what osx is showing me. > > > Could it be that .20 is actually more "honest" in evaluating the signal > > strength? Or is it having "problems" in connecting also because it sees > > less signal? Not a chance in hell that it's honest, it's a rather nasty affair witch left me hardboot more than once. > > Uhmmm, don't know. As for me, the laptop is at the same distance, > position etc... from the router with 2.6.18 and 2.6.20, so this is not > the problem. > > A. Corbi. -- --- Børge http://www.arivene.net ---
