What about having a "Calibrate Autoscaling" mode. Where the user could select to run with autoscaling enabled, and would be directed to a calibration menu. The calibration would take place only while the user had the calibration window open. It could give the user a list of known adapters (generated by adapters used by the user, not all known adapters), it would give high and low values for received signal, and then the values for the autoscaling. The autoscale numbers could be edited by the user. I would also suggest that the autoscaling calibration not simply record the highest and lowest seen numbers, but rather would increment the highest known value by X or decrement the lowest by X (X might be 1 or 0.2 or something like that). This may be a simple way to watch out for RFI spikes that might occur. Continuosly high signals would increase the max pretty quickly, while erroneously large spikes would have little effect. There could be a clear option, or re-calibrate button to perform the calibration phase if anything seems to go wrong. And obviously there would always be the option to turn off autoscaling. John On 3/1/06, Robin L Darroch <robin@xxxxxxxxxxxxx> wrote: > >>That looks promising! Any idea what kinds of output various adapters give? > > > >That's just the point, supposedly, the adapter will tell you which > >of these formats its using. > > Sorry - wasn't quite clear in my question: I was more wondering > whether anyone has info on how many kinds of adapters give each kind > of output (so for example, if most adapters give type 3, then it's > most important for us to track down a type 3 to type 2 conversion > algorithm to provide dBm output). Of course, if many give type 2 > output, then we could implement dBm display for those adapters in > KisMAC almost immediately. > -- > > ------------------------------------------------------------------------- > Robin L. Darroch - PO Box 2715, South Hedland WA 6722 - +61 421 503 966 > robin@xxxxxxxxxxxxx - robin@xxxxxxxxxxx - robin@xxxxxxxxxxxxx > > >