On Tuesday 14 October 2008 01:01:26 pm Roberto A. Foglietta wrote: > 2008/10/14 Roberto A. Foglietta <[EMAIL PROTECTED]>: > >>>> Ok, making poll wait 300 miliseconds before deciding there's no next > >>>> character > >>>> in a pending escape sequence seems to have fixed it. (At least I can't > >>>> reproduce the problem under qemu anymore.) > >>> > >>> Please document this next time, or someone else might come later > >>> and delete the timeout. I did this a few mins ago :( will fix it now. > >>> > >>> Did you try something smaller than 300ms? > >> > >> As far as I understood the problem: considering a 1200bps line, 120 > >> chars per second, 40 escape sequence per seconds, then the minimum > >> timeout should be at least 1/40 sec = 25 ms. This is for a fixed speed > >> line, considering an asynchronous data line with an average speed of > >> 1200bps and a bell curve variance of 25 ms then using a timeout of > >> 100ms would catch the 0.999936657516% of the escape sequences. > >> Enlarging the timeout to 150ms 0.999999998027%. Over 200ms should not > >> make any sense any more if the variance has been correctly estimated > >> in 25 ms. > >> > >> http://en.wikipedia.org/wiki/Normal_distribution > >> > > > > Because the distribution for delays follows a poisson curve you > > should add 25ms to the timeout I previously estimated: 100 -> 125, 150 > > -> 175 and 200 -> 225. This because the after having wait for 25 ms > > (fixed speed line) the poisson curve could be fairly approximated with > > a gaussian curve. > > > > Pressing ESC each second the mean time of a text corruption event would be: > > 25+125 = 150 ms --> 2 days > > 25+150 = 175 ms --> 16 years > > I think 200 ms would be sufficient to cover the everybody expectation > life time.
In theory, theory and practice is the same... I'm curious what smallest number seems to work reliably for Rob. -- vda _______________________________________________ busybox mailing list [email protected] http://busybox.net/cgi-bin/mailman/listinfo/busybox
