On Sat, 20 Jan 2007 15:21:20 +0100 "Bin Zhang" <[EMAIL PROTECTED]> wrote:
> No problem on my ibook G4 using debian pbbuttons package. > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=389770 Thanks. I didn't remember this bug report. I think the backlight control itself is not the problem here. Since version 0.7.9 pbbuttonsd uses the sysfs interface too and depends no longer on the PMU ioctl interface (except IOC_GRAB_BACKLIGHT). But we have another side effect here. If you press the brightness up key, the kernel itself processes the key trigger and makes the display brighter. After that a key event is sent out and pbbuttonsd does the same again. You got a double brightness increase for the price of one. That sounds not very serious but if you use the display fading feature I would expect that the brightness change smoothly from one level to another after I hit the brightness up key. What happens here is that you could see the brightness jump up and then maybe the rest changes smoothly. The reason for that is that the kernel changes the brightness still hard in 15 steps (It doesn't know anything about fading) although the sysfs interface supports 127 steps (on my machine) and therefore the step size of the kernel must not be the same as of pbbuttonsd . The key problem here is if you can't safely disable old interfaces you have to keep them fully intact. Don't disable them partly. If debian doesn't add CONFIG_PMAC_BACKLIGHT_LEGACY people will recognise odd backlight behaviour on powerpc machines. I have no problem with that because I compile my own kernel and that's it. At least I hope Frank, pbbuttons' debian maintainer, would remove such bug reports instantly if someone blamed pbbuttons for that. Best Regards Matthias -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
