On Tue, 17 Jul 2001, T. V. Raman wrote:
> One of the little ones to fix that might be big for the
> ordinary user is the clock problem.
> It's now reduced to being an irritant for those of us
> struggling with bigger issues with intimate and familiar --
> but it's likely to be a show stoper for someone who boots
> familiar and compares it to wince and says "this thing cant
> even keep time"
As I understand it it does keep time. It's just a matter of putting this
little utility called hwclock into your root fs. When you set the date
just do
hwclock --systohc
(and perhaps with --utc)
and when you boot do
hwclock --hctosys
to set the system time from the hw clock.
Is there any reason why this is not the way it is done right now?
The only thing is that the hw clock will lose the time when you reset the
ipaq, which still probably happens often enough.
Also the alarm issue on the TODO list: as I understand it this should be
working now, as per the instructions in the kernel source tree under
Documentation/rtc.txt. An application should be able to set an alarm
and the ipaq would wake up from sleep at that time. Is anything missing,
except the app itself of course?
Something I noticed: the touch screen driver does not check for
pending signals during a blocking read, and thus programs waiting for
touch screen data or buttons cannot be killed when waiting on a blocking
read. Of course using select/poll/fasync avoids this problem but still...
I read in the WIKI that the power and record buttons have their own GPIOs
and it should be possible to get press/release events of combinations of
those and other buttons (and the keypad center button as well). I'm not
too clear on how this can be accomplished from userspace however. If the
mechanism is missing, this issue might be added to the list. If I'm
mistaken then please someone explain how to do it... from what I read, the
record key is handled just like the other plain buttons and is not
independent.
Also is there anyway to get a signal to userspace when the ipaq is put
into a sleeve? There's a GPIO for that, but is there anyway an application
can notice and tell the user?
-- Stéphane Doyon <s.doyon@videotron.ca> http://pages.infinit.net/sdoyon/Received on Tue Jul 17 10:29:00 2001
This archive was generated by hypermail 2.1.8 : Tue May 04 2004 - 09:38:27 EDT