Quote:
Originally Posted by baskerville
Please note that Plato relies on /tmp/nickel-hardware-status for knowing when the USB cable is plugged. KSM changes /usr/local/Kobo/udev/plug and prevents it from writing to the aforementioned FIFO. I'll suggest a different approach to tshering when he returns.
|
I think I will simply replace
Code:
[ "$(pidof nickel | wc -w)" == "0" ] && exit
by
Code:
[ "$(pidof nickel | wc -w)" == "0" ] || [ "$(pidof plato | wc -w)" == "0" ] && exit
in plug. What do you think?
Should I tread plato the same way as nickel in /usr/local/Kobo/udev/sd too?
Quote:
Originally Posted by baskerville
I forgot to mention that the wifi-{enable,disable}.sh scripts won't work when Plato is launched via KSM: the required environment variables are defined late in rcS and KSM starts early.
|
Actually these variables are available when KSM is running. Your script does work, when on inserts " sleep 2" after
Code:
insmod "$WIFI_MODULE_PATH"
As for plato.sh, if you run "./nickel.sh &" only if nickel was running before you launched plato, KSM could simply call this plato.sh. Then I would not need to change my calling script each time you change plato.sh.