CMoH
commented
9 years ago From email with @guillaumecl:
On my G13 I have warnings that the initialization does not work, the device still works though (and it powers up much faster). It looks like the G13 never sends the messages that are supposed to be received in raw_input.
I have tried to remove the wait condition, and the device still powers up fine, everything works as expected. I've noticed this initialization sequence is also found on other devices, does it give warnings on those too ?
The computer I am using has both a Linux installation and a Windows install. I have noticed that my G19 (my main keyboard) behaves differently from a cold boot to a reboot from Linux or a reboot from Windows. Also, the G19 is externally powered, and power-cycling the keyboard seems to have effects on its state after boot.
Perhaps it's worth seeing if your conclusions hold in each of these situations. If they do, then we can remove (or deactivate) that code - see if anyone complains :)
Thanks for the observations. When I had removed the initialization code, it worked from cold boot, resume after suspend to ram, plug out the cable then plug back again.
What I never tested was to boot Windows then reboot to Linux since my Windows setup is on another older machine. I'll test that too just in case.
Some "faking away with initialization" message seems to appear on all devices. The initialization sequence seems to be reverse engineered with an USB monitor from the windows drivers until the device "worked". The stages and the messages exchanged at that phase seem to differ for each device, although the results are ignored by the driver.