labscript waits: Do not save AIN values during wait (then we don't have a problem for sure if the wait takes very long).

Added parameters to pause and resume both the main
process and the TiCo processes. Moved the time increment
in the main process to earlier position, for better sync
when TiCos are resumed. Still need to test on hardware.

PID and timing improvements

See merge request !2

Moved calculating new output values to after setting AOUT modules. Write to DAC at very end of EVENT. Set PID gains and limits only once during init, because that took too much time during the cycle.

Removed remiaing code from AIN storage rate from ADwin code (we don't use that anymore because of higher workoad)

Revert "When PID is enabled, don't use the I value from previous one, that seems to enhance overshoots (added that in April, not sure anymore why)"

This reverts commit 9535cf1a.

When PID is enabled, don't use the I value from previous one, that seems to enhance overshoots (added that in April, not sure anymore why)

Removed resetting pid_sum when PID is turned on (at beginning they are initialized as 0 already, and when we update the PID parameters or turn it off and on again, this does not reset the last values).

Set DIO channels as outputs in manual mode processes (otherwise DO channels do not work after ADwin poweroff in manual mode until after first shot).

Added sorting of AIN array

See merge request !1

Moved AIN commands out of "IF (Par_11=1)" block, such that the AIN values
can be read in manual mode after finishing a shot without having any values for the ourput set.

Fixed workload issue with two addition time variables, set DIO output at t=0, use for loop for AOut channels (PIDs) --> Loop might be removed again!