Hardcoded md timestep - is it intenional?

[AnnaKravchenko]

In mdref and mdscoring, timestep is listed as a parameter, but in the code this value is overridden by the hardcoded "timest=0.002”.

Code snippet

In mdref.cns, line 24:

!==================================================================!
! Initialisation of variables
!==================================================================!

evaluate ($saprotocol.timestep=$timestep)  <------
evaluate ($saprotocol.fbeta=100)
evaluate ($saprotocol.mass=100)

And then the same file, line 357:

! Start MD refinement

do (mass =$Saprotocol.mass) (all)
do (mass=1000) (resn ANI or resn DAN or resn XAN)
do (fbeta = 0) (all)
do (fbeta = $Saprotocol.fbeta {1/ps} ) ( all )

! heat to 300 K
for $bath in (100 200 300) loop heat

    do (vx=maxwell($bath)) (all)
    do (vy=maxwell($bath)) (all)
    do (vz=maxwell($bath)) (all)

    dynamics cartesian
        nstep=$refine.heatsteps timest=0.002{ps}  <------
        temperature=$bath  tcoupling = true
        nprint=50
    end

Tested on examples/refine-complex/refine-complex-test.cfg, runs with default timestep and timestep = 0.0005 produce identical results, and the override can also be seen in the verbose cns log.

If this is intentional, then need to remove timestep from editable parameters.
If this is not intentional, then need to remove the hardcoded lines both in mdref and mdscoring.