error: fix vector and rerun

Dear all;
There is a terrible trouble when I post-processing my dump file with the rerun and fix vector commonds in a new input script to commpute diffusion coefficient via compute VACF.
The partial input script as follow:

timestep 0.25

compute vacf carbon vacf
fix vacf carbon vector 200 c_vacf[4]
variable diff equal dt*trap(f_vacf)

rerun dump.lammpstrj first 200 every 200 last 10000 dump x y z vx vy vz format native
Once input was excuted,then a error——cannot reset timestep with a time_dependent fix defined——came out.I didn’t reset timestep,it is always “timestep 0.25” whatever it in my dump file or in my post-processing input.when dumping my “dump.file”,Nevery is 200,so I continue using 200 in fix vector. The “fix vector” is the only fix used in input.I’m trying everying i can,such as I change Nevery 200 of fix vector to 1 and check the lammps/mailing lis for related theme,but all what I do is useless. But I found that when I exchange “rerun” for “run” command,everying is ok.I don’t know how it happens,is there some bug for “fix vector” command or some trouble in my post-processing input ? Can someone give me some advice or offer me a similarity input which can correctly “rerun” ? Expect your response !

PS: The version used is Aug2015,the latest.

Best
regards !
Louis-Kuan

Reset timestep does not refer to the “timestep” command

which sets the “size” of the timestep. Reset timestep

refers to the counter for the timestep (what the reset_timestep

command would set). Some fixes, including fix vector,

rely on timesteps being continuous (1,2,3,etc) so

they can trigger their operations correctly. The rerun

command violates that by reading in snapshots at

new timesteps, and resetting the timestep to that

value. Hence the error. We’d have to think if

there is someway to allow fix vector to be used

with rerun, since there is no guarantee the timestamps

on the snapshots you read in will be compatible with

the Nevery in fix vector.

Steve