|
|
|
Forum Guru
 
Group: Forum Members
Last Login: 2 days ago @ 5:23:48 PM
Posts: 137,
Visits: 380
|
|
OK, now I have my own problem to post. I have an experiment where occasionally a slide object refuses to run. I have looked at the time audit, and find that .FinishTime = .StartTime when this happens. Also, .OffsetTime, .ActionTime, and .Offset time are all 0 when this happens. .TargetOnsetTime is still what it ought to be, and .Duration is a positive value.
Any idea what could make this happen? I have run out of ideas.
Thanks,
-- David McFarlane, Professional Faultfinder
|
|
|
|
|
Forum Guru
Group: Forum Members
Last Login: 2 days ago @ 5:23:48 PM
Posts: 137,
Visits: 380
|
|
I think I found a workaround. I put a dummy Wait object with a duration of 0 (and no vertical sync) just before the failing object, so the dummy Wait now absorbs the .FinishTime = .StartTime and the next object runs as it should.
That will let me get back to work on my experiment program, but this still looks like a bug in E-Prime to me, so I would still like PST to look at this.
-- David McFarlane, Professional Faultfinder
|
|
|
|
|
Forum Guru
Group: Forum Members
Last Login: 2 days ago @ 5:23:48 PM
Posts: 137,
Visits: 380
|
|
I have thought about this further, and now I am pretty sure that this is not a bug in E-Prime, EP is working as designed and as it ought to be designed. Here is what I think...
I was using an extended response period with End Action set to Terminate. The response period extended through a list that followed. The list procedure contained a display object, and an inline script that terminated the list if it saw a subject response. So far so good. But sometimes the response came during the brief moment between the offset of the list display object and the list termination test in the script. So apparently the input mask raised a termination signal that was not caught by any display object. So when the program got to the next display object on the other side of the list, that object would catch the termination signal and not run.
I still need to test this all to be sure, but the day is late.
In the meantime, here are solutions that I can think of:
1) Use a dummy display object to catch any lingering response termination signal.
2) Rearrange my list procedure so that the list display object will catch any response termination signal that remains after the list termination test.
3) Set the input mask End Action to (none).
-- David McFarlane, Professional Faultfinder
|
|
|
|
|
Forum Guru
Group: Forum Members
Last Login: 2 days ago @ 5:23:48 PM
Posts: 137,
Visits: 380
|
|
Now that I have done some tests, I am convinced that my analysis above is correct. Of course, I would be glad to have PST staff correct or confirm as needed.
-- David McFarlane, Professional Faultfinder
|
|
|
|
