Millisecond Forums

Hello!
I am creating an experiment using a dual-stream version of RSVP paradigm. The task is to identify the T1 and T2 targets that are preceded and followed by distractors. 
The script for the distractors looks like this: 

<item DYS>
/1="Q"/2="W"/3="E" /4="R" /5="T" /6="Y" /7="U"
/8="I" /9="O" /10="P" /11="A" /12="S" /13="H"
/14="Z" /15="X" /16="C" /17="V" /18="B" /19="N" /20="M"
</item>

<text DYSL>
/ items = DYS
/ position = (25,50)
/ fontstyle = ("Helvetica", 6.38%, true, false, false, false, 5, 1)
/ select = noreplace

</text>
<text DYSR>
/ items = DYS
/ position = (75,50)
/ fontstyle = ("Helvetica", 6.38%, true, false, false, false, 5, 1)
/ select = noreplace
</text>

And the script for the first trial looks like that:
<trial LAG1T1LT2L>
/ stimulustimes = [1= FIX, DYSL, DYSR;
130=FIX, DYSL, DYSR;
259=BLANK, BLANKL, BLANKR;
260=FIX, DYSL, DYSR;
289=BLANK, BLANKL, BLANKR;
390=FIX, DYSL, DYSR;
519=BLANK, BLANKL, BLANKR;
520=FIX, DYSL, DYSR;
649=BLANK, BLANKL, BLANKR;
650=FIX, DYSL, DYSR;
749=BLANK, BLANKL, BLANKR;
750=FIX, T1L, DYSR;
879=BLANK, BLANKL, BLANKR;
880=FIX, T2L, DYSR;
1009=BLANK, BLANKL, BLANKR;
1010=FIX, DYSL, DYSR;
1139=BLANK, BLANKL, BLANKR;
1140=FIX, DYSL, DYSR]

/ response = noresponse
</trial>
My problem is that when I run the script the first pair of distractors are being repeated for the whole trial, and I would want to get a random selection of items from the DYS directory on each stimulstimes involving distractors. How can I fix this problem? 

Any help would be appreciated!

Thanks,
Szymon.