Millisecond Forums

Yes, selection at the <block> level works according to the proportions given. See the examples in the documentation for the /trials attribute for details. One short example here (many more can be found by searching the forum):

/ trials = [1-10 = noreplace(a,a,b,b,b)]

will result in 4 'a' trials and 6 'b' trials in random order.

Another option is to use a <list> element for trial selection in the respective <block>s' /trials attribute. The desired trial-ratio can be specified via the <list>'s /itemprobabilities.

I am having a 100ms gap between the instructions and the beginning of a blank screen.

The instructions end when 0 key is pressed.
I need it to be instant.

Looking through my code I do not see what the cause could be. It's just a random gap of ~100ms between marker data.

Does Inquisit put an automatic 50ms gap between stimulus???

Sorry, I'm not following. (a) What's the relation to the trial-ratio question in this thread and (b) how would anyone know what code you are referring to (I don't see any of that posted)? Can you clarify? Otherwise I can't help here.

No. But you may be mistaken about how trial timings work / the interplay between the various components. For that, see the "How to Control Trial Duration and Inter-Trial Intervals" topic in the docs.

Here is the trials and blocks:

<trial instructions>
/ stimulustimes = [1=instructions]
/ validresponse = ("0")
/ errormessage = false
/ recorddata = false
/ responsemessage = ("0", tPort, 50)
</trial>

<trial break>
/ stimulustimes    = [1=blank, bPort; 50=zeroPort]
/ trialduration = values.break
/ errormessage = false
/ recorddata = false
</trial>

<trial warnsound>
/ stimulustimes = [0=warnsound]
/ trialduration = 500
/ errormessage = false
/ recorddata = false
</trial>

<block instructions>
/ screencolor = (0, 0, 0)
/ trials = [1=instructions]
</block>

<block warning>
/ onblockbegin = [values.break=200]
/ screencolor = (0, 0, 0)
/ trials = [1=break; 2=warnsound; 3=break; 4=warnsound; 5=break; 6=warnsound]
</block>

When I run this there is a gap of 100ms between Instructions and the first break in warning block.

And why is that a problem? Also, why are you running the instructions trial in a separate block? Note that there'll be at least a single frame between <block>s for cleanup.

Whoops. Put it in there. But there is still a delay of about ~88ms from beginning of instruction marker to the beginning of the break marker.