To play a sound file as a question, then the subject is asked to read
the sentences to answer the questions.

Materials:
==========

(1) over 100 sound files (*.wav), named as q1, q2, ..., which are the
questions.

(2) one text files (*.txt) containing the sentences, s1, s2, .....,
which are the answer sentences.
	FORMAT

The question sound files and the sentences are linked by using the same
number.

Procedure:
==========

(1) show "ready?", then wait the subject to click "Y" to run the experiment

(2) show the sentence on the screen (It is important that the whole
sentence fit in one line.)  One of the sentences is shown all the time.

(3) two buttons: "(Q)question" and "(N)ext" 

"question" is to play the question sound file, 
which should be allowed to click again as many times as they like 

	if question sound already played
		enable "next" is go to next sentence, 

	if subject clicks next,
		go on to next sentence


(4) when all the sentences are finished, show "thank you".

Requirement:
============

(1) the sentences should be randomized.

(2) to give an output file of the orders of the sentences.
	FORMAT?

More ideas
==========
In addition to starting experiment with Y,
subject also types their name.
output (log) file records the name, and time of day, etc.

Make sure log file is never overwritten.

Implementation
==============
program name
	QA Paradigm

File locations
	sound files?
		In same directory as the input text file.
	output files?
		suggests subdirectory "log" of same directory

Information to be set by the operator
	input file
	font and font size

	Subject's Name
	Age
	Sex
	Languages/Dialect spoke
	do they study linguistics?

Robustness
==========
Should not fail during an experiment, due to missing file etc.

	So run through list of sentence numbers on opening, 
	and check all the cooresponding sound files are there.
	If not, raise a dialog.

Fonts supporting Chinese
	how to choose/find fonts on different systems?


