[Coco] Decypher this code!

[Robert Gault]

J.P. Samson wrote:
> On Oct 1, 2008, at 6:16 PM, Robert Gault wrote:
>> It would help to see the DATA statements but without them, my best 
>> guess would be:
>>
>> 95 CLS(0):FORX=1TO26STEP2:T=(X-1)*16:T2=(X-1)*16+16:PRINT at T,P$(X,1);:
>> PRINT at T2,P$(X,2);:NEXTX:END
> 
> Why use (X-1)*16+16 when you could use the mathematically equivalent 
> X*16.  But who knows if the programmer bothered to optimize!
> 
> -- JP
> 

Tim asked to translate the scan. There is no way the scan could be X*16, 
so even though that would be better than my proposed translation, it is 
not relevant.

However, the choice for the character between the two 16s is relevant 
and significant. So what do you think it is?