Best way to do free-form realtime analysis?

[Stefan Schulz <prog1@xxxxxxxxxxxxxx>]

Jack Griffin wrote:
>In the event (I should say "unlikely event", if my previous trading skills 
>are any guidance) that I ever devise a really profitable system I would 
>like to program my computer to place trades for me and head off to the beach.

Don't.  Let it tell you when to enter and exit by all means but you still 
need to do a sanity check.  That's not because you should second-guess a 
working system - you should not - but because faulty data can sometimes 
give a signal when reality would not have done so.  If, say, you see a $10 
stock move $100 in a minute and back again you know that it's a data 
glitch.  An automatic system will not.

>So far, it seems that the best way to have the sort of "direct access" to 
>realtime data would indeed be with the Excel spreadsheet program. Is this 
>thinking reasonable?

Absolutely.  Its drawback is that you have to program everything from first 
principles, which is time consuming, but no canned market data app can 
touch it in terms of versatility.  No prizes for guessing what I use ;-)

>Am I totally missing anything?

Possibly, but I'll tread carefully here:  Software is frozen thought, and 
it's the thought that counts.  Choosing package A over package B is a only 
matter of comparing features and benefits.  My system development happens 
with coloured marker pens on *printouts* of data or graphs.  Only when I 
have a working idea developed this way do I code it and throw it at more 
markets for checking.  A simple $100 colour printer is the best development 
tool I ever bought.

Regards,

Stefan Schulz
Suaviter Limited
prog1@xxxxxxxxxxxxxx