From: Rob Sacks (editor_at_realization.org)
Date: 2001-05-20 21:56:47
Hi Waldemar,
> I suggest using right now not final software yet ,
> but instead diagrams
> and pseudocodes and language C.
I would suggest that we start by talking about our
design goals. I've been assuming all along that we
are basically writing a biofeedback program with
a flexible threshholding (protocol) interface but
perhaps we disagree.
The program should be designed to be extremely
extensible and flexible. As time goes on, new machines
and interfaces should be easily added.
I'd like to suggest the following goal: let's make it
possible to add modules to the finished program
without changing existing source code. In other
words, let's maximize the amount of functionality
that can be added as a plug-in.
Let's also maximize dynamic programmability.
IN other words, as much as possible, things
should be changeable at runtime by the user
or by add-on modules.
I also think we should support all EEG machines,
not just our own. And we should support other
kinds of input.
I'd also suggest that we maximize peformance.
We should look ahead to supporting 3D animation
in the user interface.
After we agree on the design goals, I'd suggest that
our second step should be to design the overall
program architecture. This includes:
1. Define the modules (which functionality goes
where).
2. Design the data structures.
3. Design the APIs through which modules
communicate with each other.
I think pseudocode is mainly useful at the stage
when a programmer is implementing a module.
I'm not sure we need to worry about that now.
I agree about C. Actually, I think we should use
C++ compilers which of course allow the use of
C.
Regards,
Rob
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:30 BST