Ideas about the Software Development Cycle



First, most software development tools require very precise
description of the problems, and precise syntax for certain
languages. And if we do not give enough details, boom.
We got suck anyway and don't know where to go.
But guys like me do not have precise details when I began.
I have very vague ideas like this one. A little by little
these ideas begain to take shapes by itself.

Can we translate these idea evolving process to the software
development cycles? I think there may be some way to do so.
One aspect of this model allows partial specification of ideas,
and some details can be left out. How this is possible is
well lets figure out later, but I think there could be some
method to do so. Partial specification naturally introduces
partial computation or partial mapping.

We can use
1. some symbolic computation method or
2. lazy evaluation or
3. use default value or
4. compute by example method or
5. ask questions when you don't know how to do method.
(some idea about partial computation.)

Each computing entity(or module or instance of class ...) consists of
1. message receiving part
2. parsing and figuring out what that message means part
3. providing service for that received message (body)
4. when needed we may ask some other service from other computing entity
5. delivering result by message part
6. when we are delivering results we can deliver to
several relavent entities so each entity can use
these result when they are ready or ...
-----------------------------------------------------------
7. each entity can act as a client or as a server or both
8. entity is not sub program or module
they are rather like COM - server or CORBA - server
9. they are similar to web browser and web server combined.
10. message is consist of xml style document
11. each entity contains xml-parser
12. there are usual CASE-tool like services for message passing
and receiving...

You may wonder well I have seen all of these things before,
so what is big fuss about this model or new methodology things?

I think this model is different from other models.

1. This model is guranted (can be mode ) to run no matter what
the specification of your entity is.
2. This model allows partial specification and later can introduce
more detailed specification.
3. You can play with your ideas easily.
4. You can observe that how your rough idea is translated to some
beginning stage of each entity and how those entities are affected,
in other words dynamic nature of your specifications.
5. Even you didn't provide the body of entity, default body is created
and all the information is logged automatically for later use.


July 8, 1999 Sang Cho

Dear Uknown

The ptb's know all this btw, and that is the master plan: You see Digital sequence is 101010101010....Nil sum Game

Anologue seguece is 1234567890.....random kaos theory...Nil sum gain

Digital sequuence is reversalable and infinite......you can see them, but you can't see them seeing you? Gognito ergo sum?

XfactorY over Zz's top! e=mc squared time 3 is dialectical equality?

Cooollllll?

I do know you....you're ummmm......ummmm...nah, lost the plot again sorry?

Love Jasper