On Feb 13, 2007, at 6:24 PM, Christian Heller wrote:
Dear Excalibur and former Avalon Developers,
Hey Christian,
some years ago, I started coding on the http://www.resmedicinae.org
open source project, at that time in Java.
I investigated software patterns and also found Apache-Jakarta-Avalon,
whose ideas I implemented as good as possible. The lifecycle idea
helped me a lot to receive clean source code.
I then found out that using interfaces often caused overlapping or
redundant interfaces when inheriting classes.
This is how I decided to ban interfaces at all and stick with pure
class hierarchies. It was the idea of ontological structures
(pure hierarchy of objects grouped into layers/ levels by their
granularity, with only unidirectional relations between layers)
that helped me here.
I also struggled with the number of parameters to hand over to a
lifecycle interface in Component Oriented Programming (COP).
So I thought a lot about nature, cell separation, DNA etc. and this
is how I got the idea to keep just *one* "Configuration" object
serving
as knowledge tree (comparable to the DNA which is forwarded from cell
to cell) and hand it over to all procedures. And so on and so on ...
Heh. Interesting. PeterD did something not unlike what we called
"Avalon5" for a while, which had a piece to it called DNA.
Coincidence? :-)
After five years I now managed to put all ideas into a book, which
I would like to announce to the developers of this list, because
I was greatly influenced by the ideas of Avalon. Here is the book:
http://cybop.berlios.de/books/cybop/index.html
It is published under the GNU FDL license.
Congratulations on getting a book out...I've never done it but from
what I hear it is usually *tons* of work.
I have signed this list some days ago and will stay here for some more
weeks, in case somebody wants to discuss the CYBOP ideas, COP, IoC
etc.
I wish I had the time! :-)
These days, when I talk about "metadata", it tends to be about RDF
and not so much about custom-parsed-javadoc hints to container
software...which is quite a different world (graphs, or rather a
"really big graph", instead of trees and containers)
best regards and good luck marrying paradigms!
/LSD
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
