[self-interest] partial conclusions about design patterns

[Brian Rice]

Hi Albertina,

>By now I think I know quite a lot about Self. For example all
>the methods connected with collections are simply not more than
>a list in the tutorial.
>What you may not understand is that Smalltalk was such a sort
>of OO language that I could not read.  I mean I knew nothing
>or very little about Smalltalk. When you state things about Self
>and you assume knowledge of Smalltalk I translate it as not
>explained things. This hinders the understanding of  Self itself.
>And many things that
>Jecel  designed for me in Self began to seem clear after my readings
>about graphical editors, graphical user interfaces and the  careful
>reading I have been undergoing  again of the design patterns.

I found Self much more interesting after studying the advantages and 
shortcomings of the Squeak libraries and environment, so of course I 
recommend this to anyone learning Self.

>I feel when I will start programming  (rather designing)  in Self 
>by March I will know
>where to start things.  Reading through all the tutorials written by 
>the Self group
>I could only experience how expressive and fit to my needs Self is.
>And of course Jecel is able to translate my needs into Self perfectly fine.
>But I do not understand what goes on in Jecel's head. I have to figure
>out things by myself. And all these readings are making me  feel confident.
>I know to write a tutorial for applications in architecture in Self seems to
>me as a big hurdle...
>Well in 1995 I would never imagine I could develop such a full-fledged
>work with symmetry groups of the plane and the dotless plane. Since
>Brazilian mathematicians failed I had to do it by myself! And I am extremely
>happy with the outcome.
>So by now I feel why not to be successful with Jecel's help and the help
>of the rest of the Selfers of this list!

If you don't mind, I'm working with groups and algebras in a language 
somewhat similar to Self, so I was wondering if you're willing to 
publish the source code (for academic interest only). I'm curious to 
see what methods and patterns you've used, and how the language 
negatively and positively affected the design.

Thanks in advance.

>Only after programming in Self for example  the last poster I developed
>about the relationships of the subgroups of the crystalographic groups
>pmg and pmm I will understand if the stratety pattern is not necessary.
>
>Having heard of the strategy pattern made me happy, because I have
>to face here with many unanticipated ways of unfolding the tilings,
>I mean the transformation of one subgroup into another.
>
>The geometric pattern changes depending where one begins, by the
>full geometric basic lattice, half of it, a quarter of it, and so on.

This definitely interests me, perhaps you've written down some 
material for a paper that you could share?

>Now I will start reading another interesting paper.
>And I am extremely happy that finally there is a book
>about knowledge based systems  that recognizes this
>may not be very different from an OO software!!!
>I knew this since 1990! KBS's community cared
>very little about OO programming.
>Anyhow I love Self and the OO community!
>
>Cheers
>Albertina
>

I love it too. :)

Thanks,
~