Prism criticism

[Jim Little]

Maneesh Yadav wrote:
> 
> The example you provide only shows how nice it'd be to write dynaimc HTML
> in that manner.  Now tell me how does prisim help me do this?  All the fun
> stuff like interfacing calls between the layers, parameter passing and
> possible format conversion that you have to have setup to do the type of
> stuff in your example (on top of all the glue that has to sit in the web
> server), I don't see anywhere.

All of those things would be handled under the covers when the code was
compiled.  That's the whole point -- given a sufficiently specialized
language, a lot of things can be handled automatically by the compiler. 
In this case, the "language" is actually two languages, but they're
compiled into one program.  The compiler takes care of the gluing the
two languages together.

> Nor do I see how Prism helps me write the above routines.

It's true that each of these little languages has to be written by
somebody, as must be the individual compilation routines.  But it's like
a library... once a mini-language and compiler have been written,
anybody can use them, saving everybody a lot of time.

My eventual goal is for the mini-languages and mini-compilers to
interoperate.  That's what all the different pieces of Prism are for. 
The meta-metamodel provides a standard format for program models.  The
global heap provides a place for the different parts of the model to
reside.  And the interpreter provides a compilation engine.

At this point, I'm still researching how to make everything
interoperate, so don't ask me how I'll do that -- I don't know yet!  :) 
But the infrastructure is in place.

Jim

Prism is at http://www.teleport.com/~sphere