[plt-scheme] A crazy idea (fMUMPS)

From: Greg Woodhouse (gregory.woodhouse at sbcglobal.net)
Date: Thu Dec 29 13:14:57 EST 2005

--- Matthias Felleisen <matthias at ccs.neu.edu> wrote:

> 
> In that case, you have to find two things:
> 
>   1. Something people need but doesn't exist and
>      appears to consume too many resources given
>      the existing infrastructure.


That's not hard. I've already developed what I called the "generic
query". After developing several HL7 interfaces (HL7, or Health Level
7, is an application protocol widely used in the field) I realized I
was doing essentially the same thing each time, so I wrote a program to
drive interfaces from templates that were just XML documents. It would
be easy to optimize by generating code directly from the template and,
in retrospect, it occurs to me that this is just a special case of
partial evaluation (http://partial-eval.org/). I've also optimized
queries by creating a subscription registry and "pushing" updated
values during otherwise dead time, gaining about an order of magnitude
in performance. Another area in which I've done a fair amount of work
is writing parser driven tools to work through the code base and build
depency graphs. It's the very necessity of doing this that led me back
FP because of the improved compositionality. I find it exceedingly hard
to sell users (doctors) on the idea that investing in infrastructure
will reduce future development time when they want a working product
NOW. Of course, I could extend this list, but findintg problems to work
on is not terribly hard.
> 
>   2. A physician/med person who's willing to work
>      with you on a test basis.
> 

That's harder. Much harder. I've got to figure out how to make a
compelling case for building the infrastructure to support new modules,
because there isn't much point in proceeding without physician "buy
in". Something for me to think about, I suppose.

> That's how Motorolla started using Chez Scheme for its catalog.
> 

I didn't know that.

> -- Matthias
> 
> 


===
Gregory Woodhouse  <gregory.woodhouse at sbcglobal.net>
"Interaction is the mind-body problem of computing."
--Philip L. Wadler


Posted on the users mailing list.