Misc ideas & comments

[strandh@cs.utexas.edu]

> >         1. I tried RScheme, but it seemed to me too complicated, at
> > first in a first glance. After compiling everything, the whole tree
> > ocupied about 70MB of disk space. I had some difficulty to find the
> > main executable in the dense tree structure. I had to use "find
> > . -perm 111 -type f" to find it! But of course, this is my ignorance
> > speaking... The bottom line is that RScheme might be too complex to
> > allow a fairly good degree of exploration. And by the way, does
> > RScheme uses a virtual machine?

RScheme is not very complicated in my opinion.  But there are still
many places in which the documentation is not very explicit. 

RScheme does use a virtual machine. 

> This is my impression as well.  I looked into the RScheme sources
> because I was looking for a Scheme system to which I could add a nice
> MOP based object system.  In RScheme this is somewhat tedious to do
> because you have to update the Scheme code and quite a lot of
> hand-written C as well.  

I am not sure what you mean here.  RScheme already has an object
system.  Why was it necessary to update C code to add an object
system?  Would it have been easier in some other system?

> Furthermore my (admittedly not very thorough)
> experiments with RScheme 0.7.1 indicated that performance was not too
> hot, even for compiled code.  

True, but we are working on improving performance.  The emphasis so
far has been on clarity and simplicity. We think we can combine the
two, however.

> (RScheme can compile to C via the rsc
> module compiler, however you have to `make rsc' in the "src" directory
> to generate rsc, 

We are working to simplify this procedure. 

> furthermore linking modules compiled with rsc v0.7.2
> to the main executable fails on my Linux/glibc system.

This should be reported as a bug.  

-- 
Robert Strandh

---------------------------------------------------------------------
Greenspun's Tenth Rule of Programming: any sufficiently complicated C
or Fortran program contains an ad hoc informally-specified bug-ridden
slow implementation of half of Common Lisp.
---------------------------------------------------------------------