[Openmcl-devel] Scheme vs. Lisp [was A plug for nx1-combination-hook]
dlw at itasoftware.com
Fri Sep 11 14:47:31 EDT 2009
Alexander Repenning wrote:
> On Sep 10, 2009, at 9:49 AM, Daniel Weinreb wrote:
>>> The OpenGL examples in PLTscheme are downright embarrassing.
>>> One can speculate why Scheme is is loosing ground. I am guessing it
>>> has more to do with actual functionality, e.g., continuations really
>>> are not a substitute for OOP, than with syntax.
>> I don't know that it's losing ground. If you mean MIT's
>> freshman core curriculum, the reason for the change
>> is nothing like that; see my blog writeup at
> As far as I can tell the the reason implied in the interview does
> match the general observation that students do not pay too much
> attention to theoretical concerns but are interested in the concrete
> aspects of programming.
Actually Gerry didn't say anything about what students want.
What he was talking about was what it means to train engineers.
There have been other debates that are different from this one,
but do impinge on some of the same issues. When we say someone
goes to college and comes out with a "computer" degree, what does
that mean, or what should it mean? The field is very large; what do
we consider the subset that every graduate should know? When I
was in college, if you did not understand the "machine language"/
"assembly language" level of abstraction, you were really not
competent to do anything in software. But nowadays, how many
people program in, write code that generates, or debug at the level
of Intel assembly language? I sure don't know Intel assembly
language, and I don't see any reason I should bother to learn it.
All I'm saying is that it's not entirely clear what subset of
computers everybody has to know.
It's also possible that there is room these days for more than one
degree. MIT has long had a "Chemistry" degree and a "Chemical
Engineering" degree, for example.
Anyway, this is pretty off-topic...
> Students want to work on personally meaningful projects and,
> typically, not to explore interesting theoretical tidbits. Are
> continuations conceptually as powerful as, say, OOP in CL? Most
> students simply do not care. Can they hack an app for their phone, or
> make a 3D game? Completely different story. Send a student to explore
> something like PLTscheme and have them run the OpenGL examples. Seeing
> that it can barely handle OpenGL 1.0 students will file that
> experience in the same fun category of playing with punch cards and
> core memory.
> All the best, Alex
> Prof. Alexander Repenning
> University of Colorado
> Computer Science Department
> Boulder, CO 80309-430
> vCard: http://www.cs.colorado.edu/~ralex/AlexanderRepenning.vcf
More information about the Openmcl-devel