[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: several hundred ugly things


We've reached the odd state where "aesthetic" is used perjoratively, as in
"merely aesthetic". I think of aesthetic considerations as strongly related
to the properties of elegance, ease of teaching, learning and
understanding. I tend to enjoy using computer systems that are
aesthetically pleasing. 

Personally, I reached the conclusion long ago that it would not be possible
to make major inroads into the aesthetics of Common Lisp by any small set
of "cleanups". What makes Common Lisp hard to teach and learn is not
necessarily even caused by a lack of elegance or simplicity. There is some
evidence that Scheme is hard to teach and learn, although this has been
mitigated by the excellent effort that has gone into teaching materials for

I think I will claim that I was misquoted in my reference to ISO --
although perhaps I merely mispoke. I think there is much interest in
defining a language that is smaller than Common Lisp, e.g., that works well
with the current and pending generations of laptops and PCs.  While I heard
this interest most recently expressed at an ISO meeting, it is of serious
concern in the US as well, e.g., by those considering Lisp for applications
in embedded systems. 

However, on reflection, I believe that the issues of determining those
features of CL that should be declared "obsolete" is nearly orthogonal to
the issues of "delivery subsets" or some such. I think we can define a
class of features of CL that should be Obsolete. How these are to be dealt
with in any implementation is an important concern, but will be easier to
determine once we have the list of "obsolete" features. (I would include
:TEST-NOT and BIGNUM and STRING-CHAR and some other misfeatures into the
"obsolete" bucket; perhaps you have your own list.)