From: Robert Swindells <rjs@fd...> - 2000-11-07 02:45:39
>1 - a straw poll about whether a "case sensitive" reader would be a
> good thing.
I learnt lisp using FranzLisp, so I tend to prefer the idea to case
I always used normal lower-case-with-hyphens symbol names though, I
just prefer seeing the same symbols that I have typed in.
I can see Kent Pitman's argument that you can hide FFI calls in code
(defun make-string-i/o-port (...) (jcall "MakeStringIOPort" ...))
but you still have to remember two sets of symbols or invent a strict
I have written quite a lot of Motif code in C, but I haven't used CLM
much since I can't remember how to map from the Motif function names
to the CLM equivalents.
>2 - what would be the "cost to implementors" (of - at least CMUCL and
> CLisp) -- e.g. what would it take to change all internal symbols
> to lower case.
I once did this for AKCL - converted all the built in symbols to lower
case and added the extra CLTL2 readtable functionality.
It wasn't all that much work to do the change, I probably spent more
time getting various packages to work.
The kinds of things that break in user code are the ones raised in
c.l.l - packages and code that does (format nil "FOO-~A" bar) to
I think I stopped using it because of problems with Garnet.
>3 - whether the change could be accommodated without changing too much
> the ANSI standard. I.e. how the change can be accommodated in
> such a way to allow strict ANSI programs to be accepted.
>4 - if the ANSI standard needed to be changed, then how should a
> consensus be reached, so that it somehow binds the implementors to
> the set of agreed upon choices.
From: Bruno Haible <haible@...>
Cc: cmucl-imp@..., clisp-list@...
> 1 - a straw poll about whether a "case sensitive" reader would be a
> good thing.
YES. The debate whether case sensitiveness in languages is over.
No, it isn't. If you say "yes, it is" that will demonstrate that in
fact there is still debate. I think Lisp should be left alone. Lisp
does a lot of things right that other languages do wrong.
Among other things, I use Lisp via dictation, and getting case right in
dictation is a real pain.
From: Raymond Toy <toy@rt...> - 2000-11-07 14:10:53
>>>>> "Robert" == Robert Swindells <rjs@...> writes:
Robert> I always used normal lower-case-with-hyphens symbol names though, I
Robert> just prefer seeing the same symbols that I have typed in.
Same, but I don't care what case the symbols are print in, because I
know Lisp is uppercase.
Robert> I can see Kent Pitman's argument that you can hide FFI calls in code
Robert> (defun make-string-i/o-port (...) (jcall "MakeStringIOPort" ...))
Robert> but you still have to remember two sets of symbols or invent a strict
Robert> conversion algoritm.
What are you going to do about C++ name mangling? What about Fortran?
I've used different Fortran compilers whose resulting linker names had
a prepended underscore, an appended underscore, or was in uppercase.
Using case here only solves part of the problem.
Of course, you can always use |MakeStringIOPort| as your Lisp function
if you like. Then you only have to remember to put || around your
I think FFI is a totally separate issue from case.