Re: [Chicken-users] Please do not drop 'thread-terminate!' from the SRFI 18 impl

[Vincent Manis]

On 2008-Aug-10, at 11:31, Elf wrote:

> [3]> (car 'foo)
>
> *** - CAR: FOO is not a list
>
> in clisp's implementation.  car and cdr did work, though.  the NIL  
> is way,
> way too overloaded in CL.

I know I said I'd shut up on this topic, but... :-)

On the first Lisp system I used for real programming (Lisp/MTS, a sort  
of perverted version of Lisp 1.5, running on the MTS
operating system circa 1971), (cdr 'symbol) gave you the property list  
of a symbol. I had done some trivial programs with
PDP-8 Lisp a year or so before, and, as I recall, (car 'symbol) gave  
you the print name (as some datatype that wasn't very
useful) and (cdr 'symbol) again gave the property list. (I believe the  
plist as cdr of a symbol was official Lisp 1.5 behavior.)

The problem with appealing to prior practice in Lisp systems is that  
there's so much of it, and so much of it that is
contradictory.

On the subject of threads, I would be quite comfortable with a strong  
warning about the perils of thread-terminate!. I have
designed a fair amount of thread-based code over the years, and taught  
concurrency in industrial settings. I have never
found a problem that *required* unscheduled thread termination.  
Normally I design threads either to terminate when they
receive a special marker in their input queue, or to poll for a  
termination request.

-- v