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

Re: Strong Typing, Dynamic Languages, What to do?

Shriram Krishnamurthi <sk@cs.brown.edu> writes:

> Oscar, you are clearly a troll.  I wish you luck finding some other
> Schemer's time to waste.

I disagree with your assessment of Oscar.  He is a big advocate of PLT
Scheme in the Borland C++ groups, and has contributed to your efforts.
He ported mzscheme to Borland for you, and continues to encourage
people to use it.  He's also a very knowledgeable C++ programmer (I
see him post frequently on Borland) and to dismiss him like you did is
a bit rude, I think, and is ignoring the fact that he is asking some
good questions.  (At least in my opinion they're good questions.)

As an advanced C++ programmer myself, I think he is representing us
fairly well, and he is voicing some of my concerns too.  This thread
was a good opportunity to dispell some misconceptions we may have of
functional languages, garbage collectors, etc.  But instead, it just
got ugly.  That is unfortunate.

Oscar brought up a point I hadn't considered before and I was hoping
you'd answer it (you are the expert and I value your advice.)  Once
you start having continuations in the picture you have entire
callstacks unable to be garbage collected, but the GC has lots of
objects laying around (in the continuation) needing checking.  How
does it remain efficient to garbage collect if the continuations keep
hundreds (thousands) of objects laying around needing to be checked
for references?  It seems to me the more objects that exist that
AREN'T garbage, the worse a GC will perform.  Is that a false premise?