On Thu, Jul 31, 2008 at 09:59:17PM -0700, Andrew Lentvorski wrote:
From my point of view, the big problem with Scheme/Lisp has always been
the community. The community never seems to want to sit down and do the
hard work of playing nicely with the rest of the world (unit tests,
documentation, external FFI's, etc.).
This is basically why I'm mostly abondended trying to do anything
practical in Common Lisp. I love the language. But:
- There aren't all that many bindings. The bindings that are there
are scattered and fragmented.
- Writing new bindings is really hard. None of the implementations
have agreed on an FFI. There is CFFI/UFFI which try to bridge, but
they're very much works in progress.
- The execution model of starting a lisp "system" and interactively
developing may be productive for certain things, but it tends to
make it difficult to produce useful programs in the end. When the
end-result of my short script is a 50MB executable that takes
several seconds to start, I'm not going to tend to use it for short
things.
- Some things that are kind of considered basics today, such as
threading, are implemented as vendor-specific extensions. Some
implementations don't even have threads, and others do it poorly.
David
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
