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

Slightly Disappointing Review of MCL 2.0 in October MacWorld :-(

Hi all,

in the latest MacWorld is a half page review of MCL 2.0. MCL got three stars out
of five. This is extremely unfair. Wouldn't most of the MCL users agree
that five stars would be correct? ;-)

  "PROS: Includes easily extensible EMACS-like editor; includes good sample
  code; compatible with CLOS; excellent documentation in manuals and on
  CONS: Limited interface construction tool does not include CLIM."

CLIM is only in Genera included (as far as I know). I'm not sure
if CLIM should be included in MCL at this time. Maybe later when
CLIM is mature enough it could be interesting. I really don't
know. I like the ideas of CLIM, but it is a big and complicated
package. Does the average (?) lisp programmer need this power?
(Sure there should be a fast and stable CLIM 2.0 implementation
for MCL 2.0.) 
Peter Wayner is wrong when he writes that
  "Common Lisp includes ... a standard interface system known as CLIM".

O.k., the interface construction tool may be limited. There is a lot
to be done. The MCL team certainly knows this. But I'm not sure if 
*everything* can be provided. Large parts of interface construction is
*programming*. Especially when the interface has to be sophisticated,
one needs to extent existing tools or one wants to implement new ideas.
How extensible are the interface tools? Is it easily doable?
These are interesting questions.

(One thing that I don't like about MCL's Interface Builder is, that
code generated by it, is not structured in a way I could use it in 
a program. It has to be edited to fit to the rest of the program.)

The review lists as strengths of MCL: FRED, debugging facilities, interface
builder, CLOS, testing of code (Debugger, Inspector).

  "When Macintosh Common Lisp first arrived, its interface generator
  was one of the best in any language. ... Macintosh Common Lisp's
  function selection is limited, but it can be extended by modifying
  the Macintosh Common Lisp source code (included), or by buying either
  the Common Lisp Interface Manager ..., or Action, a set of more
  sophisticated interface-design tools... Despite its lack of
  ready-made interface facilities..."

I'm not sure what product Peter Wayner is reviewing ;-). Is it
really MCL? Sure he missed most of real strengths of MCL.
Some are (IMHO): Access to the Macintosh Toolbox, ease of usage,
extensibilty, quality of implementation, stability of the environment,
excellent support through newsgroup and email from Apple and the
user community, code provided by users, a lot of the Common Lisp AI
software is running on MCL, ephemeral garbage collection, ...

And he sure missed some of the weak points of MCL...

Maybe I should write a letter to MACWORLD1@applelink.apple.com and
ask for a corrected rating. ;-)

Rainer Joswig
Email: joswig@informatik.uni-hamburg.de