Enfin compared to other Smalltalks ...
In response to a previous posting asking about the difference between Smalltalk
s, and how easy it is to learn or port from one to the other, here
are my immediate thoughts. Others will no doubt want to add their own ...
It is easier to learn Enfin because the doc is better than other
Smalltalks. This comment is based on a European Digitalk distributor that
was asked to reimplement a large application in Enfin. They learned Enfin
and ported the Digitalk application in less than three weeks.
Smalltalks are like word processors. You learn one and you never want to
switch to another. Enfin is particularly irritating to some PP people
because of its "nonstandard" iostreams and dictionary classes. These
differences will go away over the next couple of releases of Enfin and the
syntax of all Smalltalks are virtually identical today.
For example, Caterpillar won the award for best OO application at Object
World in San Francisco this year. They implemented the system in Digitalk.
However, at the presentation, the project leader said that they had used
Synchronicity to generate much of the code. He recommended that people
used Synchronicity for building Digitalk applications, even though
Synchronicity generates Enfin Smalltalk.
So the differences are really less than you would think from reading about
the Smalltalks. Mostly, it is in the user interface code. But since most
userfaces are autogenerated today, that doesn't make much difference
All the Smalltalk vendors are on the X3J20 Smalltalk ANSI Committee and all
are committed to be compliant. All know they will have to change and
probably we are all making some changes today in preparation for the future
release of the standard, even though the standards process will take
Looking at the larger picture, good design is much more important than the
Smalltalk variant. Having a design document that is always consistent with
code is essential to obtaining consistent reusability of code. If you
cannot get good reusability of code, you fail to obtain dramatic increases
in productivity promised by the object paridigm.