[self-interest] Why did Sun stop development of Self ?

Jecel Assumpcao Jr jecel at merlintec.com
Fri Dec 29 18:24:35 UTC 2000


many people thought that Sun wasn't even big enough to impose one
language on the market by itself. While now we have examples of single
vendor languages becoming popular (Perl, Python and Rebol is coming
along nicely) which is encouraging others to try (Microsoft and C#), I
can only think of dBase and PL/I as examples of this happening before.

About the Smalltalk market, it didn't look too good back then. Both IBM
and Parcplace-Digitalk seemed to abandon Smalltalk to focus on Java.
And I would say that the Smalltalk people could be divided into two
groups: those who liked the idea and those who had to abandon Cobol but
were afraid of C++. The commercial success of Smalltalk in the early
1990s made the second group the largest by far and these people were
more likely to want to move on to a "nicer C++" than a "better
Smalltalk". Now that most of these people have gone away, the level of
interest about Self among those who remained with Smalltalk is growing.

> Examining other prototype-based languages like Agora (Vrje University)
> or Kevo, they are a far cry from the simplicity of Self. They do not have
> the musical Kansas that I hope could be really transformed into
> a true virtual reality tool!!!

I would just like to comment that Kevo is actually much simpler than
Self. Perhaps too much.

About the nature of the written material for Self, nearly all of it was
published in conferences and workshops for language designers and
implementors. The two tutorials (Manchester and ECOOP95) are the
exceptions. The same was true of Smalltalk until the August 1981 Byte
magazine effort.

I wrote:
> > > It probably seemed easier to grow Java as needed
> > > than to slim down Self.

I should have mentioned that I didn't believe that back then - this is
a variation of the famous "Turing Tar Pit". My opinion was that a
modest investment would have solved Self's Snapshot problem while not
even all the money in the world could extend Java to do all that Self
was doing. I am still waiting for a demo of a Java product that will
prove me wrong...

About investments, it is a matter of putting some effort into
marketing. Squeak nearly faced the same fate as Self, but they were
able to move to Disney and escape from Apple's problems.

Albertina is a little optimistic - just because you see someone with
what you think are worse ideas than yours getting million dollar grants
it doesn't mean you can too. That is like saying that because some poor
kid was good at sports and was given a chance to become very rich that
every child in a similar situation should count on this. But, on the
other hand, if it did happen to someone then it is possible and we
shouldn't give up without trying.

Most potential investors have never heard of Smalltalk, much less Self.
And that is actually not too bad - those who have heard of them have a
vague impression that it is something that has been tried already but
failed miserably for some reason. Many who invested in Linux didn't
know it was the same thing as Unix, which many people thought had been
killed by Windows NT.

This is making less and less sense, so I'd better stop...
-- Jecel

More information about the Self-interest mailing list