OO future (was: An OO history)

[Jecel Assumpcao Jr]

On Monday 23 April 2001 17:05, Steve Dekorte wrote:
> On Tuesday, April 17, 2001, at 11:33 AM, Jecel Assumpcao Jr wrote:
> > [...] I have found it nearly
> > impossible to change people's minds, to "covert" them...
>
> Right, that's the problem I'm getting at. It seems there's so many
> great technologies/ideas that sit on the shelf because inventors
> don't understand marketing. It's been said that in engineering there
> are no technical problems, only people problems. For those of us that
> want to see the innovations of Smalltalk and Self in widespread use,
> our biggest problem isn't technological. It's figuring out how to get
> people to use new technologies.

Easy: get everyone around them to use them. Most people go with the 
crowd. Obviously, this leads to a recursive bootstrap problem :-(

To solve this problem, I am targetting schools which want to look more 
modern than the others. If all students in a class get a new computer 
on the same day, then they will be using what everyone around them is 
using without any of them having to be the "early adopter".

> >> In my experience, most developers find Smalltalk style code very
> >> difficult to read.
> >
> > That is very interesting to know. What is their background, in
> > general?
>
> C, C++, Java or JavaScript. (I think most developers have such a
> background.)

I would have expected that. Many people from a Wirth language 
background can't read C derived languages. See Oberon and Delphi fans 
for example. This is silly - everyone should be required to write 3 
good programs in APL and 4 in Lisp before being allowed to call 
themselves programmers...

> That's a good point. But then you have to ask, why where those
> choices made
> by Netscape or Sun, etc?

Since I have to, "why were those choices made by Netscape and Sun?"

I suppose you mean a C-like syntax for Java? Everybody was doing it 
then (see Renderman and others from those days) just as everyone was 
doing Pascal looking languages the previous decade. It made sense for 
Sun, as a Unix shop. It made sense for Netscape, which probably wrote 
all its code in C.

But I bet if Java had been Python-like and Sun and Netscape had pushed 
it, it would be almost as popular as it is now. Just a guess, of course.

> > The problem with your theory is that many innovations work much
> > better together than in isolation.
>
> I think it depends on your market. If you're after the .01% of
> developers that are into neat new things, then packaging more neat
> new things makes sense.

I mean like: persistent objects and capability based security. Adding 
just one of these to system is far from half as good as adding both at 
the same time. I am talking about synergy.

> But if it were your goal is to change the
> world(in your case it isn't),

Actually, it is. In fact, I intend to change it twice in a row just so 
that nobody can claim I just got lucky.

> you may be after the wrong market. It
> might inspire the folks who do change things, but those changes may
> only be accepted in a piece meal way anyways and not in order of
> importance.

But consider a person's first computer - there is no possibility of a 
piece meal change.

> As examples consider the Alto vs the Mac or the Newton vs. the Palm.

Alto - cost over $20k and Xerox never tried to sell it.
Mac - $2500

Newton - every time a model dropped below $1000 it was eliminated. Jobs 
killed it because it reminded him of Sculley.
Palm - $400 and lower and the batteries lasted a long time

And Dru Nelson asked:
> What about the Mac vs. the PC?

I guess that would support Steve's thesis: the Mac was a whole change 
and the PC reached the status of a Mac-clone in a piece meal fashion.

-- Jecel