[self-interest] Self & Engelbart's vision

Randy Smith randy.smith at gmail.com
Wed Feb 6 01:28:57 UTC 2019

I like this in your email:

>  it is only natural, that things converge 
> to the same point if you have sufficiently permissible system. 

... and would say that is more what was going on.

At PARC, I was interested in the “desktop” metaphor, and naturally wondered what would happen if you took things to an extreme and made everything into a physical object.  My PhD is is in Physics, not computer science, so my impulse was to look for a simple model to explain everything. Smalltalk had already shown the advantage of everything (Boolean, integers, etc) as computational objects. The physical metaphor led to the Alternate Reality Kit, and the notion of making new things out of existing stuff, so a small step to prototypes over classes. Also in ARK, representing instance variables only as accessor methods meant objects could only send messages to change state, even to themselves. 

Working with Dave we figured out how to put these ideas into a first-class OO language, Self. And Dave and his students even figured out how to make it run, which amazed me.

So as Dave said, the influences were powerfully present in Smalltalk and the creative minds in the hallways of PARC. We knew of NLS, but for me it was something one heard and read about, rather then touched and used daily as with Smalltalk, for example. It’s not clear to me how much NLS influenced Smalltalk, if it did, it was before my time there by a few years I guess. Stu Card for example was working with the PARC Lisp group and helped refine the mouse into being pretty much what we have today, which was what I was introduced to at PARC, a lot different than the NLS system as I recall it. So again, indirect influence through PARC.

More information about the Self-interest mailing list