Alan Kay: The Computing Revolution Hasn’t Happend Yet

Alan Kay: The Computing Revolution Hasn’t Happened Yet – Google Video.

Executive summary, if you don’t have an hour for the video:

The video is from OOPSLA 1997, and I suspect most of Alan’s slides predate that. Still, several of his points are no less relevant. Alan Kay is probably best know for inventing Smalltalk while at Xerox Park.

“I coined the term object-oriented, and C++ wasn’t what I had in mind.” (paraphrasing from memory) No, while a common theme of the presentation, ripping on C++ was not one of the timeless lessons contained therein (still fun though ;^) ) The lesson here is what ‘object-oriented’ is (Alan actually suggested that OO was a misnomer; I’ve forgotten if he offered an alternate) The big example was data sharing in the military way long time ago – there was no internet so they had to send data tapes. The problem is, you had various kinds of data records being sent around, and any system might need to read from any kind of data. The solution was to store three sections on the tape: a data section with the actual records, the procedures (program code) to access this data, and a standardized set of pointers into common entry point routines. Thus the receiving system just had to know to load a particular pointer for ‘read record’ and the stored routine took care of the tricky parts. The tape was fully encapsulated.

A similar point was made about HTML – it is not fully encapsulated; a HTML document says nothing about how to interpret it – all that knowledge is contained in the browser. Of course even the data tapes aren’t fully encapsulated – the system attempting to use them has to be capable of executing the stored procedures. (This last is my observation, not his.)

There was also the Japanese concept of Ma – the stuff in between things; the best english translation he could offer was interstitial. He pointed out that the length of a word is often inversely proportional to it’s importance.

There were several references to biology. Silicon computing power pales in comparison to biological (i.e. chemical) computing power. Comparison of encapsulation to a cell, with the cell membrane forming the interface. Comparison of C++ etc. to a bunch of organelles floating disconnectedly about.

Also, paradigm shifts – current thinking as a ‘pink box’, but with a little blue knowledge, you might have a blue insight and have a chance to jump to the blue box, (the ‘boxes’ were presented as intersecting planes) which may be a more appropriate way of looking at the problem.

Posted Sunday, February 25th, 2007 under Review.

Comments are closed.