October 03, 1986
OOPSLA '86 Reflections
OOPSLA 1986 Zeitgeist: Objects are good for: Nice User Interfaces / Disributed Processing
An irony: OOP does a much better job of allowing one to solve "Pascal-ish" problems than does Pascal.
People like "readable" papers. (People evidently thought these were "readable")...
Distribution of library materials is important. Someone called Click Art the only successful attempt so far at distributing a library of anything...
Randy Smith's (XEROX) Alternate Reality Kit was perhaps the single most impressive thing I saw at OOPSLA '86.
The Most Impressive List:
- Danny Bobrow
- Henry Lieberman
- Peter Deutsch
- Alan Borning
I was taken with Bobrow's notion of Linguistic Imperialism.
Both Kristen Nygaard and Alan Kay gave good talks.
Peter Deutch observed that it is necessary for us to borrow from the graphic arts (where have I heard this before). This was the other impressive thing.
Mark Sherman (CMU-ITC) gave a convincing testimonial for MacApp.
Other notions propounded: OOP isn't important; OO Evnironments are.
Parallelism / Concurrency was underrepresented.
It's one thing to rearrange things, but a new idea, that's Christmas.
--Bob Weir
Further, it is impossible to hoarde knowledge. Don't be afraid to be wrong now and then, because you will be anyway. The alternative is that there is no correction in the loop.
Trust thy judgement, for thou art stuck with it in any case, and it is more comfortable to do so.
--Thomas Jay Peckish II
Best Exchange:
Kay: Good ideas don't always scale well...
Lieberman: So what do we do? Just scale the bad ones?
Tim O'Shea is quite funny.
A lesson from OOPSLA: The "big boys" skim nearly everything, and read only things in their specialty...
Software Maintenance: People percieve maintainers sort of as software busboys; cleaning up after the designer and analyst have eaten. Academic work in this area seems to have been retarded by the fact that people seem reluctant to become the world's most famous busboy...
--Thomas Jay Peckish II