Wherein we ask some questions

If you have questions for our conversation then list them here!

(Patrick) As a graduate of an information "science" program, I was really struck by this quote from a 2011 interview you did with Peter Seibel:

Seibel: So if computer science, like all other fields with science in the name, isn’t a science what is it that we’re doing? Engineering? Art? Math? Something completely new?

Abelson: Well, it’s this funny mixture of formalizing knowledge and thinking of ways to do that and then looking at the results of this information communication stuff mixing up with culture and society and the way people interact with each other. And that’s all very complicated and very deep. I think the future of computer science is at places like Cornell and Michigan and Maryland, where they figured out that they should merge their computer science departments with their information schools.

Seibel: Information schools as in library science?

Abelson: Yeah, they come out of library science. Cornell is a really good example of that. The computer science department, I believe, actually moved out of the school of engineering and now there’s an information/computing merger. I think that’s the direction this is going to go, and some of it will happen formally. Some will happen by people shifting their research focus.

Seibel: In the beginnings of computer science, it seems there were two ways it emerged: it either came out of the math department or it came out of the electrical engineering department. So you got either theoretical, mathy computer science departments or very systems-focused departments, like at Berkeley, building things like Unix.

Abelson: Right. And neither of those are where the important stuff is now. I think the interesting stuff is in this third area.

Seibel doesn't really bite here, but I'm wondering if you could talk at all about this "interesting stuff" that is coming out of places like information schools, if your perspective on that has changed in the past 3 years, etc.

(scott)

• How did you learn to program? On what machines?
• What was the reason for writing the original? How did you get involved?
• How did you decide to organize the book the way you did?
• Ch 1 and Ch 2 purely functional, then Ch 3 involves state and assignment and it's like a bomb drops then Ch 4 is both theoretical and exploratory, and Ch 5 is super fiddly, and is actually closer to my experience of where most of my time actually goes.
• What were the most major changes in emphasis over the time that this book was used to teach CS students?
• It was eye-opening how all programming models were presented as facets of the same thing in sicp, variations on a theme, however, that isn't the standard way to teach people programming today, even at university, why not?
• Some of the sections seemed like they were intended to be explorations or thought provokers, amb evaluator, event driven digital signal simulation, constraint propagation, garbage collection, how did you decide what to include?
• What's up with all the bank account examples? I mean my father is a banker so I have an excuse, what's yours?
• Where do you see the book today?

(Leo)

• Why no demonstration of macros in the metacircular evaluator?
• What should I read next?

(Andrey)

• Thoughts on the MIT switch from scheme to python in the introductory course? Has there been feedback from the results of that change?

(Ian)

• Thoughts on Logo?

(Dann)

• The Scheme chip in Chapter 5 is interesting -- any stories about it?
• Changing the meta-circular evaluator in Chapter 4 to highlight non-determinism, logic programming, and laziness is amazing. Where did that idea come from? How useful is that approach of modifying the evaluator in practical code -- is it mostly a pedagogical technique?