Genius research. A meta-analysis of the language with which linguists and computer scientists construct language for use. What a breadth and depth of historical and applicable knowledge!
This should be mandatory viewing for anyone trying to read CS papers. The notation really hindered me after reading physics stuff first. Some additions: - He didn't talk about geometric/Clifford algebras which are like more abstract quaternions - this language does have some implementations more or less in the Racket dsl Redex, and also in the K framework
Nice talk. I wonder if Guy Steele is aware of the Shen programming language (a lisp) which uses CSM code (sequent calculus) for defining types. Shen has an integrated prolog interpreter so I guess it wouldn't be too difficult to rewrite Guy Steele's code in Shen, thereby making it possible to directly import CSM from computer science papers into Shen.
18:47 zsh actually still has # and ## as postfix quantifiers in addition to the more common ksh/bash prefix notation, was the only time I had ever seen it until now lol
very nice talk! what I noticed, at 24'30, is that the Pascal notation, for describing types are readable and understandable by almost everyone; the others one's, are uselessly combersome; what is the main difference? the universal power of a picture, far better than any "blabla". people know it since they begin studying geometry : draw a picture to figure out the thing if far usefull than any gossip about the thing. infortunatly, in CS, they don't learn how to draw picture : in all other sciences like physics, chemistry, biologie, drawing pictures in mandatory.
Typical programs use far too many levels of nesting over too many lines for that to be viable. The Lisp solution to this problem at the meta-level is what inspired Steele in the first place.
@@0rkk0 It's much the same of the undocumented, arbitrary, context-dependent notation that's used in math and physics, and I would imagine that it's because the people using this kind of notation tend to be more mathematically-minded rather than linguistically-minded. (Hence Lisp already having the solution)
Genius research. A meta-analysis of the language with which linguists and computer scientists construct language for use.
What a breadth and depth of historical and applicable knowledge!
Great talk. Nice to learn about the origins of the computer science languages we used today.
This should be mandatory viewing for anyone trying to read CS papers. The notation really hindered me after reading physics stuff first. Some additions:
- He didn't talk about geometric/Clifford algebras which are like more abstract quaternions
- this language does have some implementations more or less in the Racket dsl Redex, and also in the K framework
Nice talk. I wonder if Guy Steele is aware of the Shen programming language (a lisp) which uses CSM code (sequent calculus) for defining types. Shen has an integrated prolog interpreter so I guess it wouldn't be too difficult to rewrite Guy Steele's code in Shen, thereby making it possible to directly import CSM from computer science papers into Shen.
18:47 zsh actually still has # and ## as postfix quantifiers in addition to the more common ksh/bash prefix notation, was the only time I had ever seen it until now lol
Thanks!
very nice talk!
what I noticed, at 24'30, is that the Pascal notation, for describing types are readable and understandable by almost everyone;
the others one's, are uselessly combersome;
what is the main difference?
the universal power of a picture, far better than any "blabla".
people know it since they begin studying geometry : draw a picture to figure out the thing if far usefull than any gossip about the thing.
infortunatly, in CS, they don't learn how to draw picture : in all other sciences like physics, chemistry, biologie, drawing pictures in mandatory.
Imagine if Lisp used overlines rather than parentheses...
Typical programs use far too many levels of nesting over too many lines for that to be viable. The Lisp solution to this problem at the meta-level is what inspired Steele in the first place.
There is an editor created by some guy in Seattle... forget the name and can't find it right now... but it uses nesting in boxes instead of parens.
9,5k views. No comments. This stuff is real scary.
Well, it's so good it doesn't need commenting :P
at best it illustrates that notation in scientific CS literature is a zoo :-|
@@0rkk0 It's much the same of the undocumented, arbitrary, context-dependent notation that's used in math and physics, and I would imagine that it's because the people using this kind of notation tend to be more mathematically-minded rather than linguistically-minded. (Hence Lisp already having the solution)
@@0rkk0 The history of math notation is an equivalent mess.
Borrowing an idea from quasiquoting comes at 52:52
2:20