Cat Presentation to SeaFunc Users Group
This Wednesday, on April 25th, I had the good fortune and opportunity to give a talk to members of the Seattle Functional Users Group at Bellevue Library. A big thank you to Ryan Elisei for supplying a projector and Frank Krueger for videotaping it (I’ll announce the video on my blog when it is edited and available for download).
I was fortunate in that the audience wasn’t shy to ask questions and share their thoughts. I found the discussion both interesting and informative. Many of the questions and comments were very helpful and I feel that the next presentation I give will be much better for it.
The powerpoint slides are available for download here: Powerpoint Slides - Cat Presentation on April 25th to SeaFunc Users Group. Note, that only about 2/3rd of the slides were presented.
A big thank you to everyone who showed up!
On slide 49, I haven’t broken the code on the definition of curry. There is a function, dip, that is not defined (or is it a typo?). Also, is it the case that lists read from right to left? E.g., you say (2, 1, 0) is the same as nil 2 cons 1 cons 0 cons so it would be ‘(0 1 2) in LISP?
Makes sense, but I wanted to be sure. So [2 1 0] list car would be 0?
Comment by orcmid — April 28, 2007 @ 9:20 pm
Continuing (on slide 54 now):
Or would you say [2 1 0] list head is 0? tail is [2 1]?
Oh, OK, dip is on slide 58 and now I see how curry works. I’m not sure on the sufficiency/necessity around Turing completeness. I gather this is because the input type and the result type might each involve multiple stack positions?
I definitely don’t understand bury1 on slide 59. Is application of [] the identity function and you are just being clever to have harmony with the remaining definitions?
The one time I looked into these kinds of stack-based languages (before anyone ever heard of Forth), I called your production the draw and the consumption the drag (not great terms, I admit, because I have to “count on my fingers” to remember which is which). It seems to me you have them backwards according to the slide on p.40. I would think the left of the error is what is consumed, and the right of the arrow is what is produced. I started remembering that foray when I saw your lang.net presentation and I find it very interesting.
Oddly, I have abandoned stack-based languages and now prefer ones that do not reveal the mechanism. I have considered stack-based IL from compiled forms of expressions, but those are never intended to be visible to the programmer. At least so far in my thinking.
Thanks for being so prolific on this blog. It is very engaging for me and much easier than flaming you on whatever list you did not include on your slides!
Comment by orcmid — April 28, 2007 @ 9:41 pm