Comments for carlo angiuli (blog)

Comment on Progress report by Mathias

Mathias — Mon, 04 Jul 2011 01:20:42 +0000

As a comp.sci. student (Aarhus University) who just did a Programming Languages course featuring Scheme primarily, it’s safe to say I’m looking forward to some math/prog lang theory blogging!

Comment on My computerweb by Gentoo

Gentoo — Fri, 18 Feb 2011 05:17:38 +0000

You scrub, real men use Gentoo.

Comment on Mutable pairs in Scheme: a brainteaser by Dustin

Dustin — Thu, 03 Feb 2011 18:21:45 +0000

Very nice post and comments, thanks for sending me this Carlo. In regards to the diagrams, I haven't used it in a while, but you can make similar stuff with Ditaa. But you have to type everything out and that takes time and yea....

Comment on Mutable pairs in Scheme: a brainteaser by Lindsey

Lindsey — Wed, 02 Feb 2011 21:00:16 +0000

I use OmniGraffle for this sort of thing. It's Mac-only and costs money, but I like it enough that I'm willing to put up with that. And, yep, what I said was exactly what Greg said. (His comment wasn't there yet when I started mine! That's the downside of taking 20 minutes to make a dumb diagram...)

Comment on Mutable pairs in Scheme: a brainteaser by Carlo

Carlo — Wed, 02 Feb 2011 20:01:29 +0000

Greg: It does! I believe Lindsey’s diagram says the same thing. Re: mutations, I’m in a CS department that’s huge on Scheme, and I’ve found that occasional (!) use of mutation is actually useful. Besides, I’m in a compilers class this semester, and we’re writing a Scheme compiler from the bottom up. Right now, our input language is basically a collection of set! operations, so I’m a bit desensitized to it anyway.

Comment on Mutable pairs in Scheme: a brainteaser by Carlo

Carlo — Wed, 02 Feb 2011 19:55:23 +0000

Lindsey: I drew that same diagram to figure it out, except it was much messier. Thanks for putting it up! How did you draw those nice cons cells and pointers?

Comment on Mutable pairs in Scheme: a brainteaser by Carlo

Carlo — Wed, 02 Feb 2011 19:53:43 +0000

Adam: Good point; you would need to use Racket’s mutable pairs, of course. Racket isn’t exactly a Scheme anymore, it seems.

Comment on Mutable pairs in Scheme: a brainteaser by Lindsey Kuper

Lindsey Kuper — Wed, 02 Feb 2011 18:50:06 +0000

Oh, yeah, that’s a fun one! It’s because when we create ls2, the cdr of ls2 is just a pointer to the head of ls, so if we mutate ls, ls2 can’t help but know about it. But if we change the cdr of ls2 to be something else, there’s no longer any connection between ls and ls2. I had to draw a diagram to see it: http://i266.photobucket.com/albums/ii256/lindseykuper/mutable-pairs-brainteaser.jpg

Comment on Mutable pairs in Scheme: a brainteaser by Greg

Greg — Wed, 02 Feb 2011 18:45:35 +0000

I find this stuff difficult to explain without using diagrams, but this makes sense to me. In both cases, the cdr of ls2 points to ls. When you mutate the cdr of ls in part 1, it shows up in ls2 because ls2 is pointing to ls. But in part 2, you’re actually changing where the cdr of ls2 points to: it’s no longer referencing ls, but instead is pointing to the new list ‘(0) that you just created. So ls2 changes, but ls stays the same.

Does that make sense? It’s been a little while since I’ve Schemed. (And even when I did, mutating stuff was a big no-no.)

Comment on Mutable pairs in Scheme: a brainteaser by Adam

Adam — Wed, 02 Feb 2011 18:29:34 +0000

I’m pretty sure mutable pairs are disabled by default in Racket, so it is important you’re using Chez.

> (define p (cons 0 1))
> (set-cdr! p p)
> p