a shower thought

the SICP uses numbers to illustrate linear processes and tree shaped processes

it calculates lots of math functions: Fibonacci, square roots, lenghty sums, lenghty products...

But operating on numbers is easier because more often than not, operations on numbers are commutative

(+ 2 3) and (+ 3 2) yield the same result

If you operate linearly on lists, the order in which you carry on your operations matters

on tree structures it's even worse

(cons 2 3) and (cons 3 2) yield DIFFERENT results

the recursive version of append does the right thing

the iterative version reverses one of the lists

the order in which you perform the cons operations does matter !

so they are NOT equivalent

these things don't happen (as often) when operating on numbers

I'm wondering if doubly linked lists could be functional (requiring no mutation in place)

does the r7rs mention any continuation (delimited or not) ?

I remember reading a blog post about how the use of language/compiler extensions in was making landing of newbies more difficult and so damaging the haskell world

I'd love to find it again

This procedure does work

(define (my-append a-list b-list)
(fold-right cons* '() a-list b-list))

And it also works with more than 2 lists

Now I'm confused

I need some time to figure out why

in procedure.source always returns

I defined a procedure in a file, I loaded the file, it compiled the file


scheme@(guile-user)> (procedure-source mah)
$1 =


you were right, my video about continuations is about vanilla continuations

Your video is about delimited continuations instead

I started watching it and I noted some code, here


The last procedure is discussed at about 11:56

This little thing here

My god

(define (five)
(display "First\n")
(shift k
(display "Second\n")
(k 5)
(display "Fourth\n")
(display "\nThird\n"))


naming is hard

and the scheme community is particularly bad at that

See SRFIs, for example and "ice-9" modules

I mean, really ?

I started to review this video in

It's the only content I know about conts that is expressed n a human language

I highly recommend it, it does provide huge a ha moments

It's easier to process delimited conts when you have regular conts settled



I'd like to have a window with a grind in it and a game of life running in the grid

A visual representation of a game of life, that is

Do you think Chickadee could be fit for such a thing ?

some years ago I run into a paper about a new kind of user interaction in a terminal implemented with in

I'd love to find it again

Show older
Functional Café

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!