2025/11/11

Newest at the top

2025-11-11 03:50:21 +0100 <geekosaur> typeclasses
2025-11-11 03:50:18 +0100 <jreicher> Took me ages to unlearn that expectation
2025-11-11 03:50:14 +0100 <geekosaur> and patsyns
2025-11-11 03:50:11 +0100 <jreicher> The thing that really broke my brain when I looked at Lisp (after doing functional) is that there isn't an implicit eval of an expression in head possition.
2025-11-11 03:49:46 +0100 <EvanR> up to three
2025-11-11 03:49:33 +0100 <EvanR> shoot
2025-11-11 03:49:32 +0100 <geekosaur> module name/qualifier
2025-11-11 03:49:14 +0100 <EvanR> every [capitalized] name has two independent bindings, type and constructor
2025-11-11 03:48:50 +0100 <monochrom> Criminals.
2025-11-11 03:48:17 +0100 <jreicher> That's Lisp-2. Lisp-1 is not like that
2025-11-11 03:48:16 +0100 <monochrom> BASIC: Every name has two independent bindings: number and string.
2025-11-11 03:48:03 +0100 <monochrom> Lisp: Every name has two independent bindings: value and function.
2025-11-11 03:47:39 +0100 <monochrom> No no no, the original sin belonged to BASIC and Lisp. Yes I'm putting them on the same line.
2025-11-11 03:47:37 +0100 <geekosaur> awk doesn't have arrays/lists, so you use its "associative arrays" as if they were
2025-11-11 03:47:19 +0100 <jreicher> Personally I'm a fan of that bad boy.
2025-11-11 03:47:07 +0100 <geekosaur> yes
2025-11-11 03:47:00 +0100 <jreicher> I thought the original sin belonged to awk?
2025-11-11 03:46:11 +0100 <geekosaur> you would not normally use a hash as a list unless it's sparse
2025-11-11 03:45:15 +0100 <geekosaur> there are three types: scalar, list, hash
2025-11-11 03:45:13 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 256 seconds)
2025-11-11 03:45:02 +0100 <geekosaur> you're still stuck somewhere insane
2025-11-11 03:44:31 +0100 <monochrom> What have I done?! >:)
2025-11-11 03:44:20 +0100 <EvanR> even better
2025-11-11 03:44:09 +0100 <EvanR> oh list is a hashtable?
2025-11-11 03:43:32 +0100humasect(~humasect@dyn-192-249-132-90.nexicom.net) humasect
2025-11-11 03:42:42 +0100 <geekosaur> they called it list
2025-11-11 03:42:31 +0100 <geekosaur> perl didn't call it array, @x is not %x
2025-11-11 03:41:38 +0100 <EvanR> so perl and PHP were justified in calling their hashtable "array" xD
2025-11-11 03:40:29 +0100monochromis looking at Python and Perl.
2025-11-11 03:40:25 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-11 03:40:10 +0100 <monochrom> Hell, s/array/hash table/g
2025-11-11 03:39:49 +0100 <EvanR> even in normal programming, you rarely care about what an array "really is", hence the big O 1 array myth
2025-11-11 03:38:58 +0100 <EvanR> even in "normal math" you really rarely care about what a function "really is"
2025-11-11 03:37:56 +0100 <EvanR> instead of the implementation
2025-11-11 03:37:42 +0100 <EvanR> the structural set theory with sets elements and relations gets closer to type theory by emphasizing the domain and codomain of functions (relations) as the important part
2025-11-11 03:37:27 +0100 <jreicher> I'm tempted to say that's because the type theory doesn't need to know the details... ;)
2025-11-11 03:36:35 +0100 <monochrom> Oh, type theory says functions are primitive. :)
2025-11-11 03:36:31 +0100 <EvanR> which gives haskell room to call functions functions
2025-11-11 03:36:21 +0100 <jreicher> EvanR: true, but I think even for relations I've seen more set-theoretic presentations than not.
2025-11-11 03:35:58 +0100 <EvanR> it could even be abstractified using category theory
2025-11-11 03:35:45 +0100 <jreicher> Leary: yes, thank you.
2025-11-11 03:35:43 +0100Googulator53(~Googulato@2a01-036d-0106-0180-8127-ba79-55a7-6f29.pool6.digikabel.hu) (Quit: Client closed)
2025-11-11 03:35:42 +0100 <EvanR> e.g. it could be implemented as a relation plus the functional condition, where relations are primitive
2025-11-11 03:35:41 +0100Googulator92(~Googulato@2a01-036d-0106-0180-8127-ba79-55a7-6f29.pool6.digikabel.hu)
2025-11-11 03:35:14 +0100 <jreicher> monochrom: yeah, that's what I expected. Can't shake the feeling it deserves explicit discussion, but maybe not.
2025-11-11 03:35:13 +0100 <Leary> jreicher: You might be interested in my (as yet unpublished) https://github.com/LSLeary/native-cont library; it provides a safe non-IO interface to the primops.
2025-11-11 03:34:36 +0100 <EvanR> jreicher, a function is not always implemented as a set in math
2025-11-11 03:33:48 +0100 <monochrom> Captured in the middle. The control searches for the matching prompt, captures everything between now and that.
2025-11-11 03:33:25 +0100 <Leary> Of course, it's captured.
2025-11-11 03:32:56 +0100 <jreicher> Alexis