00:00 <aeth> s/successfuly has/successfully have/

00:01 <ZombieChicken> or it could be that programmers (especially younger ones) are very into trends, and a 50+ year old language family isn't 'hip' anymore.

00:01 <ZombieChicken> At least in this case Haskell is a somewhat interesting language

00:02 <aeth> ZombieChicken: one of the hippest trends is static type systems

00:03 <ZombieChicken> like in C, right?

00:03 <aeth> ZombieChicken: you can see this in e.g. the JS community. 10 years ago, CoffeeScript was the thing people used to compile to JavaScript if they wanted a decent language. Now, they use TypeScript.

00:03 <aeth> ZombieChicken: well, C doesn't have a particularly good static type system because you can break anything by casting it to void and other tricks

00:04 <ZombieChicken> Well, yeah. It's a weak system to be sure

00:04 <aeth> think more along the lines of Rust or TypeScript as what's currently trendy. Or probably Typed Racket.

00:04 <ZombieChicken> I havn't touched any of those

00:04 <ZombieChicken> but thanks for the chuckle; compiling something else to JS and thinking your product is going to be good is funny

00:05 <aeth> ZombieChicken: If you want to see what all of the 23 year olds in San Francisco are coding with on their Macbook Pros to write yet another piece of adtech that no one outside of San Francisco wants, just read HN to see those trends in tech.

00:05 <ZombieChicken> I get an RSS for HN

00:05 <ZombieChicken> Sometimes useful things pop up

00:06 <aeth> ZombieChicken: And TypeScript is, well, not particularly good imo because it's still stuck with the primitive types that JS offers like "number" (i.e. double float). Compare this to the numeric tower that Lispers expect.

00:06 <ZombieChicken> I never assume anything will be like a sane Lisp

00:08 <aeth> The first thing I look at in a language is how it expresses its numbers. There are different ways to do it that all have their tradeoffs, of course.

00:08 <aeth> I'm sure for a lot of people strings are more important, though.

00:09 <aeth> JS JITs afaik handle strings very elaborately.

00:11 <aeth> Afaik, Lisps kind of have a luxury to not focus on strings quite as much as some dynamic languages because symbols take the role that strings would in some places.

00:16 <bremner> lots of languages intern strings. It's not that different from scheme symbols

00:16 <bremner> it's more or less an implementation detail. Arguably the seperate syntax is a bit unfortunate

00:17 <bremner> don't @ me :P

00:25 <aeth> bremner: I agree with the first part, I disagree with the separate syntax bit being unfortunate. Just because a langauge is dynamically typed doesn't mean it can't have, well, types. Some languages (I'm not talking about JS here, but I'm not naming names because this is a channel about Racket) oversimplify their types too much and rely too much on magic for efficiency imo.

00:26 <aeth> I mean, I guess I can say Tcl because it should be fairly uncontroversial that Tcl is an "everything is a string" language, making it the most extreme here.

00:28 <aeth> There have been a decent amount of attempts (usually on top of Racket because that's the sort of thing that Racket's for) to do something Tcl-like, but with lists, i.e. everything as lists. e.g. Strings and even sometimes numbers. These haven't taken off.

00:29 <aeth> I guess arguably if symbols are strings and strings are character lists, it's in some way "Lispier"

00:50 <ZombieChicken> Wasn't the orignal Lisp a 'Everything is a CONS cell' language?

03:16 <erkin> Yes, but there exist Lisps whose lists aren't based on cons cells. (Like Clojure)

03:16 <erkin> And Lisps without sexps altogether. (Like Logo and Dylan)

03:40 <johnjay> samth: ah, i was in classes with some of the authors of that paper

03:40 <johnjay> thanks!

03:41 <samth> vraid: mostly the answer is slack

15:10 <kritixilithos> huh in the repl (read) takes input from the next line, but (read-line) starts before the next line starts, ie (read-line)this is input