brendanzab.bsky.social
World Builder, Artist, Programmer. Thinking about programming languages and imaginary landscapes. he/him
- https://types.pl/@brendan (preferred)
- https://github.com/brendanzab
116 posts
159 followers
224 following
Regular Contributor
Active Commenter
comment in response to
post
Has mttd posted recently? š
comment in response to
post
Ah yeah, probably a bit of a stretch now re-reading the original post. Was thinking about how you could play around with logic and type theory stuff in a way that might not be sound or correct(?), but there was still a ton of value in that.
comment in response to
post
Iāve found Holbert fun for experimenting/learning about type theory stuff. Not sure how similar it is to what youāre imagining though!
comment in response to
post
I stupidly cancelled my preorder a number of years ago when it was much cheaper. Curious to see how you like it!
comment in response to
post
Oooh, it was my first time there! I think we picked a good time to wander around it. Sadly the tower was currently under renovation, so we werenāt able to see the views from there, which I assume are spectacular.
comment in response to
post
Iāve seen a lot of this in the Haskell world too. š©
comment in response to
post
You can read more about the details of the specification language here: github.com/Wasm-DSL/spe..., and see how it is used to define the actual WASM specification here: github.com/Wasm-DSL/spe...
comment in response to
post
The biggest issue for me is that they donāt seem to have a dedicated forum space, and so it can be a roll of the dice whether your question is answered on Matrix before the conversation moves on.
comment in response to
post
cppnix seems pretty abandoned at this point, from my understanding. I donāt want to give determinate nix the light of day, and have not had much in the way of issues with lix personally.
comment in response to
post
As somebody who ditched Twitter a couple of years ago, I eventually just accepted letting the hangers-on go, and it wasnāt as terrible as Iād anticipated! If theyāre worth following, theyāll eventually move to a different platform.
comment in response to
post
There's some good documentation here doc.rust-lang.org/std/ops/trai..., if it helps at all. But yeah, while thereās a lot of temptation to use it for subtyping-style coercions, but Iād be wary of it.
comment in response to
post
Iād recommend only using deref for smart-pointer-like things.
comment in response to
post
I agree that dimensional analysis is beyond many type systemsā¦ but what if?? š¤
(no need to wait, Kennedy has you covered www.cl.cam.ac.uk/techreports/...)
comment in response to
post
Andrew Kennedy might like to have a word š¤©
comment in response to
post
Chrome is making it harder to implement ad blockers. Iād probably recommend switching to Firefox tbh (though I realise it can be a pain).
comment in response to
post
ann-zee
comment in response to
post
I thought it affected if you were relying on the order of nominal variants though (at least it did in my experience). TBH I need to check with records, but Iām pretty sure it checks them too.
comment in response to
post
I just add lots of type annotations anyway and itās not a _massive_ issue for me.
comment in response to
post
If you add explicit type annotations it can figure out which one you want. But yeah it's not great. Thereās a compiler flag, `-principle` that will give an error if you are depending on such behavior, but IIUC it cause a performance hit to type checking for some reason.
comment in response to
post
Damn, I don't have something to play this on, but it looks so great!
comment in response to
post
Love this so much! š¤©
comment in response to
post
I think others recommend uniplate or multiplate or whatever in Haskell, but I haven't yet figured out how to translate that to OCaml. I also don't really understand the performance tradeoffs of the plated approaches.
comment in response to
post
The tricky thing with recursion schemes often comes when you get mutually recursive typesā¦ and also all the names are incredibly obscure. Iād love to see somebody come up with a nice surface syntax for them so I didn't have to remember all of them, and figure them out later.
comment in response to
post
Yeah, a PPX might be good for automation, but I think you could just hard code some of them if you like in the interim. Here's an example of a simplified version: gist.github.com/brendanzab/7...
comment in response to
post
I approve :3
comment in response to
post
I wonder if one could implement message passing stuff with effects and handlers š¤
comment in response to
post
Yeah! OCamlās row polymorphic object system is actually quite interesting, even if the syntax is incredibly goofy looking. Some more resources and background:
- dev.realworldocaml.org/objects.html
- dev.realworldocaml.org/classes.html
- ocaml.org/docs/objects
- caml.inria.fr/pub/papers/r...
comment in response to
post
I get a similar feeling whenever I see RCT screenshots :)
comment in response to
post
One way is recursion schemes, which involves separating out the recursive layer into a new type. They come with their own challenges though and are not for the feint of heart, though perhaps you could glean some inspiration from them.
comment in response to
post
Silly, weird, odd, strange, goofy, wild, ridonkulousā¦ depends on the context though.
comment in response to
post
- OāHaskell web.archive.org/web/20090517...
- āHaskellās overlooked object systemā arxiv.org/abs/cs/0509027
- āObject Oriented Programming in Haskellā www.well-typed.com/blog/2018/03...
comment in response to
post
I'm assuming itās a joke š
comment in response to
post
Not sure if that is related at all to your point, but it made me think at any rate! Thank you!
comment in response to
post
Now Iām pondering this in relation to programming language designā¦ in my communities I often see feedback about that seems overly focused on adoption and commercial use, where as I think sometimes it pays to first inquire about the context, and what the creator is intending exploring in their work.
comment in response to
post
I looove stuff like thisā¦ itās so cool what you can do with a minimal palette of things! Very inspiring!
comment in response to
post
Feed readers are great! I tend to read the posts on the site as opposed to in the feed reader ā I mainly like getting a list of new posts every day without having to doomscroll or check a link aggregator, potentially missing something cool from somebody Iām a fan of. :)