anfelor

7 ms·

1.
by anfelor 2y ago
Yes, that is true! Koka's constructor contexts also allow you to do this. A constructor context is a data structure with exactly one hole which you can fill in with an arbitrary value (or even several arbitrary values, copying the cons
2.
by anfelor 2y ago
Partially instantiated data structures are also available in Haskell (via Laziness), in OCaml (via tail modulo cons, https://inria.hal.science/hal-03146495/document ) and Koka (via constructor contexts, https:/&#x
3.
by anfelor 2y ago
Disclosure: I work on Koka's FBIP optimization (Option 4). > The most efficient data structure to use here would be a mutable dynamic array and in an imperative language that's what pretty much everyone would use. But if you as
4.
by anfelor 2y ago
Perhaps contrary to most people in this thread, I think you should avoid learning lenses or category theory too early. These are great tools, but they take months or even years to master and are not required to write useful code in the lang
5.
by anfelor 3y ago
Another interesting data structure related to skiplists but not mentioned here are zip trees: https://arxiv.org/abs/1806.06726 The are a tree-based version of skiplists and thus more suited to functional programming &#
6.
by anfelor 3y ago
You can use more external displays using a dock and DisplayLink Manager though. I got three extra monitors to run on my M1 Air with no problems that way.
7.
by anfelor 3y ago
Some of the research behind Koka was discussed at: FP2: Fully In-Place Functional Programming [pdf] - https://news.ycombinator.com/item?id=36471591 - July 2023 (24 comments) Perceus: Garbage Free Reference Counting with
8.
by anfelor 3y ago
The headline "GPT-4 increased BCG consultants’ performance by over 40%" is misleading since it implies that they became more productive in their actual work, when this is a carefully controlled study that separates tasks by an &qu
9.
by anfelor 3y ago
Sure, it would be great to chat! If you are interested in contributing to Koka at some point, there might also be opportunities there :)
10.
by anfelor 3y ago
These are two different languages, see https://effekt-lang.org/ and https://koka-lang.github.io/koka/doc/index.html -- although I believe that the FP^2 work mentioned by OP currently only works in
11.
by anfelor 3y ago
Thanks, I am surprised and delighted to see our paper here! We also wrote a follow-up recently, which you may enjoy: https://www.microsoft.com/en-us/research/uploads/prod/2023/0...
12.
by anfelor 3y ago
I would also recommend Koen Claessen's simplified finger trees ( https://dl.acm.org/doi/abs/10.1145/3406088.3409026 ) and Zip trees ( https://arxiv.org/pdf/1806.06726.pdf ) for purely f
13.
by anfelor 3y ago
On the contrary, I think this is a great example for how similar co-working spaces look. Even though this company apparently tried hard to make them look different, the pictures are so similar: - All the ceilings have exposed "industri
14.
by anfelor 4y ago
Intuition from category theory has certainly been useful for some programming language features (eg. algebraic effects, modal types or lenses), but I don't think it has helped with the things you cite. Rusts ownership model was created
15.
by anfelor 4y ago
Ah thanks! I didn't find that paper, is it online somewhere? You may also be interested in the stack allocation system in OCaml: https://www.youtube.com/watch?v=yGRn5ZIbEW8 In particular, Stephen gives a nice example i
16.
by anfelor 4y ago
Another question: In the section on eliminating heap-allocated captures, you discuss creating a datatype for captures that is passed by-value. But what if that datatype is recursive? For example, how would you compile a CPS-transformed `rev
17.
by anfelor 4y ago
Thanks for confirming, and that is a good point! However, I think that for monomorphizing compilers, programmers often write code in such a way as to avoid triggering a larger rec-compilation. With monomorphization this might be easier to a
18.
by anfelor 4y ago
Thank you for this cool blog post! > Each syntactic function (a lambda \x -> ...) gets a unique name > Determine the stack size of its captures based on the largest captures of any lambda in the lambda set it's involved
19.
by anfelor 4y ago
https://www.jasper.ai/
20.
From delimited continuations to algebraic effects in Haskell(blog.poisson.chat)
2 points by anfelor 4y ago|0 comments
21.
by anfelor 4y ago
This should have an (2011) in the title. luu, could you say a few words on what you found interesting about this thesis? The first chapter sounds impressive, but later it only seems to be defining a monad around typical operating system tas
22.
by anfelor 4y ago
Also, you can produce pretty horrific stuff by asking GPT to create a poem "celebrating", "glorifying" such things. You can also ask it to write a short story told by a general or Auschwitz guard etc. When doing so, you
23.
by anfelor 4y ago
Which programming language is best? There is no single programming language that is the "best" for all situations. Different programming languages are designed for different purposes, and the best language to use depends on th
24.
by anfelor 4y ago
The best attempt at creating a learnable programming language I've seen is this design document by Bret Victor: http://worrydream.com/#!/LearnableProgramming It is now ten years old but I don't know if it has
25.
by anfelor 4y ago
I am not sure the "Geeks, MOPs, and Sociopaths" model applies here since the most significant actors in this development came from the geek phase. SBF started out earning to give and was even featured on the homepage of 80.000 hou
26.
by anfelor 4y ago
No matter its technical merits, I would never use a programming language where I don't know the dominant (natural) language of the community. Every project has bugs and discussions how to work around them. This project has 12 Github is
27.
by anfelor 4y ago
Sure, they are in different language families. But a chart showing the "top dozen functional-programming languages" should include all languages that were given a separate workshop at the International Conference on Functional Pro
28.
by anfelor 4y ago
Even though this article comes from a reputable source, it should be pointed out that the author is not a researcher in the area -- and the decision not to include various MLs, OCaml, Scala, or F# in the chart of functional languages seems
29.
by anfelor 4y ago
There is also a nice new datastructure called Zip Trees [1] which is isomorphic to skip lists in the sense that it performs exactly the same comparisons on insertion/deletion (but it can skip some redundant ones). I would expect zip tr
30.
by anfelor 4y ago
Some counter-points: In Ocaml-world it is customary to write .mli files that specify the types of exported functions and modules. Those are then checked by the compiler against the .ml file with the implementation. In the .ml file you indee

More ›