Filippo Sestini

🇮🇹 PhD in Computer Science (type theory). Functional programmer (Haskell/OCaml/Agda). Software Engineer @Arm

[Filippo Sestini @FilippoSestini]

@TheEduardoRFS @KleeneAlgebra What is the repeating state in

x = 1; while (x != 0) { x++ }

assuming no overflows

[Filippo Sestini @FilippoSestini]

@_Felipe @dcolascione @joseph_h_garvin How is OCaml any different from Haskell, when it comes to currying by default?

[Filippo Sestini @FilippoSestini]

@TravisMWhitaker @ZuriHac That’s the plan

[Graham Hutton @haskellhutt]

Just a few days left now to register for the Midlands Graduate School (MGS)! Eight fantastic courses on category theory, type theory, coalgebra, semantics and more. 7-11 April 2025, Sheffield, UK. Registration closes Monday 24th March. Please share! https://t.co/g2x6pXwvVo

[Filippo Sestini @FilippoSestini]

@lillybilly299 What learning material are you using? Books and tutorials for beginners are designed to explain things in small digestible steps, so that you don't feel overwhelmed. They certainly don't teach you to copy paste (I mean, what would you be even copying, and from where..)

[Filippo Sestini @FilippoSestini]

@uwukko OCR.

[Satnam Singh @satnam6502]

Graham Hutton @haskellhutt, my PhD brother, has had an admirable academic career doing exactly what he wants to do, pursuing research to the highest quality without bending to the gravitational pull of grant body hot topics and the pressure to grow large long running research projects. His razor sharp focus on developing surprisingly simple expositions of quite complicated concepts is a rare and wonderful skill, very much in the style of Bird. His dedication to teaching is utterly inspiring and he says what I say: "Don't tell my employer, but I would do this for nothing." I worry that this kind of academic career with immense amounts of freedom that he has had is becoming harder to achieve for the next generation. I hope not, the world needs more Huttons. Yet another great @HaskelInterlude hosted by @kosmikus and @wouterswierstra.
https://t.co/ksF6qZkLdL

[Filippo Sestini @FilippoSestini]

@OliverJia1014 @EmperorQuintana It would seem you, quite ironically, missed the sarcasm of that post.

[Filippo Sestini @FilippoSestini]

@VictorTaelin @SirProdigle Japanese animation as a medium is extremely diverse and multidimensional actually. The fact that it just happens to not explore musical theater-like expression that much hardly makes it "unidimensional".

[Ben Gamari @bgamari]

GHC 9.10.1-alpha1 is now available!

Lots of goodness in this release including a few great steps on the road to Dependent Haskell, the long-awaited introduction of exception backtraces, and Javascript FFI support in the Wasm backend.

https://t.co/KMPKn7dMZB

[Computer Science, Uni of Nottingham @UoNComputerSci]

PhD Studentship: Multiple opportunities within School of Computer Science!

Applications are invited from Home and International students for up to 10 fully-funded PhD studentships offered by the School of Computer Science starting on 1st October 2024.

https://t.co/2XrRj2eXTm

[Filippo Sestini @FilippoSestini]

@krismicinski @tritlo Oh, likewise, I always appreciate the opportunity to have an interesting conversation and think things out loud!

[Filippo Sestini @FilippoSestini]

@krismicinski @tritlo Your OP presents two statements that cannot be obtained from each other via CH. One is about which terms of a type can be defined with different schemes, the other is about which types are inhabited (regardless of the inhabitants). These are different concepts even under CH.

[Filippo Sestini @FilippoSestini]

@krismicinski @tritlo In any case, I don't see how C-H brings insight here. The comparison "primitive vs general recursion" is a statement about how many distinct functions (i.e. proofs) of *the same type* (i.e. proposition) can be defined. It's not a statement about how one schema proves more thrms.

[Filippo Sestini @FilippoSestini]

@krismicinski @tritlo Yeah, I think it's hard to follow without more precise definitions or examples. I don't know what you mean by "the obvious induction principles", for example. And in sufficiently extensional systems, like set theory or ETT, most of these schemes are provably equivalent anyway.

[Filippo Sestini @FilippoSestini]

@Krever01 @wollantine If we are talking about using classes as a module system, I don't even disagree. They are pretty convenient for that.

[Filippo Sestini @FilippoSestini]

@Krever01 @wollantine What's the problem with typeclasses being visible per function? Not sure I understand what you mean by "dependencies are noise and they break abstraction", and it also sounds like an issue that is orthogonal to the OOP/FP discussion.

[Filippo Sestini @FilippoSestini]

@Krever01 @wollantine Interfaces are great, but it's not like pure FP languages don't have similar facilities: things like typeclasses, modules, and abstract types are all about programming against interfaces. Also, literally nobody uses "records of functions" in pure FP as an established pattern.

[Filippo Sestini @FilippoSestini]

@Krever01 @wollantine I'm curious about what "FP inside OOP classes" really buys you vs "real" FP. If you don't use objects to encapsulate mutable state (and I assume you don't, otherwise it's not FP), then your OOP classes are basically a glorified module system as you'd find in any pure FP lang.