Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Jan 18 08:58
    brendanzab commented #242
  • Jan 18 07:12
    brson opened #242
  • Jan 17 00:49

    brendanzab on main

    Remove lifts from specification Cleanup specification of lexica… Elide explicit lifetimes and 1 more (compare)

  • Jan 17 00:49
    brendanzab closed #241
  • Jan 17 00:43
    brendanzab synchronize #241
  • Jan 17 00:33
    brendanzab synchronize #241
  • Jan 17 00:31
    brendanzab synchronize #241
  • Jan 17 00:05
    brendanzab opened #241
  • Jan 15 09:44
    nguyenphuminh starred pikelet-lang/pikelet
  • Jan 09 12:21
    Michaeleiner starred pikelet-lang/pikelet
  • Jan 06 23:44
    mbugrayildiz starred pikelet-lang/pikelet
  • Jan 02 13:50
    Superstar64 starred pikelet-lang/pikelet
  • Jan 02 02:16

    brendanzab on main

    Remove universe levels for now … Merge pull request #240 from br… (compare)

  • Jan 02 02:16
    brendanzab closed #240
  • Dec 30 2020 06:20
    brendanzab edited #240
  • Dec 30 2020 01:43
    brendanzab opened #240
  • Dec 21 2020 14:40

    brendanzab on main

    Rework local variable API and i… Remove local_levels vector in e… Merge pull request #239 from br… (compare)

  • Dec 21 2020 14:40
    brendanzab closed #239
  • Dec 21 2020 10:29
    brendanzab synchronize #239
  • Dec 21 2020 08:55
    brendanzab synchronize #239
matrixbot
@matrixbot
suhr segeljakt: I have to figure out a mysterious lalrpop error.
suhr The problem with lalrpop freeze can be fixed by disabling lane table algorithm.
matrixbot
@matrixbot
segeljakt SQL seems like hell to parse
segeljakt "Structured" Query Language my a$* 😂
matrixbot
@matrixbot
brendanzab This was pretty cool: https://www.youtube.com/watch?v=vK_cg8UP6gU
brendanzab * This was pretty cool - using multi-stage programming to improve the performance of datatype generic programming: https://www.youtube.com/watch?v=vK_cg8UP6gU
brendanzab * This was pretty cool - using multi-stage programming to eliminate the runtime overhead of datatype generic programming: https://www.youtube.com/watch?v=vK_cg8UP6gU
matrixbot
@matrixbot
suhr SQL is a COBOL-style language.
suhr Kinda strange, it's like a young dinosaur that still lives today.
matrixbot
@matrixbot
brendanzab well, got the range information into the core language
brendanzab now looking at trying to do the '{ x = 1, y = x } : { x : I32, y : I32 }` thing 😅
brendanzab * now looking at trying to do the { x = 1, y = x } : { x : I32, y : I32 } thing 😅
matrixbot
@matrixbot
brendanzab might have to roll back my thing where you can construct records with out of order fields
matrixbot
@matrixbot
brendanzab Kind of wish I could have something like Tuple [ I32, String ] == { 0 : I32, 1 : I32 }
brendanzab And Array 3 I32 == { 0 : I32, 1 : I32, 2 : I32 }
brendanzab also [ 1, 2, 3 ] == { 0 = 1, 1 = 2, 2 = 3 } : Array 3 I32
brendanzab and [ 23, "hello" ] == { 0 = 23, 1 = "hello" } : Tuple [ I32, String ]
matrixbot
@matrixbot
brendanzab re. the unordered construction - they issue is that I have a btreemap for my record terms
brendanzab * re. the unordered construction - the issue is that I have a btreemap for my record terms
brendanzab that lets me have { y = 23, x = 1 } : { x : S32, y : S32 }
brendanzab because when I type check the term I can pick off entries at my leisure
brendanzab but that might be getting in the way of other experiements
matrixbot
@matrixbot
brendanzab ie. in the way of making record construction more like let expressions
brendanzab sorry I realise it might have been confusing that I mentioned that then started talking about another form of ordering - ie. tuples and arrays
matrixbot
@matrixbot
ratmice Yeah, so the way i've been thinking about it is ';' vs ',' where (a, b) specifies the canonical form and ';' canonicalizes based on some convenience like alphabetical order to binding name (or anything), So you get things like [a, b] and [a; b] being like the difference between array, and set
ratmice It's a bit weird with records, because then "," with records implies some kind of unbalanced raw tree structure or something.
ratmice Anyhow that might be a way to salvage your existing ordering
brendanzab interesting - have you seen how logic languages use ; for 'and' and , for 'or'?
brendanzab eg. Prolog, Mercury, Makam
ratmice no, I have a kind of weird prolog phobia, something to do with backtracking 😭
matrixbot
@matrixbot
brendanzab hehehe
brendanzab I like the logic languages where you get to say whether you backtrack or not
brendanzab like mercury
brendanzab but yeah I have pondered using ; and , for or some other symbols like that for types because of that
brendanzab also been pondering the 'key set' of a record as kind of a finite set or enumeration 🤔
matrixbot
@matrixbot
ratmice Yeah, Ur/web is neat it basically allows you to fold over row polymorphic record keys, and like concatenate them if they are disjoint
brendanzab need to attempt to read the 'gentle art of levitation' again
matrixbot
@matrixbot
ratmice Been reading through the compositional compiler correctness things here.
brendanzab and yeah +1 to Ur/web
brendanzab nice!
brendanzab the next 700 compiler correctness theorems
brendanzab * the next 700 compiler correctness theorems?
ratmice Yeah, and same authors earlier paper "FabULous Interoperability for ML and a Linear Language"
brendanzab "FabULous" lmoa
matrixbot
@matrixbot
segeljakt do you know any good resources on declarative programming?
matrixbot
@matrixbot
segeljakt btw, have you seen this language? https://www.minizinc.org/
brendanzab yeah I've heard of it!
brendanzab never played around with it though
matrixbot
@matrixbot
brendanzab Was messing around with this today :P pikelet-lang/pikelet#222
brendanzab dunno if it is worth it though haha