Arnaud Spiwack

@aspiwack

Multi-classed Software Engineer/Constructive Mathematician. I torture types for a living.

Vrijeme pridruživanja: ožujak 2019.

Tweetovi

Blokirali ste korisnika/cu @aspiwack

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @aspiwack

  1. proslijedio/la je Tweet
    22. sij

    libgen/scihub are of such monumental social/moral worth that their illegality delegitimizes the law as a whole

    Prikaži ovu nit
    Poništi
  2. 2. velj

    In fact, I believe that French had very few words of Greek origin prior to this.

    Prikaži ovu nit
    Poništi
  3. 2. velj

    This long-winded story was only to get me to a point: when a French-speaker sees the ‘ch’ digram, they instinctively read it ʃ (like ‘sh’ in English). It reads k mostly in words imported from Greek at this very time (or since!). (as pointed out by )

    Prikaži ovu nit
    Poništi
  4. 2. velj

    Nevertheless these fellows (keyword: La pléiade) had a tremendous influence on the French language.

    Prikaži ovu nit
    Poništi
  5. 2. velj

    The plan kind of failed, because the new word rarely displaced the original word, rather took a different, more specialised meaning (“ausculter” comes to mind: it was meant to replace “écouter”, but it now only pertains to using a stethoscope).

    Prikaži ovu nit
    Poništi
  6. 2. velj

    Somehow, we kept the two forms for slightly different usages. I will decline all responsibilities.

    Prikaži ovu nit
    Poništi
  7. 2. velj

    This is how French got stuck with “credible” for “believable” but “incroyable” for “incredible”. The former has travelled through centuries of language shaping, the latter jumped directly from ancient text books into modern French.

    Prikaži ovu nit
    Poništi
  8. 2. velj

    I don't know. Anyway. They set to rebuild the French language, to make it more betterer. To that effect they mostly decided to ignore 1000 year of French morphology and re-import words from Latin and Greek (antiquity was cool in the 16th Century, much like the 1970s are now).

    Prikaži ovu nit
    Poništi
  9. 2. velj

    [Caveat emptor: I am not a medical doctor, this and the rest of this thread should not be construed as a medical diagnosis or advice]

    Prikaži ovu nit
    Poništi
  10. 2. velj

    There was a bunch of jolly chaps in the 1500s who decided that French was butt ugly. I think it was a mix of inferiority complex (Renaissance had started in Italy), and teenager syndrom (Renaissance is not like the Middle Ages, you don't understand anything Daaaad).

    Prikaži ovu nit
    Poništi
  11. 2. velj

    To be fair, French is weird. Like, very weird. Except weirder. I mean, I realise that your native language is English, but humour me.

    Prikaži ovu nit
    Poništi
  12. 31. sij

    I'll be streaming a game of Zelda: a link to the past, randomizer tonight. I'll be trying a logic with glitches taken into account. And chatting about combination locks. Stream starts at 20.45 UTC

    Poništi
  13. 31. sij

    Only it hadn't been connected to the data/control intuition before. It's a cute little story. The strength requirement is often forgotten because of how all monads are strong (aka enriched) in Hask.

    Prikaži ovu nit
    Poništi
  14. 31. sij

    This do-notation business has been known since before the monads were a thing in Haskell: Moggi needed *strong* monads to give semantics to effects. In a self-enriched category, strong monads and enriched monads are the same.

    Prikaži ovu nit
    Poništi
  15. 31. sij

    I didn't speak about `Traversable` at all, but traversable functors do belong in the data hierarchy (I don't believe there is any value in coming up with a control equivalent of traversables but I'd be thrilled to be wrong!)

    Prikaži ovu nit
    Poništi
  16. 31. sij

    This also proves, incidentally, that data applicative are not a super-class of data monads since lists are the latter, but not the former (control applicative, however, are a super-class of control monads).

    Prikaži ovu nit
    Poništi
  17. 31. sij

    With linear types, lists form a data monad, but not a control monad. This is witnessed by the fact that the do-notation, on lists, lets me make cartesian products, which are definitely not linear.

    Prikaži ovu nit
    Poništi
  18. 31. sij

    This is why I don't speak about data monads too much in the blog post: they are less evidently useful. You can still use them to model substitution (e.g. in first-order terms).

    Prikaži ovu nit
    Poništi
  19. 31. sij

    The distinction is about what I can do with a functor, rather than what it is. That's why individual functors can be control or not depending on the situation.

    Prikaži ovu nit
    Poništi
  20. 31. sij

    What distinguishes data and control, essentially, is that you can use the do-notation with control monads, but not with data monads.

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·