James Wood

(2022-11-08) Programming Languages at Huawei. Sticking around for a PhD with @mspstrath, mechanising linearity. Banner from #四畳半神話大系. @[email protected]

[James Wood @laMudri]

Linear lambda memes for context constrained teens

[James Wood @laMudri]

@tldrjack Have you tried https://t.co/wRT0IoS5nx as a source of pronunciations? For a selection of 19 popular languages, you can search for words, and it'll give video clips from YouTube of native speakers of that language saying them.

[Andrej Bauer @andrejbauer]

@EgbertRijke @taooftypes @JDHamkins Uhm, the Bauer reals?

[James Wood @laMudri]

@mattecapu Presumably \mathit in the middle

[James Wood @laMudri]

@heades [email protected]; will follow when I'm back at the machine I'm logged in on!

[Sinéad Crowley @SineadCrowley]

Ok, I’ve joined #Mastodon but also this

[James Wood @laMudri]

@AndeolEvain @ARGleave Still, what's “pass-alive-territory”? If white keeps passing from the board state shown above, then black can eventually pick off everything.

[James Wood @laMudri]

@ProfMaxNew @c0b1w2 I guess the category of renamings is a cross-correspondence concept that fits here.

[James Wood @laMudri]

Next door got a new pet air compressor.

[James Wood @laMudri]

@anormalform @jfdm Why's it in the Idris 2 repository? Is it supposed to form part of the implementation at some point?

[James Wood @laMudri]

@ProfMaxNew are not logical/type-theoretic.

[James Wood @laMudri]

@ProfMaxNew Is there some next-level generalisation of limits and colimits that fits with what I said? I feel as if I know a “positive” or “negative” universal property when I see it (based on whether it's in the domain or the codomain), and there must be some universal properties which…

[James Wood @laMudri]

@ProfMaxNew Oh, you were talking about this in the other tweet. https://t.co/6rr69kbKiE

[Max ⊶ New @[email protected] @maxstewartnew]

My current view of Curry-Howard-Lambek:

Judgmental structure = (very) generalized multicategory
Connective = universal property
Logical relation = (multi-/bi-) fibration

[James Wood @laMudri]

@ProfMaxNew What's the implicit claim of the Curry-Howard-Lambek correspondence, anyway? That limits and colimits correspond to negative and positive type formers, respectively?

[James Wood @laMudri]

@anqurvanillapy Is it also freely generated from reflexivity?

[James Wood @laMudri]

@andrejbauer @EgbertRijke @wtgowers @panlepan @mathforge @achambertloir @johncarlosbaez @noamzoam An incident that put this discussion into perspective for me recently was witnessing someone using the phrase “reductio ad absurdum” to mean a silly argument.

[Antonello Guerrera @antoguerrera]

This video is an absolute gem. ❤️🇬🇧

In 1967 John Reith, the first ever Director General of the BBC, spoke to Malcolm Muggeridge about the "BBC accent". - via @BBCArchive

[James Wood @laMudri]

@ProfMaxNew You mean automating the encode-decode proofs? Last I heard, it was in the “should, but doesn't yet, exist” category.

[James Wood @laMudri]

Naming things is one of the great hard problems in computer science, but placing things can be pretty difficult too.

[James Wood @laMudri]

More of this trivia: When was the last time the monarch appointed a new prime minister before their coronation?

[James Wood @laMudri]

Before last week, when was the last time the monarch and the prime minister had the same name?

[James Wood @laMudri]

@jonmsterling @i2talics @boarders__ @ProfMaxNew I've used `Wrap` in my work just because I couldn't be bothered to write out a verbose record declaration and have to think about overloading, but it should be ideal for univalent mathematics, where you benefit from the uniformity given by not introducing new record types.

[James Wood @laMudri]

@jonmsterling @i2talics @boarders__ @ProfMaxNew I introduced Data.Wrap to stdlib as an attempt to make this style compatible with unification-based resolution. https://t.co/iJS9CIvtji