Richard Eisenberg

I'm excited to be part of a panel on careers in PL at PLMW at @poplconf. But I'm weighing whether or not to stick around for all of POPL. I'd love to -- but I went 2 for 2 in 2022 for getting covid at conferences. Will there be e.g. a testing protocol? That would go a long way.

And speaking of unboxed types... check out my talk next week at the ML Workshop: icfp22.sigplan.org/details/mlfami…

This looks like an exciting program to be a part of. Come join us! I've been at Jane Street for only three weeks and am already deep into unboxed types. You'll have a chance to collaborate with a clever team and make a quick impact, too.

@AndersMiltner @ulysses4ever @JaneStreetGroup Yes yes yes. I know.

@RaeHaskell @ulysses4ever @JaneStreetGroup Modular implicits please 🥺

I'm thrilled to share that today was my first day working at @JaneStreetGroup. First task: work out unboxed types for #OCaml. Do not fear, though: Part of my job is to continue contributing to #Haskell. What will the two languages learn from each other? So excited to find out!

@gusbicalho @tweagio I will no longer be posting videos for @tweagio, but I expect the videos to continue: others at @tweagio are queued up to make videos (hopefully announcement later this week), and I will likely keep posting some on my own, but elsewhere.

@RaeHaskell @tweagio I'm going to miss your Haskell vids on the tweag YouTube channel! I learned a lot from them

@Profpatsch The VSCode command is called "Go to Next Error/Problem", and I just assigned a key binding to it. Ask again if this isn't enough information to get you moving.

Oh, also: how do you assign a keybinding for displaying the current line’s error in a CodeLens? I searched a bit but no luck.

Hi @RaeHaskell, what's your microphone/audio setup? I'm thinking about recording similar explainer videos and yours sound sublime.

@Profpatsch I use a Shure MV7, as recommended by Roman Cheplyaka, with an arbitrarily chosen pop filter. (Having the filter is important, in testing.) The mic is plugged into a powered USB hub, as I've read that doing so reduces electrical noise. My stand is a Gator Frameworks.

@ChShersh I agree that TTG is hard to navigate. I like the idea of it, and it would be very cool if TTG gives rise to a native language feature (in some language) and makes it better. But I've argued against expanding TTG into new areas, precisely because it's so opaque to contributors.

🌳 "Trees that grow" and GHC API show no mercy to GHC plugin developers 🥲 This little fella below is supposed to be AST pattern-matching on a function applied to two arguments 2⃣ It's not something they teach you when advocating "Haskell is great for compiler development" 😢

@Iceland_jack I haven't gotten explicit permission from co-authors to post the work, but it was rejected for fair reasons. The paper was about a cool new library and focused too much on its implementation... which was creative, but not really ground-breaking. So: not enough science.

@RaeHaskell That sucks. What paper got rejected?

Back after a hiatus for Haskell Symposium submission (rejected 😢) and then holidays.

Simon Peyton Jones’ work to deliver outstanding computing education is recognised in the Queen’s Birthday Honours ahead of the Platinum Jubilee celebrations. 👏🎊 Read more: computingatschool.org.uk/news-and-blogs… #PlatinumJubilee #Congratulations #OBE

@fancytypes I'm excited about the possibility of unrestricted representation polymorphism. Imagine ∀ (r :: RuntimeRep). KnownRuntimeRep r ⇒ ∀ (b :: TYPE r). b → … If we know r in advance, we get fast code. If we don't, we do a runtime lookup of how to work with a value of type b.

Happy Wednesday! What is your favorite use of dependent types in programming?

I'm writing a paper for Haskell Symposium, using @kosmikus's lhs2TeX for code formatting. My editor: VSCode. And, amazing to say, HLS is finding type errors in the middle of my paper. Very cool!

Apologies to my loyal fans, but this week got away from me -- no @tweagio video. :( I'll make it up with 2 next week! :) One will use criterion to run microbenchmarks on the Wordle code. Don't worry if you haven't kept up on the (long) series -- it will be relevant regardless!

@monadlife @tweagio Thanks for the ❤️! Glad you're enjoying it. It's been fun to put together. :)

In awe of the 8-part, 6.5 hour Wordle in Haskell series that @RaeHaskell from @tweagio has put together. And still going! It’s great watching Richard’s thoughts on a single problem evolve over an extended horizon. You don’t get that with normal tutorials. youtu.be/nr6rYvhACHs

Alert: The GHC Steering Committee is contemplating a proposal that will cause some breakage. We want community feedback to inform whether we should proceed. Read all about it -- and share your thoughts -- at Discourse: discourse.haskell.org/t/feedback-req…

Well, video will be up in the morning. Turns out the problem was much easier than I thought (`cabal repl test`), but it came at the end of a painful recording session, with lots of silly mistakes, and I was out of creativity/patience by the end.

I must apologize to loyal watchers of my videos made with @tweagio -- there will be no video this week. Since getting going in Oct 2020, this will be the first week missed. The problem? github.com/haskell/cabal/…