mangoiv

mangoiv, 5 months ago to haskell

I feel like although for nerds like me putting more types in a language is a cool thing, what would be much more important as a topic of research, in practise, is reducing the syntactic overhead of type systems. I recently heard so many complaints about this, I feel like we’re not there yet; I do agree.
#haskell

mangoiv, 8 months ago to haskell

Squeal is so fun!
But also so slow to compile!

https://flora.pm/packages/@hackage/squeal-postgresql

#haskell

mangoiv, 2 months ago to NixOS

meeting notes from today's board meeting, the only topic they talked about was the sponsorship policy:

https://discourse.nixos.org/t/open-board-call-2024-03-20/41898

#nix #nixos

mangoiv, 22 days ago to NixOS

Oh we’re so back, girls #nixos

https://discourse.nixos.org/t/nixos-foundation-board-giving-power-to-the-community/44552

mangoiv, 3 months ago to random German

I don’t remember @bobkonf to be that expensive - have you changed something? 🥲

mangoiv, 3 months ago to random German

Hey @kosmikus perhaps something you will like: https://mangoiv.com/ghci-module/ depending on what you said on the podcast. Still a bit barebones still but the idea is what you’re describing :3

mangoiv, 2 months ago to random

@maralorn did you do something about the icons in nom? It looks much better now for some reason! Thank you for fixing that <3

mangoiv, 11 months ago to random

Hi everyone!

I have written a little snippet in which I explore some "free" generic programming in Haskell, where you define your types as polynomials and then can use a lot of their structure to program on them, this enables more powerful things than deriving Foldable because the structure of the types actually enable the folding, not some compiler magic. I can already show that this works better by fmapping and folding types for free that we cannot (stock/ newtype) derive Foldable or Functor instances for in GHC. (thinking about newtype instances where the type variables don't occur in the correct spot)

I wrote a little snippet of what this can do (for example) at the very end. I would be happy about some feedback and especially whether this is novel enough to be worth a blogpost or something like that :)

https://git.sr.ht/~mangoiv/haskell-sandbox/tree/main/item/Recursion.hs
or, if you just want to skip to the chase,
https://git.sr.ht/~mangoiv/haskell-sandbox/tree/main/item/Recursion.hs#L443-510

mangoiv, 1 year ago to random

Idk if it’s window manager/compositor specific but xmonad is 20x more stable and less buggy than hyprland. We really need waymonad.

Things I like about Wayland:

• Gestures, less tearing
• everything is configured within the compositor

Things I don’t like about hyprland/Wayland:

• monitor support is abysmal (can’t bring my mouse from laptop to external monitor??)
• x Wayland looks like garbage on high dpi displays
• need to enable “experimental” features to use Wayland with certain programs, electron apps are one example
• shit doesn’t work out of the box with display managers. Neither of them. Tried sway and hyprland

mangoiv, 1 month ago (edited 1 month ago) to haskell

Is it just me or is this whole thing about „don’t do X if you don’t really know what you’re doing“ really counter productive? Many people live in fear of X now and don’t start thinking if X is an actual problem or learn what the problems are that could rise from X.

I see this very often in the #haskell ecosystem where the solution to doing X properly is often learnt in a few minutes investigation.

mangoiv, 1 year ago (edited 1 year ago) to random

Another possible way of representing existentials more modularly, (this composes quite well) which is slightly different to what Edsko showed:

Thanks for this Haskell Unfolder @kosmikus and Edsko, it was great again :)

mangoiv, 10 months ago to random

@kosmikus have you thought about writing functions in Terms of Proxy# and then providing a „common use“ wrapper that uses a Proxy instead?

That is just one more reminder that it would be so nice to have an {-# AMBIGUOUS #-} pragma that does AllowAmbiguousTypes for just one definition. (To avoid accidental pollution with definitions that can’t be inferred or hide type variables)

Thank you for the episode, it was once again quite nice :)

mangoiv, 10 months ago to haskell

Who’s coming to MuniHac 2023?
Next week in - who would’ve guessed it - Munich!

https://munihac.de/2023.html

#haskell

mangoiv, 8 months ago to random

Wide awake after 4 hours and 30 minutes of sleep to head off to #NixCon

(I really, really hope I’ll be able to sleep in the train) :blobcatcomfslep:

mangoiv, 8 months ago to sircmpwn

Every time I see things like this, I feel very sad that people choose to be like this on the internet. I do use Hyprland but I will definitely keep steering clear of their surrounding community. This is really depressing.

https://drewdevault.com/2023/09/17/Hyprland-toxicity.html

Also thank you @sircmpwn , this is a really good post and it's good that this gets some attention!

mangoiv, 9 days ago to random

Hey guys, I made a thing:

https://git.mangoiv.com/mangoiv/haskell-note

It is a (somewhat wip) pastebin microblogging style written in Haskell using fused-effects, servant, one-time-password, lucid2, htmx, styled with bulma v1.

There’s a nixos module you can use to get it running on your machine.

My instance is hosted at https://bin.mangoiv.com/

mangoiv, 3 months ago (edited 3 months ago) to random

random stuff:
I have long ago started writing Haskell types like: newtype A = MkA {unA :: B}, to avoid punning. So far, so good. However, someone told me that writing this: f (MkA b) = ... is somewhat odd because you don't MkA, you only do that the other way around. So I have started doing this very cursed thing everywhere I have a MkA-like constructor: f (unA -> b) = ...

mangoiv, 2 months ago to random

Don’t host your own mail server.

Don’t do it.

If Google decides it will be the only one not accepting mails from you 50% of the time, then you won’t be sending mail to 90% of your customers. 🥲

mangoiv, 5 months ago to random

Is functional.cafe going to federate with threads?

mangoiv, 1 year ago to random

@yvan what do you think about this: https://www.well-typed.com/blog/2023/03/purgatory/ ? Does it align with your work? Can you perhaps even merge it?

mangoiv, 26 days ago to NixOS

sudo is unmaintained on #nixos

all thanks to the great work of the determinate systems community.

mangoiv, 6 months ago to random

@kosmikus about you mentioning rel8, have you actually been able to use this in a reasonably sized project? I have found the compile times for it prohibitively high, even for trivial modules and on a relatively small project. Now I’m using squeal which is still very slow to compile but it’s several times faster than rel8.

mangoiv, 16 days ago to haskell

Out:
(A -> D) -> B -> C -> D

In:
(A -> (D, B, C) ) -> D

#haskell

mangoiv, 3 months ago to random German

Here’s mine #40sday :ablobcatheart:

mangoiv, 12 days ago (edited 9 days ago) to random

Who’s coming to ZuriHac?! I’m so excited!

#haskell