> Engelen: ‘Die anekdote wilde ik net gaan vertellen, maar het is helemaal goed joh.’
> Thieme: ‘O nee, vertel, vertel.’
> Engelen: ‘Het is wel een beetje mijn anekdote, dus je moet er gewoon met je tengels van afblijven.’

Jezus wat tenenkrommend

True, but it was also expected that the VVD would get decimated these elections, which did not happen. So it could very well embolden them. I guess we'll see, but either way I still firmly believe negotiations will be extremely difficult. Our last three(!) formations were over half a year long.

VVD-leader Yesilgoz has however said she did not want to form a government with GL-PvdA. I also have a hard time seeing D66 together with JA21, so I feel like we're going to have a very difficult formation, as is tradition at this point.

At least I'm pretty sure Clippy didn't cost extra?

how haskell-specific is this problem? it feels like something most type systems would struggle with, since you need to have the forall somewhere.

Make your own text editor 🥸

also, I personally quite like Kate, but I think it's best if you use KDE, and there is no plugin ecosystem to speak of, nor do I think making one is easy (pretty sure it needs to be done in C++...)

Seems like it does, it probably defines foldMap and then goes left-through-right:

Fedora offers an atomic Sway spin: fedoraproject.org/atomic-deskt...

Might be a nice way if you do want to give i3-style tiling a try!

I knew you could derive Functor, but I didn't know it also worked for Foldable. True 🤯 moment for me, tbh.

I wonder though, will Foldable behave differently if Node is defined as `Node (Tree a) a (Tree a)`?

Unicode is annoying because it's hard to type. And Haskell typically is pretty rich in symbols (I mean, what the hell is <*>?) so it makes sense to me that they don't use a keyword like fn.

Rust's |x| would work, I guess, but has the same problems. Just -> would make the syntax more complex IMO.

Firefox does not share my posts publicly or anything like that. The fact here is that people are mad about the JSON file being distributed publicly on HF. The file hosting does not federate with Bsky so I don't think it's covered, but just saying "JSON file" is a reductio ad absurdum.

The dataset does not include the post author name in a human readable way. As such, searching your or someone else's username is completely pointless.

Alpindale uploaded the file to Huggingface, a SERVICE to share AI content. Doing that provides a SERVICE to AI developers.

So I guess the strictness analyzer succeeds here? That's pretty neat!

Still not sure how I feel about relying on a compiler optimization to get good asymptotic memory usage though, especially since it may also require inlining? Guess I'll be trying a few things today!

I'm also curious to see if the i3 lineup will get a cluster of E-cores, especially now that hyperthreading is gone. Making a 4c4t CPU seems kind of not-the-move even for a budget chip, since all i3's since 10th gen have been 4c8t. 4p+4e 8t would be fine and probably a really nice MT uplift too.