I just got C FFI working in my LLVM based language last night so I'll need to take a look at playing around with BCHS :)

This one is why I never stuck with unity. You were always either learning something that was going to be replaced, but the replacement never really came. Add on top a lack of CI friendliness (it might be different now but years ago it was clearly not designed to let you build projects from a CLI or script), install sizes, slow starts, still using a mono fork with patches instead of .NET core, etc I just couldn't stick with it

It feels like this really awkward place now where Godot is much nicer and faster for when you don't need all of the features that unity provides, while Unreal arguable offers more depth and features. The one thing Unity had going for it there was that .NET and reloading code was better than (in my limited experience with unreal) C++/blueprints

Godot hit this really nice sweet spot of consistent UX, complete enough features to get you by as an indie, and a clean enough codebase + design that I felt like if something was missing from the engine that adding it myself was reasonable (ie: at one point I was writing a raymarching based renderer for fun and adding it as a new render seemed like a reasonable project even solo)

I do think the ECS direction Unity started to go was a good call, and I actually did enjoy it, but after waiting for long enough I sort of just forgot about it at this point

Its fun playing around and seeing how much has stayed the same in desktops since then, and so far the most surprising thing to me was that a scrollable text area does not re-render the text as you move the scroll bar with your mouse

Totally makes sense that it wouldn't be worth the cycles back then, but its one of the few things that was actually surprising to me today

I was only able to do that after I had solved multiple related problems in different places and started introducing subtle bugs by accident / had difficulty detecting all edge cases

I've noticed whenever I use LLMs they introduce the same kind of thing but at much smaller scales than I would. They often suggest solving the wrong problem when I prompt them to diagnose specific bugs too. Usually opting for a shortcut that introduces its own issues and ironically calling the proper direction "too complex" when it's really not.

I am working on a compiler for an OCaml inspired language (I am on a sabbatical) and used them a lot in the early phases to check my ideas and search for prior work. I use them less now (mostly debugging at this point), but have made use of LLMs to write some pieces like source map generation, a couple of small and well known algorithms that were new to me, and an analyzer to dump a schema from a typescript module. Other times I've tried to use it for other things, I have ended up regretting it because of subtle bugs

I don't think it made the design phase any shorter but it definitely helped save me energy in researching and made that period feel less like a fever dream. Design phases can be draining for me so that was welcome

I started using harnesses because they are good for when something breaks and it's not trivial to investigate so I'll have the agent tell me what's happening, then using that to produce my own change