Sophie Kirschner
pinemach.bsky.social
Sophie Kirschner
@pinemach.bsky.social
I'm just learning as I go.
I think you mean Walmarepλtes ain't nothing to fuck with.
November 24, 2025 at 9:51 PM
Ha I actually made a mod for that bananas.openttd.org/package/newg...
World Town Names - BaNaNaS - OpenTTD
bananas.openttd.org
November 24, 2025 at 9:37 PM
As it is, it feels like the number of cases where just carrying around and using a normal laptop isn't strictly better than carrying around a DevTerm is vanishingly small
November 24, 2025 at 9:07 PM
I bought a DevTerm years back and it's a beautiful little device. But in practice it ended up being too big and heavy to be as portable as I really wanted, and at the same time too small to be at all useful as an actual dev device. I wish there was a lighter and more handheld UMPC like the DevTerm
November 24, 2025 at 9:07 PM
One day. One day I will make my dream automation game. It will be voxel based, you'll have workers to manage and autonomous cities to help make prosperous, and although there will be factory-building the greatest focus will be on building vast transit networks, including road and rail...
November 24, 2025 at 9:01 PM
Whenever I play OpenTTD I'm sad that the logistics aren't more in-depth. With Factorio I'm sad I don't have a Z level, and also automated road vehicles. With Satisfactory I'm...just sad. I so want to take them all and put them in a blender. (With a dash of Dwarf Fortress.)
November 24, 2025 at 9:01 PM
It's also the only one with a clearly defined border/boundary! I think it's extremely sensible that the IBM design was chosen over these others
November 24, 2025 at 7:26 PM
Came here to say Spring. That was a heck of a movie
November 24, 2025 at 7:22 PM
THIS is what y'all are hung up on? At least Olivia is over the age of consent.
November 24, 2025 at 7:05 PM
...resisting...my impulse...to infodump...about coroutines....
November 24, 2025 at 4:39 PM
(Kind of like how in zig you can have bigints if you want, you just have to use them explicitly instead of writing only `1`. And you can have wrapping int math if you want, you just have to use explicit wrapping operators like `+%` instead of only `+`, since int overflow is usually an accident.)
November 24, 2025 at 4:26 PM
(Inb4 you say what about writing an infinite "while" loop doing the same thing, I would 100% agree with you that loops which can't be statically proven to have a reachable exit condition should be an error by default, unless you mark it as intentional, because ime it's an accident 99% of the time.)
November 24, 2025 at 4:22 PM
(Inb4 you say well it's like that with pointers and segfaults too in C, I would 100% agree with you that pointers should be safer and a little more fenced off than in C, so it's a little harder to blast your foot off by accident. I am hardly upholding C as a model language here.)

(Maybe zig tho.)
November 24, 2025 at 4:22 PM
Basic simple everyday arithmetic expressions having the potential to monopolize all your RAM and/or hang the process and/or maybe crash your computer is a huge footgun. Especially when Python is extremely unusual in how this is the case.
November 24, 2025 at 4:22 PM
My thinking is that if you want Bigints, you should be making that explicit somehow. Instead of writing `x = 1` you'd write, for example, `x = bigint(1)`. Or even just `x = 1n` as in JS, that's better than Python. Not that they shouldn't be available as a feature. Dunno where that came from ???????
November 24, 2025 at 4:22 PM
But I like asm macros. They're neat :(

Understanding how higher-level code translates to lower-level concepts is cool and fun and I think a big part of what makes me good at what I do! For better or for worse though it does make Lua kind of painful and harder to adjust to
November 24, 2025 at 3:51 PM
Two's complement integer arithmetic: Simple, elegant, can be fully explained to anyone with a fair grasp of math in maybe an hour

Arbitrarily large integers: I've been writing code for 20+ years and they're still arcane and frightening magic to me. Let alone beginners starting coding with Python!
November 24, 2025 at 3:48 PM
Integer overflow is maybe insane and bad, but the better solution to that is checked integers by default (produce a traceable error upon integer overflow) as they do in zig (I really like zig). Arbitrarily large integers are SUCH a huge and terrible can of worms to dump on everyone by default
November 24, 2025 at 3:43 PM
Anyway I have spent a truly silly amount of my limited time on this earth into learning how computers work, and any time a language wants to pretend like it doesn't actually run on a computer, that trips me up. I just prefer it when languages don't do that!
November 24, 2025 at 3:39 PM
But...closures are just functions accepting a pointer to a heap-allocated context object. You can 100% do that in C or asm, there's just not syntax candy making it quite so convenient. Which is maybe for the best! Lots of programmers don't understand closures, the costs and caveats, and misuse them
November 24, 2025 at 3:38 PM
I'm a big fan of explicit variable declaration personally, and assigning to an undeclared variable is just an error rather than implicitly defining a new local or global variable. But I'll take Python's local by default over Lua's global by default, at least.
November 24, 2025 at 3:26 PM
(I'm looking at you, Python 3 using bignums instead of bounded ints. I also disapprove of this)
November 24, 2025 at 3:23 PM
0-based indices are a leaky abstraction like number primitives being integers or IEEE floating points is a leaky abstraction. You can build abstractions for other kinds of indices, or other kinds of numbers, but that's the kind the computer fundamentally uses and so makes sense to use as default
November 24, 2025 at 3:23 PM