@RedwolfProgrammed Yeah, the ice is the big problem. Six inches of snow in one storm is pretty respectable, though. If it were Nebraska, that'd be on the upper end of "medium." Then again, it was probably wetter, heavier snow, so six inches of that is more than six inches of the lighter, fluffier variety.
I always find it interesting that the truth machine question has so many more answers than the other trivial challenges (excluding HW, and infinite loop w/o output)
@cairdcoinheringaahing I bet it's because the truth-machine concept is doable for all sorts of languages, even those with weird I/O formatting. You can take input as an integer, string, character, single bit, and probably other ways as well. And there are only two possible inputs, so the logic is very simple.
I never answered Add Two Numbers in BitCycle because it requires supporting negative inputs, which makes it a stealth "add under some conditions, subtract under other conditions" challenge.
It's a trivial challenge in all but the most esoteric languages, and so the fun of the challenges comes from answering in those esoteric languages. But its made unnecessarily complicated by requiring negative numbers, but only for the esoteric langs, as every other language naturally supports negative numbers
And because it completely "covers" a question about just adding two positive numbers, we can't even post a better one, cause it'll be closed as a dupe (at least, not without meta discussion, which is unlikely to be unanimous)
For some reason, my course gives us our overall marks as "total marks scored on all assessed work" instead of like. a percentage of correct marks vs total possible marks. So I have a Current Mark of 149 in Probability O.o
@cairdcoinheringaahing My initial reaction (without thinking the ramifications through) is that it could be replaced with two questions: Add Two Nonnegative Integers, and Subtract Two Integers (Possibly Negative)
@emanresuA I don't know, but I'm going to guess not: on a first readthrough, I don't see a way to access data arbitrarily deep in the stack, and I don't think it has enough math operations to simulate a tape/queue/pair of stacks using the unbounded integers. But I could easily be wrong.
> Better rules of thumb are welcome.
Feel free to edit one in if you come up with a good one (maybe get feedback here first)
@cairdcoinheringaahing My one problem with that phrasing is, what if there's only one person here who uses the language, but they answer a LOT of questions? I'd still support that language being LotM. Might encourage more people to start using it.
@Anush I used Julia for a while a few years ago. There were some things I liked about it (for example, the ability to easily vectorize anything by adding .). But ultimately, there were other things I didn't like about it, and I went back to Python.
The things I ended up disliking about Julia were 1) it puts all those overloaded functions in the global namespace together, so it becomes nearly impossible to sort through the 138 different definitions of + to find the one you want, and 2) you can write code at the top scope but it's way less efficient or something? I don't remember the details.
Julia claims to be significantly faster than Python, but I wasn't seeing much of a difference for what I was doing (an implementation of tinylisp). It's very possible that was due to my lack of experience with the language, though.
How do you mean "really" big? Scala has BigInt which has worked pretty well for me so far, though when the numbers get past a certain size it gives up and throws ArithmeticError.
But any language is going to have problems when the numbers get so big they don't fit in your computer's memory.
I believe Scala's BigInt is basically a wrapper around Java's BigInteger, so you could probably add Java to the list (but why would you want to use Java)
SWI-Prolog seems to support big integers with no problem, although I'm not sure how fast it is or whether the logic programming paradigm makes any sense for your purposes
@RedwolfProgrammed Yeah, I was going to say, I would expect any high-level, general-purpose, practical language designed in the past 20-30 years will have some kind of bigint support.
I suspect that by the time a pattern is uniquely recognisable, it is pretty much complete.
For example, if I write ⊢⍤ then I surely want the next character to be / or \ or ⌿ or ⍀ but it could be any of those, and once I type that one I want, the pattern is complete.
