#advent_of_code — Public Fediverse posts

Some serious health problems are messing with #Forth and #Advent_of_code plans. I was already moving slow but now I find that I can't work the problems and learn Forth at the same time. I love how the language works (OG assembly guy) but I can't make the jump from "writing words" to "writing phrases" of code--real programming requires phrases and then sentences.

I'm tired of Pascal, Go does little for me and often annoys me (all loops are "for", really?). I could pick up Scheme again but I don't trust that I'll have the capacity to move from phrases back to sentences.

Out on eBay I saw some old books that I liked at the time--the #BASIC era. One of them was a book I regretted losing in move (BASIC With Style). They are ordered and I've dusted off #QB64. I was more of a boot to Basic guy myself, but I can deal with QBasic style.

It's chicken soup for the ill programmers soul. Easy peasy comfort coding while enduring treatment.

https://qb64phoenix.com/qb64wiki/index.php/Main_Page. QB64PhoenixEdition

@kimvanwyk I have been having so much fun with Advent Of Code 2025 (https://adventofcode.com/2025) -- thanks for putting me on to that!

I'm done with Day 10 (so 20 puzzles). I was able to solve 18 completely on my own. I only used AI for lookup-type questions, e.g. easiest way to create a list comprehension that takes stringA and turns it into a list of sets. (silly made up example, just to illustrate what I used AI for -- i.e. no help with the design of the solution itself, only language specific help, not puzzle logic help.)

Day 7 Part 2: After a couple of days of frustration, I had to peek at your solution in GitHub, and immediately grok-ed the approach you took and was able to manually replicate it.

Day 10 Part 2…. I hate to report… is the first one I eventually had to turn to ChatGPT to for the entire solution. It required higher order math-combination logic that I am not at all familiar with. I first wrote logic that was able to correctly brute force the example data, but the real data took the number of combinations to ridiculous levels -- well beyond my PC's reasonable abilities. So it required algorithms that I know NOTHING about. I am VERY impressed that ChatGPT was able to first give a brute force method when I gave it the example data, and when I gave it one line from the real data, it went “woah there buddy, that’s not possible with brute force, so let me give you a mathematical algorithmic way” and it then gave me a solution which worked first try, and solves the full problem in 2.5 seconds. Hate that I had to turn to AI, but there was no way I was going to solve that.

#advent_of_code #advent_of_code_2025 #adventofcode #adventofcode2025

Readings shared October 7, 2025. https://jaalonso.github.io/vestigium/posts/2025/10/08-readings_shared_10-07-25 #AI_coding #Advent_of_Code #CommonLisp #FunctionalProgramming #Gemini #Haskell #ITP #IsabelleHOL #LLM #Math

Using an LLM on the Advent of Code. ~ Joe Marshall. https://funcall.blogspot.com/2025/09/using-llm-on-advent-of-code.html #Advent_of_Code #AI_coding #CommonLisp #Gemini #LLM