💾 Archived View for tilde.cafe › ~spellbinding › gemlog › 2022-04 › 2022-04-20.gmi captured on 2024-08-18 at 17:29:30. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-01-29)
-=-=-=-=-=-=-
This is a wordlog for the SpellBinding puzzle (in which you construct words using the specified letters; each word must contain the anchor letter).
EULOGIZE GEOLOGIZE GIGOLO GLOGG GLOZE GOGGLE GOOG GOOGOL GOUGE IGLOO LOGE LOGO LOLL LOOGIE OGEE OGLE OLEO OLIO OLLIE OOZE OUZO
Pangrams: EULOGIZE
Yikes! This is exactly the puzzle from 04-06! Either I won the lottery (there are like tens of thousands of possible puzzles), or something is screwy with randomness - or much more likely, I somehow goofed and overwrote a file. And I definitiely misspeled the letters as 'EGULZU-O' in ghe wordlog of that date...
Let's see - how many possible puzzles are there. Cranking up Emacs, starting SBCL, loading dictionary code... My computer went down a few times in the last couple of days - what gives?
M-x slime
(ql:quickload cl-spelling-bee)(in-package :bee)(bee-init)
#<HASH-TABLE :TEST EQUAL :COUNT 7108 {10037C9C63}>
7108 pangrams in my purse! That means there are 7108 * 7, or almost 50,000 unique games. Although many, more often than not the puzzles are too big to be useful - who wants to slog through a 500-pointer with 150 words? So I skip these, opting for the 100-200 point range. Sometimes I manually trim the solution set, with special one-day-only rules, but that turned out to be error-prone.
Anyway, it is pretty unlikely the duplication was an accident - much more likely to be a human error. How did I manage to overwrite a file?
Well, I make a puzzle in Lisp, check for missing words, then copy into a separate file using Emacs. I usually open a file first, so normally overwriting would not happen as Emacs would open an existing file and I would see it. The only way I see it happening is that I was called away after saving a puzzle, left Lisp opened, then next day opened another session in a different workspace, saved a few puzzles, then was distracted again, then came back to the first session and said "oh, that is a good puzzle", and saved is as a new separate game. That is kind of weird and unlikely.
Each game is saved under the name of its scheduled date, so it's difficult to check for duplicates without a shell script of some sort, and it is so unlikely that I don't bother. I do have a little C program that is similar to the final game CGI to check for the validity of the puzzle file, but it would not catch a duplicate.
Weird. Some day I will figure it out. Or it will remain a Murphy's law mystery forever. We shall see.
Another day.
contact: spellbinding@tilde.cafe