I've put off release due to feature creep, but what I think is the final version of Shuffling Around is now at itch.io. It was conceived over 10 years ago, in April 2012, and it was submitted to IFComp in September 2012. Apparently release 4 occurred in July of 2015. Which is a while back. At least I've had other rewarding projects on the way, too.
DrkStarr, Olaf Nowacki and Amanda W. helped me finally push this through in 2022 with testing efforts.
The major changes to the story and puzzles themselves are a much smoother introduction and endgame. There are now three solutions in the final room to win. I kept the original, which feels strained compared to the two, but I like the end. A personal favorite is that each liver in one room now has a description that conforms to the game mechanics.
There's been a long journey to get here. I've gotten better at Inform coding--well enough to look at stuff from 2012 and see why the stuff that I just needed to make work was, well, not perfect, but I'm glad I got things started as quickly as I did. The whole process of testing has helped me grow, I think.
That's the quick stuff! The TL;DR is below.
I vacillated on when I should release the next version. It seemed when I'd get close, I'd find more bugs. But Zarf's Python regression testing scripts helped me push a lot of bugs out of the way. I achieved a huge brass ring in the last two weeks before release: I'd never used custom check classes before, but I tweaked regtest to detect lines with just a period in them, or lines without adequate punctuation at the end. I felt like I'd really arrived at that point, both with Python competence and knowing that, well, a lot of the text's obvious errors were caught.
GitHub also helped me organize things, and I learned so many commands to do what I wanted, or where I said "It'd be nice if git could do this, but nah ... wait, let's just check and make sure on StackOverflow." And then I stumbled on a question with a hundred or even a thousand upvotes, asked back in 2011.
Originally I was just going to fix one bug Sean M. Shore found: if you eat the saltine and use your cheat powers on something you already know (e.g. you find what Store R should become before clearing F, I and M), then you have wasted your cheat.
Of course it expanded into much more: renaming one room, then another, then another. Some NPCs and scenery got renamed, and I also implemented bounding scenery for fun. It became a case of "I like this anagram, but I think I can do better." Then, later, I did.
Then, a bonus point here, text cleaned up there, hint features streamlined elsewhere, code refactored for readability, Zarf's test scripts to help make sure nothing broke, and I learned Python along the way to make sure that the test scripts covered all the special flavor and hint text I'd laid out there, or that the clues from the Tagged Gadget were in fact all accurate. I even wrote up flowcharts for solving the various areas using Trizbort.
The website with the downloadable binary and feelies (walkthroughs, PDF maps, etc.) is at https://andrewschultz.itch.io/shuffling-around.
The GitHub site is at https://github.com/andrewschultz/stale-tales-slate/. The sub-code for Shuffling is at https://github.com/andrewschultz/stale-tales-slate/tree/main/Shuffling.
Though I want to move on, this was worthwhile for me, and I encourage anyone who wonders if they should re-release that IFComp game they had some fixes for to do so, big or small.
(For anyone interested in Roiling, well, I wouldn't mind a tester or two. It's in very human-testable shape, but I can't quite push the button, yet. Even help with one or two areas is big.)