hutt's puzzle dump

 I had the good fortune of noticing this was a new puzzle type on the Puzzle Square home page only about a day after the first one was posted there. I really like how Kurarin feels new yet natural. The clue type synergizes nicely with the inherent constraints of a loop.  These first four puzzles were made on 6/26: https://pzprxs.vercel.app/p?kurarin/7/8/i2j4g8r4zh84h8o8h8h2zh2m https://pzprxs.vercel.app/p?kurarin/8/6/k1p2z2g422gc2z2p3l https://pzprxs.vercel.app/p?kurarin/7/7/n8g4gcmct8p4hcs4n8r https://pzprxs.vercel.app/p?kurarin/8/8/o8h4gcn8cm2l2h8g8zi4gcg2l2n48n8g8h8o These next four puzzles were made on 6/27, have only gray clues, and are notably more difficult: https://pzprxs.vercel.app/p?kurarin/8/5/n8h8h8n8h8h8n8h8h8n8h8h8m https://pzprxs.vercel.app/p?kurarin/7/9/s222222u82l2zi2h2h22222zi2j2h2h https://pzprxs.vercel.app/p?kurarin/8/8/t8o82g2m2j8g8i2i2t2j8g8g8g2l82g2g8p8j8n https://pzprxs.vercel.app/p?kurarin/10/10/p8m8q8222h8s8g8g8g8g2x2k2j8g8g8i2k2x2j8g8g8s822g2g8q8g8k8...

 I just made this while thinking about interpreting a three-dimensional puzzle with two-dimensional cross-section rulesets. I'm pleasantly surprised it worked out to only need one clue. penpa link (answer check) Solve the Skyscrapers puzzle in such a way that each digit counts the amount of squares that should be shaded in that digit's position in the five incomplete shading puzzles below. A square may only be shaded if it is also shaded in the puzzle immediately to its left, with the exception of the leftmost puzzle, which is entirely shaded. Each layer uses a different puzzle type from the layer candidates. No clues or given edges have been revealed for the shading puzzles. All rulesets can be found at  https://pzprxs.vercel.app/list.html

 I recently filled up one of my graph paper notebooks. This inspired me to go back through the notebooks I used before that one to see if I had any cool old puzzles lying around that I had forgotten about. The following puzzles were probably drawn in late May 2022, except the rightmost puzzle here was re-clued today to avoid what might've been a nonuniqueness. I have suspected that I first came up with this puzzle idea as far back as 2016, making this the ultimate recycling of my ideas. These puzzles do use a slightly adapted ruleset, but I think it's better than the original rules. Shade some cells to form walls, such that the unshaded cells form rectangular rooms. Then, connect the rooms by drawing hallways in the pre-marked light gray cells. A hallway is either a straight line or a turn which connects two rooms. All rooms must be connected through the hallways. The letter "h" has been used to represent chairs, which cannot be shaded.  penpa link (no answer check)

This was made on 5/29/25. The idea was to see if the structure could give the feeling of exploring connected rooms. https://puzz.link/p?fillomino/13/13/-17-17-17-17-17h9-17-17-17-17-17-17-17-17-17-17h3-17-17-17-17-17-17-17-17-17-17i-17-17-17-17-17-17-17-17-17-17-17g-17-17-17-17-17-17i-179i3-17i147-17k-17121l2m2g-17k-17j1g-173i9-174g4-17-17-17-17-17-17g-17-17-17-17-17-17-17-17-17-17-17i-17-17-17-17-17-17-17-17-17-171g1-17-17-17-17-17-17-17-17-17-17g4g-17-17-17-17-17

This is the project I've been teasing! You could call this page a devlog. I've tried not to spoil too much. Since every post here should include a puzzle, I've put the link to the project at the bottom.      This all started about two years ago. I had an idea for a puzzle game where the player solved paper puzzles with input methods restricted by the tools they had available. These tools, called stencils, would be used for solving the puzzles as well as marking the puzzles on the grids. This led to me trying to create a language suited for describing puzzle rulesets. It would have to be nuanced enough to describe complicated ideas, but simple enough to be conveyable to the player.      Stencil was a short game and did not take very long to create. It was especially convenient that I did not need to program because I had decided to present Stencil in a folder system. The main reason for doing so was because progression could be locked behind password-protect...

 Recently, I decided to revisit pzprxs to see what had been added. Since the last time I checked was before the latest merge to puzz.link, there were plenty of new types for me to explore, most of which I hadn't heard of before. One in particular that I underestimated at first was LIX Loop. The example in the rules was pretty straightforward to approach like a Yajilin until the ending deductions, so I started to appreciate the nuances of its logic while constructing this doppelganger. These puzzles were both made yesterday. https://puzz.link/p?yajilin/8/8/e31d11i41e31a41m31l31c11b11a https://pzprxs.vercel.app/p?lixloop/8/8/e32d12i42e32a42m32l32c12b12a