jeffbaumes

I love putting together puzzles, but I hardly ever get a chance to do puzzles with others. I made this simple game to experiment with a new type of puzzle experience at a grand scale. Everyone works together on a single puzzle. You log in with a color hex as your username. Each piece is a pixel of a 1000x1000 picture, which reveals as it gets built. Pieces are scattered randomly at first, and lock in when placed in the correct spot, at which point you automatically pick up the piece that was there and continue. Arrows guide you to the general area (hold a direction to move faster), but you still need to search the 5x5 target area for the correct exact location. The puzzle was constructed so each pixel is a different hex color, so it uses 1 million of the ~16 million possible colors in the picture. To make the image, I took a seed image and greedily assigned the closest available color to each pixel.

> I also see more and more pull request summaries clearly written by machines for machines. They are rich in detail: this was changed to that, these things were split, those things were merged, this was left untouched, tests were added for this, and so on. But I’m left asking: Why are we doing this? What is the value? How risky or urgent is this work? Where do you want my input? What should I pay attention to?

My own solution for this (for now at least) is to add a preamble section “AI;DR” to PR descriptions, filling what a TL;DR section would have done in the past. I’ve socialized to my peers that no part of such a section will be written by AI, and it should contain exactly this missing high-level content described by the author.