Summer of Math Exposition

A fun exhibition of the game/platform and exploration of a single problem. Kind of on the boundary of promotional.

Pretty cool. My opinion of games that try to be programmable has been coloured by a number of them that ended up being terrible games, or not really programmable (there's only one possible solution so it's really just "find the thing the level builder thought of"), but what you've done does look interesting.

I like the presentation of different algorithms and quantifying their performance. I wonder whether the presentation would have been more effective without involving ABI-DOS. That seems an arbitrary way to limit the possible algorithms for a puzzle that is already interesting in the abstract. Instead of 8x8 or 8x11 in this game that requires significant thought from the reader, it could be sufficient to see how effective an algorithm you could develop with minimum components (registers / pivots, etc.), and then how it could be improved by additional components.

Thank you for your article and the link to ABI-DOS.

This was well written and included clever solutions to a seemingly simple problem. A couple of suggestions: - It was not clear why you were using this video game programming thing instead of writing things in pseudocode (or even python). I did not find it helpful and I thought it would be more clear to have your algorithms in pseudocode. - You could have a been a bit more clear on proving the algorithms actually worked - it wasn't too hard to piece together, but you might as well write it down. - It would have helped to make explicit at the start the tradeoff between space, speed, and accuracy and have a quantitative metric for each that you could write down. You acknowledged this later, but for a while I was reading your algorithms and wondering why you insisted on using almost no space or were suggesting an algorithm that would not always work.

I thought the topic was pretty cool, but in the end there wasn't really anything that made this stand out.

Writing everything in terms of ABI-DOS (which pretty much no one heard about, since it's still in development) made this way more confusing than it should be.