I first started to play around with Puzzlescript because I liked the idea of HTML5 games - Love2D was a pretty fun venture, but it felt like it wasn't going anywhere fast due to being a huge pain to export.

I think puzzlescript has a kind of perverse appeal for giving you the opportunity to create something way too complicated for the engine's intentions.  With Middlegrounds I certainly did that. I think I worked around the limitations of the engine pretty well; the colour scheme is fairly distinct, the characters are reasonably varied and contextual controls work pretty well. There is some slowdown with the way that water is handled - its pretty hard to limit the amount of times the rules to animate it are called - but the nature of puzzlescript means that the player will barely notice the wait unless they are playing on a potato.

I think working out the scripting issues made me a stronger coder in general - so that I will probably be able to handle future problems a little easier. In a lot of places I had to figure out which checks would stop what, and how I could use those checks in the context of Puzzlescript's logic.

Where I think my real lessons were learned, is in the field of paring back an idea. My original plan was to make a gameworld more the size and grandness of the original Zelda - but as my schedule narrowed I had to cut things back everywhere.  

The game world is probably a third the size now of what I had originally intended - which I think is a bit disappointing for the sake of the final sort of obstacle.What was originally to be a kind of mystery that is unravelled over the course of a longer adventure became more the sole obvious objective.