esper VS python-tcod-ecs

That's cool, I didn't realize you had implemented an ECS library for tcod. At first glance it seems a bit more featureful than esper. The tags and relations functionality seem particularly useful for queries. l'll consider using it for the tutorial.

If you don't mind the broken type hinting then you could go with Esper.

I've been closely looking at the Python ECS libraries Snecs and Esper. I was hoping I could use one of them to fix the Python tutorial without making the tutorial about writing an entity-component engine. What I actually want would be a Python ECS library with NumPy for static types, but a library I did find for that didn't have a good enough license for a tutorial. For now I'll be okay with dynamic types just so I can fix my current problems, and I'll think about performance some other time.

I've been attempting to do this using an entity-component-system (specifically the Esper system). The difference being that I'm trying to explain my processors in separate documents first.

Not much to see/to do yet. I just wanted to try Esper and LDtk.

I'll probably have a lot to say about the Python tutorial as it progresses. I'll likely talk about the most recent deprecations in Python-tcod and how to rewrite entities using tcod-ecs.

tcod-ecs | GitHub (with examples)

Not much news with libtcod itself, but I've continued to mess with Python ECS stuff. I think I've done the most I can do with the EC objects, so now I need to store entities in a global registry to support more complex tasks. Mainly things involving ECS queries. The current experiment is here. I'm not focused on performance, my current code is very dictionary heavy. In theory I can later add code to optimize queries based on usage.