I've been investigating RPGMaker MV and its limitations and I've decided against using it. The main issues are the mapping editor, specifically the way priority works in the tilemap. It's very difficult to make objects walk-aroundable like a tree for example where a character is more than one tile tall as all tiles are either above or below, there is no dynamic ability to determine what should be in-front and behind based on where the two items are located.
Instead I've decided to use
MKXP-Z, and I'll be making some changes (hopefully) to support uncapped framerate and any tilemap size.
A few weeks ago, I spent some time converting MV's code to C# so I can run the interpreter code remotely so the client is only a visual representation, with no important code client-side. As part of that, I ended up attempting to copy everything, including stuff I didn't need, which sent me down rabbit holes and eventually got stuck. As with this shift to MKXP, I'll also be rewriting the server again, but trimmed down to only support the bare minimum of what I need for now. Also, because this client will be desktop only, I am able to leverage UDP as well, which I'll be using for non-critical information (movement updates, etc.) and possibly input state (naturally with additional systems on top), I'll investigate that when I get there.
Because it's desktop as well, I'm hoping I can run much larger maps without issue. I want to avoid map transitions for the most part, similar to Skyrim where the outside is mostly the same map including houses, and dungeons etc. are their own maps. Hopefully because the client is going to be super dumb, I don't need to run event update code as frequently as might be otherwise required, so most of the heavy lifting will be the tilemap and network stuff.
Also currently thinking about the most efficient way to set up visual equipment. I'm thinking I'll have a front, back, left, right for each weapon, and I'll specify an anchor point on the sprite and on the character. Then I break the character into many layers and interleave them in the correct spot. Different stances could be an issue though, for two handed weapons. Also attack animations relative to worn armour. Hopefully if I throw everything into a function that renders a spritesheet, I can just add the additional stuff later without too much breaking.
I've decided to entirely not worry about people opening the game files and having a look around, and therefore I don't care about any sort of obfuscation even if it's a desktop app. My goal is for everything to be so server sided that modifying the client does nothing, even with full source code access. If they break visuals and make everything bug out, it just means they'll have a self inflicted poor experience. And hey, if people want to improve my client and make it better, less work for me!