Tiny but lots of steps

Ever since taking on my minimalist stance of just getting everything working no matter how bad it works, I've been making great progress! This has probably turned out to be my most important lesson over the years. I could fuss on making the sunset perfect for months if I wanted to, but that's no way to go about developing a fun game.

This week I got basic saving and loading in, a simple starter screen, monster spawning, resource generation (requires some tricks for terrain based trees), weapon damage, some crafting features, encounter spawning(towers), and a simple recall spell in. I did spend a little too much time messing with biomes and terrain generation, but I learned some lessons.

This isn't the best screenshot... but I don't like posting without at least one picture. Pretty much everything you can see in this shot in unfinished, but the basic programming is all in and working. I have a speedtree license, and hope to use all original trees, but for now have some basic Unity Trees spawning in.

Dimension crystals, like the one in the screenshot, are the basis of long range travel, respawn points, and several other features. That's mostly what I hope to be working on next week. I have a 2d minimap I was using to show crystal locations throughout a world, but I might change that to an easier to use list.

The towers in the screenshot are far away, but you can absolutely run straight to them and climb up them. They don't spawn their own monsters and loot yet, but that's their main purpose and we'll get there soon enough. Eventually I plan to have different towers for each biome such as pyramids for desert, and endless mineshafts in the mountains (which might be hard to see from far away).

The biggest problems looming on the horizon at this point are monster movement/combat AI and pathfinding. I'm pretty confident in the rest of what I'm doing, I might be looking into some third party tools to cover those that area up rather than take up a bunch of my time at this point in development.

Lots of tiny steps, but all necessary, and adding up towards the big picture!

Monsters, Programming, and Epiphanies

As a crafting heavy game I've spent hundreds of hours thinking of, taking notes, and testing ideas about how to make crafting more controllable. One thing that I knew I wanted to improve on from the beginning, though I didn't figure out how until this week, was a problem made very clear in The Elder Scrolls V: Skyrim. I love the Elder Scrolls series and of course enjoyed crafting my own gear in Skyrim, however, with limited resources and a little testing it quickly becomes clear that repeatedly crafting daggers is the best way to level up smithing.

Now, we're not trying to make a "realistic" game, this isn't a simulator, but crafting 300 level 1 daggers to become a master smith just seems silly. But how do you do better? Game development is not easy. I tried requiring five different items crafted per level. I tried requiring different categories within a profession to be leveled up individually. I tried many things. But ultimately it comes down to how much metal did you bend. If a 3 metal dagger gives 3 exp, why not make a 6 metal sword give 6 exp. We also have tiers of metal, and other factors, but I'm finding this to be a -now- obvious system, though it took a long time to stumble across.

Speaking of crafting, I rewrote all of the crafting code this week into a much more compact system, in some cases reducing over a thousand lines of code into roughly one hundred. How? I'm now using what are known as machines in the code. Rather than "hard coding" the stats for a Sword, a Longsword, and a Shortsword each individually, several variables are input and the machines outputs what is needed. The greatest part of this method is that a new item can be added into the game with one a few lines of code. I hope that this eventually leads to opening the door to community created items by the dozens.

Monsters are in and working well. Spawning with different, skills, stats, and colors based on their biome. They're not the brightest when it comes to pathfinding, but pathfinding in a procedural world is never easy. If anyone is a genius in that field please let me know ;p

The UI has improved a lot. I hope to get more tooltips in over the next week to show information when requested rather than filling the screen up with levels and stats.

My wife and I are buying a house! So I've been pretty busy. And Terrain Composer 2 has been delayed, so I'm not currently working on level graphics. I hope to have more flashy graphics to show off soon =]