Finally getting some time...

[Ian Badcoe]

Hi all,

It's been a while, but I retired a little over two weeks ago (unless I get really bored) I've been a professional GameDev for the last 9 years (and a further 5 or 6 years, back in the '90s...) but now I plan to finally write the games I want to play...

And, as you'll know if you remember me at all, that will _definitely involve PGC...

I drew up a list of the first 5 game ideas I had been brooding over for some time, and when somebody asked "how many involve procedural content?" it turned out to be all of them.  I didn't plan it that way, but...

I am writing them up here: https://emnf-ideas.ianbadcoe.uk/welcome although I only got as far as expanding the first when I got distracted by wanting to prototype it...

I've been revisiting subdivision surfaces:

(The bottom image was just me pushing the limits of my control-mesh construction system...)

And a thing I like about those is I can make tubular environments (top image) and they are mostly laid-out with a square grid, which will be very useful for positioning gameplay elements...

If you read "Turrets" off my ideas page, and think about the top image as being a section from the inside of a "run":

• the small (smooth) hole in the foreground might lead to a side-tunnel
• the large holes in the roof might admit shafts of light... with different flora growing beneath
• this is just subdivided twice, but I am thinking that if that were a progressive refinement process, where I stop between subdivisions to further tweak the mesh:
• parts of the environment destined to be rock might have some random offsets added to make them chunky
• playable areas might be made flatter still
• additional detail (such as raising or dropping the playable areas relative to the surrounding terrain, injecting a sharp vertical edge, or adding cubical "stairs" to the side-walls
• and obviously the whole thing needs texturing/decorating/populating

Anyway, that is just one idea I've been toying with, I've also historically played with 3D Voronoi processes, which have the nice properties that everything is convex, and you do not have to divide up the whole of the space: in an unvisited area you can leave space empty, and then as the play approaches inject new centers, which can peal pieces of it off for generation...  and quite complex subdivisions can be done in real time pretty easily...

Anyway, just some thoughts, just putting this out there to see if it triggers any interesting discussion at the moment...

Cheers.

Ian