And then I cheated.

At the point of terminal frustration, I imported the JPG of the revolver into Blender and cut out the shape of the grip and trigger guard like taking scissors to a magazine. Whatever. A win's a win.

Grrrr

I took another crack at making the grip of the gun, this time drawing a shape with vertices which would then be, in theory, filled in to make a polygon:

It seemed to be working until I realized I'd somehow managed to make it a 3D shape instead of a 2D outline:

Back to the drawing board.

Complicated Shapes Are Complicated

I learned how to apply beveling to round out this...shape. What is this thing, in front of the cylinder?

Spending a lot of time making sure the proportions are accurate. I want this to be a portfolio piece that shows attention to detail and discipline (i.e. it's not a cartoon ray gun).

It Took Some Tough Love

Goodbye, long sword. It's time for me to move on with my life.

Oh My Fucking God

This is going to be (another) do-over next week on the hilt, but in the process of making this piece of shit, I have some ideas for how I'll make the real one.

Complicated shapes are complicated.

On Symmetry

No pretty pictures this week.

• Took a couple of shots at the cross guard on the sword using subdivision surface modifiers. Both ended up looking like double-sided dildos.
• Research and practice with the symmetry functions (side note: GRRRRRR)
• Tutorials on above.

Well It Starts with a Bloody S.

The google image search find.

Photoshop file for reference and proportions.

First fucked up attempt using a surface subdivision modifier (Yeah, not so much.)

Working.

Finished blade.

Let Me Axe You a Question

And the fucked up render. Talk about taking a knee at the 1 yard line.

Skyboxes are Hard and For Boys

So it turns out there isn't an easy mode for adding Skyboxes to a Leadwerks scene. The easiest way is not to easy, and involves using a script to create it on the fly during execution. So LUA isn' t something I can avoid anymore.

I spent a lot of time this week reading up about the language and basic syntax. It's not what I want to be doing but I think getting my head back in the scripting game will be a good thing in the long-term.

(Incidentally, I tried just copy-pasting someone else's skybox script into my terrain scene. Not only did it not create a Skybox; the map wouldn't even compile. Sadface.)

Holy Shit

I can walk, run and jump around this fucker. Can't get a skybox to work yet. That's next.

Save Early, Save Often

I spent a lot of my working time in frustration or watching tutorial videos to alleviate frustration. However, I did successfully employ two techniques I'd been confused about: Symmetrical editing and combining/merging vertices. As with most things software-related, the devil was in the details.

This is a rough draft of a Starship concept. The idea here is that most of this thing's mass will be its power plant. I'm speculatively saying that something which can affect faster-than-light travel or simply leap between two points light-years apart will require so much energy to do so that that mechanism will eclipse whatever other functions the craft serves.

More details and fiddly-bits pending.

You really wouldn't believe how hard it is to make a completely hollow cylinder.

3D is Hard and for Boys

First, my design document:

• Re-familiarize myself with 3D modeling using  new software and "modern" techniques. I'm finding that a lot of the terminology I'm used to from a few years ago is gone, but so are a lot of the obstacles. For instance: If I wanted to make a character starting from a few basic shapes (primitives), I had to create those primitives, know ahead of time how many edges and faces it would need, and then commit to that. If, after 15 hours of work, I realized a cube (let's say the torso) should have had 18 horizontal edges instead of 16, I was pretty much fucked and hard to start over. It's become easier to make minor adjustments on the fly. Also, Blender is free and does everything and more than the $6,000 software suites I used to work in.
•  Gain a working understanding of Leadwerks 3. To put my environments and other models on display, I'll need a game engine. This one was $100, is intuitive easy to learn.
• Create an "Exploration Engine" in Leadwerks 3. As I fine-tune my modeling back to the diamond-edged efficiency I was operating at in 2005, I'll also be developing a basic platform with which to showcase what I make. This exploration engine will have:
• Third-Person character controller
• A basic character with animations for walking, running, jumping, climbing and crouching/sneaking.
• Create discreet playfields using Blender. I'll be creating the environments using Blender and importing them into Leadwerks. Leadwerks makes this insanely easy, including mapping a Nav Mesh across everything it detects as a flat horizontal surface. The character controller script uses this to decide where it can and cannot let the player character go.
• Textures? Ugh. I have the ability to(via photoshop), and a basic understand of, applying textures to 3D models. What I haven't decided yet is how hard to go on this, because it represents a significant time investment. Do I leave the models naked with basic colors and lighting to show off shapes and composition, or go all the way and create fewer but more complete environments? And when I say fewer, I mean much, much fewer.
• Add interactivity. Over time, I'll add things like switches, elevators, doors, moving trains, etc. to the environments in Leadwerks. These things are relatively simple; other things it will take to make a playfield really seem alive, like ambient weather effects, NPCs with convincing AI, etc, are significantly more complicated and time-consuming. I wonder if this is why there are so many indie games on Steam that fall into the "Alone on an island / in an abandoned mansion / after the world has ended" genre. No need for AI if there are no people.
• Build an updated portfolio along the way. All of the elements from the steps above will become separate pieces in my portfolio. This will include their own showcases including simpler set pieces, lighting and simple animations exported into a JPG or video file. 

My current challenge in Blender is to learn how to build models iteratively, so that I don't have to manually repeat identical operations over and over to build, for instance, a tower or skyscraper.

 

I wanted my first real foray back into 3D modeling to be something fun and conservative. This tank will get animated, and I'm going to give it mechanical spider legs instead of treads, because robots.

I've got the character controller working, sans the animations, although he does functionally run, jump, and crouch. That green guy on the left is him. He misbehaves sometimes, but I can compile the game and move him around, so there's that.