I am now officially on hiatus from Second Life. What’s that mean? I dunno, but when TV Guide said my most favorite TV show of all time was ‘going on hiatus’ I knew I’d never see it again. :(
WARNING: totally depressed whiner ranting post ahead. it may get messy.
I’m going into another pretty heavy depression, and I can not hack this any more. MESH. I got into the mesh pre-beta, the ground floor, working with mesh since like March of last year. A great head start. And I am STILL at square ZERO! Why? Because my life is JUST that effing hard, and nobody wants to help me.
Okay, okay… a couple of people… sporadically might sorta want to help me. To be honest. It ain’t helping, though. Let’s take the wonderful rigged mesh avatars! It’s very easy to do, so I am told. You use the right version of Blender (which I do); you click this, this, and this; you hit export, and it works. Except… IT DOESN’T!!!
Maybe it works for, say, THE ENTIRE UNIVERSE, but it just does NOT work for Bloodsong. No. There is some kind of Cosmic Law that says Bloodsong’s life has to be HARD. Bloodsong cannot just press three buttons and have a working result. Oh, no. Bloodsong has to spend THREE FREEKIN WEEKS experimenting with these few little steps; getting bugs, errors, or no results; filing reports; only to find out, NOBODY can reproduce these results. Sure. The universe is purring along fine for everybody else. No problems, there.
So I can’t effin’ rig an effin’ mesh avatar. Fine. I have another big project I have been wanting to do for years. I want to make rideable horses. These are still attachments, so the prim count is not (yet) limited. And with mesh, I can get beautiful details! No more sculpty peanut-heads!
Except I stopped to think about one thing…. You understand that each mesh model has a ‘cost’ or ‘weight’ in prims. And you know high prim count is bad. If you take a sculpty (one prim) and convert it into a mesh, it actually costs 4 prims. So, to get a junky, detail-less peanut-head, it’s 4 prims, and to get even three or four times the detail of a sculpty, you’re talking 12-16 prims.
Legs were coming out to about that much, let’s say 15 prims. There’s 4 legs, so that’s 60 prims just for your horse to stand there. Then, of course, each leg has several frames of ‘animation.’ Let’s say five frames of animation while the horse walks or runs. That is 4×5=20 separate legs at 15 prims each is… 300 prims.
Think about that number for a second. First of all, back when we didn’t have sculpties, prim horses made of spheres and cylinders and toruses did not have that many prims for the legs! In fact… since the maximum size of a linkset is 255 pieces, 300 prims is more than the entire prim count of a full-prim horse! And this is just for the mesh horse’s legs. We haven’t EVEN started adding the bloody flexi-prim mane and tail!
Here’s another Second Life Fact you may not know (because I don’t think Wynd published this finding anywhere)…. Switching a prim’s alpha (transparency) state is one of the highest script-lag and render-lag inducing things you can do. Flipping alphas, rapid-fire, on the one (4) prim sculpty legs caused a noticeable performance hit. Still, it was better than trying to simultaneously flip the 5 or 6 prims that made up a prim horse leg. Now you’re going to ‘blink’ what essentially amounts to 15 prims at a time? Wouldn’t this cause a severe performance hit?
So I asked it at the Mesh Pre-Release Office Hours. The answer? Um, yeah. Yeah, it would.
So I shelved that project. But you know what? Does anybody even pay attention to that? I don’t think so. I think that people are probably developing these lovely full-mesh rideables, and don’t give a hoot about the script lag or performance. No, seriously. And if I were just so ‘ignernt’ and didn’t care to make my rideables actuallly WORK in real situations… I could, too. And things would be a lot easier! I wouldn’t need to worry about causing sim lag, or viewer lag, or script lag. I wouldn’t need to worry about making my stuff compatible with, say, jousting lances. I wouldn’t need to worry about how to allow people the choice to have a speed-boosted horse without them being banned from races for fear of cheating.
Hell, all I’d need to worry about is making a pretty horse that looks good in the product shot and gets people to plunk down the moolah for it! Does it lag? Blame Linden Labs! Does it slow down your client? Get a faster computer! Why do I have to have this Vision of something Perfect, that would be Really Cool, and people would enjoy? I mean, REALLY have fun with? And would Look Killer! And be like riding a real horse — or even better!
Because the World is Not Perfect. My freekin PEFECT mesh creation app can’t even bloody UV map. Well, it can, but it creates non-human-editable uvmaps. (That means, Lady Jess can’t paint a pinto horse flank on a square map and have it work.) It won’t even respect the uvmaps I import the object with! Then my app that can unzip perfect square maps can’t work with the bloody triangles the other pefect app creates. EVERY DAMNED THING IS A FREEKIN PROBLEM!
So….. yeah. I need to chuck it all in the trash and take a break from it. It’s just grinding me down. I have to give up.