Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[GalenEvil]

Ah okay, that's where my lack of research failed me. I didn't know that rain shadows followed the terrain. I thought they were just like cloudcover over an area. In that case I will go with reseeding to the terrain seed. I think I can keep it decently low res if instead of say a 513x513 grid (terrain) I use a 129x129 grid and scale it up to 513x513. That would be a reduction in detail of 4x, and would lead to a similar distribution of data points over that grid. Another way to do it would be to keep a 513x513 grid for the rain shadow but instead of going all the way to stepSize < 1 just go down to stepSize < 4 and then fill in the holes with the square's average values.

[Virex]

Well, rain shadows don't exactly follow the terrain, but you need the terrain data to compute them. What happens is that rainfall gets higher at one end of a mountain, because the moist air rises to colder regions and gets compressed, while it's drier on the other side because the air there has largely been drained from water. You could possibly simulate this with a cellular automaton, in which the cells on one side are given a high moisture level and each step they have a chance to either rain down or transfer the water to the next cell, with the chances depending on the height of the cell.

[GalenEvil]

ah okay. sounds like cellular automata might be the easiest way to go about it. Other ways that I have read about involve implementing wind and having each point of land raycast in the direction opposite the wind until finding water, then backtracing to see if there are any obstructions (mountains) that would prevent water from getting to that location.

[Skyrunner]

All this advanced programming makes me confuse. @.@

* Skyrunner backs away slowly, not showing fear.

I'm sure you're all doing fine.

*bolts*

[GalenEvil]

Heya, just letting everyone know that tomorrow at the meeting we will be discussing how to streamline the Game Design Document. It will be a breakdown of milestones that we need to complete and how to go about getting to them. If you would like to participate the meeting will be starting at the same time as usual: approximately 1400 hours Central Standard Time, 2000 hours GMT. The meeting will last until everyone is tired of talking about anything relevant.

The meeting will be held in irc://irc.newnet.net/CodeJam so if you want to come by and participate that's cool. If not then hopefully we will be well stocked with the full current dev-team.

I also hope to have a small Q&A section for anyone that is not part of the dev-team once we finish talking about needed revisions. So if you have any questions about the project that you want to ask us please wait until the appropriate time.

Thanks for reading this! See ya'll tomorrow!

EDIT!!! Since I don't want to double post...

We added some new stuff to the google doc to prepare and organize an alpha candidate. It is mostly going to be a "proof of concept" than anything else. You can check it out at the link below!

Updated Google-Doc!

Things will get fleshed out as needed between now and release. Also, I hope that we can release some pictures in addition to more regular updates to progress.

[Kofthefens]

I have done a few more things on the doc, most notably (perhaps) the picture of what it looks like at the bottom. I also added "List of people - Revised" Add your name to it if you are still interested in working on this.

Mulletmaster - thanks for all the models. I added a few more I would like to the bottom of your list.

[Mullet Master]

Meeting tonight, at #codejam.
I have a few things I want to talk about, especially getting graphics prepared for the alpha.

[Skyrunner]

Textures offered by me:

Spoiler: Unknown silvery ore (click to show/hide)

Spoiler: Genericrock (click to show/hide)

Spoiler: Rust? (click to show/hide)

[GalenEvil]

Sorry, Internet has been dark for past several days. Need to do a few finishing touches on the Rock Generator that you asked for Kof, and I have orthogonal direction ramps done for terrain. Working on diagonal ramps now.

Details on Rock Generator:
*  It currently only generates a generic rock object.
*  On my computer it will generate up to 600 colliding rocks before it starts to chug along slower than 40fps (approx 450k vertices). If I have them just falling I can get a steady 50-60 fps at about 950k - 1M vertices using 3 generators with only spawning collisions.
*  It currently spawns rocks TOO FAST and they go flying all over the place so need to slow it down just a tad and get them to spawn a set distance away from the generator in a circle.

Terrain Details:
*   Spent a little time playing with cubes and figured out how to mess with the meshes to get them to turn out correctly for a basic cube and the 8 orthogonal ramps (NESW in Up and Down directions!)
*   Need to work on diagonal ramps for (NE, SE, NW, SW: Up and Down directions) but that shouldn't be too difficult since it just requires one more standard alteration and some extra rotations.
*   Need to apply Materials and Textures to them so that I can see how they look
*   Need to make them affected by Gravity only if they are not supported... that part sounds fun

Skyrunner: Cool textures

[Virex]

The last point isn't needed for the Alpha yet. As for spawning rocks too fast, look into Yield, which allows you to implement the spawning routine as a coroutine, so it only loops once every n steps of the simulation. Or just have a spawner object that creates one object ever so often using it's update() function.

[GalenEvil]

Yeah, it's a "for later" thing so that I have somewhere to remind me what to work on. I think I should take breaks from the internet more often. I have gotten a hella lot more done in the past two days then I would have with the use of the Internet (other than some research that I badly need to do >.<)

Oh, and as a side note about another project that I am working on: Trying to keep track of entire populations and their family trees, and making sure that they don't make family wreaths instead of family trees is actually pretty taxing on a computer. At least if you try to do it on a day-by-day basis program wise. I think I can get up around 200 years and 15-20k population before it really bogs way down and I have to quit my simulation... >.>

[Mullet Master]

From the Testing Files :

Galen's Random Ore Generator asplodes!

Spoiler (click to show/hide)

[Scelly9]

That is some wonderful random you have going there 

[Kofthefens]

Forging is up! You can mine, smelt, and forge now. This is shaping up. However, the things you can forge don't have models. I only have the model for the mining tool- and that works great. However, the other jobs won't create anything until I get some models, specifically models for the robot head, arms, legs, torso, tank, construction tool, deconstruction tool.

New picture:
http://imgur.com/PEqK6

[GalenEvil]

I am just about done with the rocks for now. Once I can take a look at the job related code to let the robots pick them up it should work! Also, got my rocks textured now, still random colours at the moment but that's so I can play with them.

Terrain is ALMOST done for first stage! For the testing room I want the room to be able to be defined via a text file with terrain type-values instead of using a heightmap since it will be a bit still before I can get a predictive generator going with my terrain models. I have all 17 terrain direction types now as well!

I will probably NOT be able to make it to the meeting tomorrow. Internet on my end is still being really really bad so I am kinda lucky to even be able to post this message I will try to get onto the channel in a little bit though.