Last night, a defining moment in this project’s history was reached. We finally made Vitaboy, the character rendering system for the game, work, as shown in the below screenshot.
This means that the Kon Tiki milestone is 83% done, and give or take another month of development, we will have released the first running version of Project Dollhouse. It won’t have any gameplay, but you’ll be able to create a character and store it on our server.
We’ve decided to go with aws for our server solution, so by the time the first client is released, you can log onto a development server hosted by us.
After 1 year, 6 months and 29 days of development, we are finally nearing land. As I am writing this, I’m sitting atop Kon Tikis mast, seeing the first birds fly by after over a year at sea.
Also, flattr buttons are now working, so feel free to flattr away!
The entire team would finally like to extend our heartfelt condolances to all families affected by the Boston bombing.
A new year is upon us, and we have some great news!
The team has gained a new member, who has already proven his worth by doing what me and Andrew from NIOTSO had put aside because we considered it too difficult a challenge to tackle at the moment.
Check out these wonderful pictures:
Of course, city rendering isn’t done yet. Stuff like foliage, roads and actual houses still needs to be added, but this is a huge step forward!
I can also confirm that porting to Monogame is going proceeding to plan. There are bumps in the road, as always, but for the most part, Monogame appears to be quite stable and appears to have all of the features that are needed for this game. The thought of being able to provide crossplatform support is a thrilling one, especially when Microsoft’s new OS, Windows Blue, will apparently not support any apps targeted specifically for Windows 8.
Fuck Microsoft, long live Project Dollhouse!
Because I still haven’t really made any progress on rendering 3D with XNA, and just got a new phone (HTC Wildfire S), I’ve decided to branch out technology development to Android using OpenGL ES (Embeded Systems). It is a somewhat unlikely branch (because the likelyhood of Project Dollhouse being ported to Android is slim to none), but my hope is that I’ll be able to learn something that I can apply to XNA.
So far all I have is a spinning triangle using a texture from TSO, but I haven’t really gotten started on the programming yet because I spent quite a while setting up the development environment. I’ll update this post with pictures as soon as I have something interesting to show!
Afr0 Games is now also on Twitter! Come follow us!
The client now supports two resolutions: 800×600 and 1024×768 – without a border! I discovered that the graphics actually scale quite well, so I’m slightly puzzled as to why Maxis originally decided to put a border around the screen when in 1024 resolution. I suspect the answer might be that today, scaling is done on the GPU and as such is much less expensive – either way, the result is that the game now looks better than ever!
CAS in glorious, mouth-watering 1024×768!
Have you ever tried hammering a nail into a wall using nothing but a rock, with both hands tied firmly behind your back?
That’s just about how it feels trying to do any 3D operations in XNA that are even remotely advanced. I have now tried for several months to set up skeletal animation, and am now in a situation where I’ve discovered that XNA even has a built-in function for parts of what I’ve been trying to do (transforming a set of vertices based on a bone’s location in 3D space), but that this is completely useless so long as I am unable to position the skeleton’s bones correctly in 3D space.
My current theory is that the bones’ locations (which are made up of a rotation and a translation) are stored as OpenGL-native coordinates. Vitaboy (the rendering system used by The Sims and The Sims Online) was originally written in Direct3D, but it isn’t unthinkable that it was ported to OpenGL for TSO. I have tried every possible matrix and vector based operation on the bones’ translations and rotations that I could possibly think of, but nothing seems to help.
I have been putting this off for what feels as long as possible, and I’m giving myself another week, but if I haven’t made any significant progress within that timeframe I’m switching to OpenGL.
I have already tried SFML (Simple Fast Multimedia Library – an engine that allows you to use OpenGL calls and combine them with calls to the engine), but it applies some kind of funky custom world matrix that fucks up animation and rendering.
The reason I’ve been putting this off for so long is that I know that rewriting the GUI-system in OpenGL is going to take time and be a learning process all of its own, but my current consensus is that it’ll take less time than trying to fight with XNA in order to make it do something it obviously wasn’t meant to do.
I am allowing comments on this post in the hope that I can get some bright ideas and/or support.
After a lot of hair pulling, teeth grinding, swearing and performing sacrificial rituals to Microsoft in my back garden, I finally figured out how to render meshes correctly on top of textures (with some help from the Gamedev.net forums).
3D head being rendered in the client
As you can see, there’s still stuff to figure out, such as correct rotation and placement in 2D space. There’s also the issue of selecting the right mesh when the corresponding button is pressed, which is currently a little off. But the rendering is in place, and that’s the important thing for now.