next »

[chromex]

The Pitch

The Alchemy Engine is an attempt to build a toolkit that will allow anyone to create their own custom co-op RPG, be it campaign oriented or re-iterative like an MMO. Focusing on the concept of small multi-player fantasy worlds like those found in MUDs, the goal of the engine is to allow anyone to build such a game and operate a server where others can play. Builders will be able to work together to create these worlds and players will be able to connect to games through a single client.

My History

I've worked a couple of game companies on a whole range of products from level editors to mobile games to Facebook poker to modern MMORPGs. However my goal has always been to create an engine where friends can work together to build their own little MMO and I think it has strong product potential besides. Not only would it be a fun thing for people who want to build games, but now there could a variety of small, modern MUDs available to players through a single client.

Where I'm At

Well, I've gotten the core of the game engine put together and it supports rendering (forward and deferred), initial HUD, input system, level loading and saving, and networked multi-player with direct component-to-replicated-component RPC. So really I've built and tested my core tech.

As for the tooling, I just released a first version of the static modeling tool. Its very pre-alpha in that its far from feature complete however it does work and you can create models and export them for use in Unity and other engines.

What I'm Working On

I got some feedback on the tool and I'm integrating it! At this very moment, I'm working on integrating FXAA to help with all the jaggies and a broader integration of Lua to allow custom tools and scripts.

As soon as the new tools and visuals are ready, I've got to write an installer script and look into getting signing set up since the lack of both was a major no-no for a lot of people with my last build.

More Info?

I have a project website at http://alchemyengine.com/ with info, downloads, and docs. I'll be posting regularly and I am a huge believer in release early and often.

For justice!

Too many projects lack transparency and when working more of less alone its too easy to get stuck in my world view. If you see something that you think is dumb, or that sounds iffy either for the the engine or its business case, please let me know.

[chromex]

I got to spend half of today integrating FXAA with my render pipeline. I want to record some of the adventure here because finding all the necessary bits online was harder than it needed to be.

To start, FXAA is an awesome anti-aliasing algorithm. It's elegant and fast. It may no longer be the best-best-best but its still the best-best. Secondly, the code is public domain. This is awesome. Finally, the published FXAA code is crazy easy to integrate, especially if you have a post process pipeline of any kind.

If you don't have any anti-aliasing and think it sounds awesome and you have an NVidia card you can try it out no effort necessary! Crack open your NVidia control panel and force FXAA on. It'll perform an injection of the process at the driver level on any renders that you start up. I used this to evaluate if I wanted to pursue integration. And I did. Be sure to turn it off when you are done though because any games you run will also have it applied and double AA can lead to a blurry mess.

Now for links.

To understand FXAA you should go here:
http://developer.download.nvidia.com/assets/gamedev/files/sdk/11/FXAA_WhitePaper.pdf

Reading that you might Google the author's name, especially since he used to have the code posted on his blog. However he nixed most of his blog history so a bajillion FXAA links all go to the same 404 page. Which made me a sad .

I eventually found FXAA 3.9 code here:
https://docs.google.com/file/d0B2manFVVrzQAZWEyNzcwZjYtMTkwNi00YWVjLWJiMTAtZGNkN2FlMmFmODEx/edit?hl=en_US

And this was awesome. There is a newer version of FXAA somewhere however I didn't need its improvements so I stopped my quest.

The linked FXAA code supports many options including a PC option and a PC option. These were the two I was interested in. To integrate it though I have to deal with a bit of an oddity. My engine is used both for the game and the tools. The tools are Qt based and, as of Qt5.0, Qt is limited to OpenGL 2.0 since it's an expected baseline across all platforms it runs on. However my game uses deferred shading and is quite fond of OpenGL 4.0. I account for this by having both a deferred and forward render. Now FXAA option is perfectly happy with OpenGL 4 (and it looks crazy good ) and FXAA is perfectly happy with OpenGL 2... except it looks wretched. I mean really bad. It might look okay if you had a fully painted world where over blurring and random color seepage were not as noticeable but in a tool with sharp lines and solid cubes... nasty

So my conclusion is that the game gets gorgeous FXAA, which really helps in a voxel world, and the tools get nothing. It's kind of frustrating since it really hurts how the tools demo but I've gone into my NVidia control panel and forced FXAA on forever for the tool exe so that it at least looks alright for me.

Anyways. Onwards. For victory.

[OneMoreGo]

A cool idea.  I'm nostalgic for MUDs too, that's part of the inspiration for my current project. I started out with just plain text, then it started getting more complicated heh. 

[chromex]

Hehe, I know what you mean. I've started this project an untold number of times as a side project. Finally quit my job and came up with a rigorous design and production path for it and started over.

MUDs were an interesting part of gaming history because the community was so involved. As time went on people seemed to want more catered experiences but at last Minecraft seems to have brought people back around. Normal players can contribute, and can do so creatively, to a games ecosystem. And they are doing it at a large enough scale that I think it might finally be possible to convince people that they'd like doing that and not get the