New Lugaru Campaign
Hey guys, skraeling from the Wolfire Forums has made a sweet Lugaru campaign called The Seven Tasks. Check it out here.
I just played through it and can vouch for its coolness. Especially the bonus zombie level. The quality and quantity of Lugaru mods always amazes us, especially given the lack of any real modding tools available for the game. Stuff like this is what has motivated us to work really hard on making really powerful mod tools for Overgrowth. In case you missed it, here's the Overgrowth map editor.
In other news, I made a quick Lugaru mod browser. I hope to use this for my upcoming post about Google App Engine.
What's cool about it is that it has a JSON interface so anyone can grab that data and make their own mod browser. Joel Levin has already started work on a hot Cocoa, OS X mod browser that will manage your mods and auto-install them for you. Please contact us if you'd like to make a Windows or Linux counterpart.
More on this, App Engine, and ultimately the hot Overgrowth mod browser, later!
Forum Management
Wolfire has a long history of taking a relaxed approach to forum management. We do in fact have rules but because we don't refer to them or enforce them much, most people don't know they exist. The precedent, with few exceptions, has been that pretty much anything goes. However, since our official start on Overgrowth last September, we've seen a gradual but steady rise in active users. While it's hard to objectively quantify "forum happiness" it appears that as the number of active users increases, so does the likelihood of heated arguments (indicated by a small rise in moderator warnings).
The Wolfire laissez-faire approach seems to work just fine for a smaller community. The question is, as we continue to grow, should we start to think about stricter rule enforcement policies?
A Case For Vigilant Police Work
-Sometimes threads get stuck in a negative cycle that can spiral out of control. In these situations, a little moderating can extinguish the flames before they spread.
-Not all topics are useful or productive (that's why we have a randomness section). Tragically sometimes useful topics turn silly. While Meebo silliness is fine with me, the forum discussions are public and have an influence on the community as a whole. In extreme cases, crazy threads might discourage newcomers from being active in the Wolfire community.
-Isn't it nice to have authorities around that can deal with trolls when they are discovered?
A Case For Relaxed Enforcement
-Everyone tends to be very well behaved already.
-Freely exchanged ideas can lead to fun discussions and awesome brainstorms. Any restrictions on forum discussions also restrict creativity.
-Forum rules are always hard to apply objectively. It's tough enough on the internet to distinguish sincerity from sarcasm.
-Sometimes it's best to just let trolls dig themselves into a ditch. Yes trolls bring negative energy to the forums but so can locking or deleting threads. The community is already good at self-monitoring.
-With a relaxed approach, the Wolfire team gets to spend more time developing and less time policing :).
What are your thoughts on the Wolfire Forums? Do you think they are generally a fun and positive place where you can share ideas and learn about Wolfire Games and game development? Are there any forum Utopias out there that we should know about and learn from?
Creating Our Own Skydomes
For Overgrowth, we are photo-sourcing all our skydomes. So far, we've been searching sites like cgtextures.com to find these images. This means we are limited to the same small set of decent skydomes that countless other game developers are looking at. We don't want to be limited, and we've noticed that the skies here in Berkeley are often more dynamic than most. So, we decided to look for a way to make our own skydomes.
What we've found is an amazing research project that resulted in a photo-stitching app called Autostitch. Autostitch has several advantages over other means of capturing panoramas. First, unlike heavy duty methods that use motorized mounts, it requires no special hardware (other than a camera) and it's not expensive (in fact, the basic version is free :)). The fact that it's free and easy to use is especially important to us since we want all our tools to be as accessible as possible to modders.
Autostitch also improves over standard software methods for creating panoramas. Most photo-stitching programs can only handle 1D, cylindrical projections. This is great for horizon panoramas, but is not adequate for 2D, spherical skydomes. In addition, photo-stitching apps often require some user interaction; for example, the user may have to arrange the photos in order. Autostitch addresses these two shortcomings. It can handle full hemispherical projections, and it does so entirely automatically. With a handheld camera, we can just haphazardly shoot a bunch of photos all over the sky, and Autostitch does the rest!
This evening, John and I drove up to the Berkeley hills to capture the sunset. Here's how we did it step by step:
Step 1
Take a bunch of photos. Try to evenly cover the entire area you want to capture. No need for a tripod or any special technique.
Step 2
Run Autostitch on the photos. This just takes a few button presses.
Step 3
Edit and extend the panorama to fill in an entire hemispherical sky.
Step 4
Load your new sky into Overgrowth! (Eventually we'll have an easy way to do this. For now, if you'd like to try this in the alphas, you'll have to edit the <sky> tag in the level xml file, and delete the terrain caches.)
Now that we have this technology, I'll be keeping an eye out for interesting skies and capturing them as they come along. I've found this process quite fun. If any of you guys want to help us and try your hand at the process, we'd really appreciate it. Feel free to send us your sky images via AIM / MSN or on the forums.
What do you guys think about this method for creating skydomes?
How I Make Overgrowth Assets: Part I, The Idea
This is the first part of a series of blog posts explaining how I go about putting a 3D art asset into Overgrowth.
Everybody has ideas about what they want to make but how do you know your ideas are good ones? If you don't have any good ideas, how do you get some?
In a lot of ways starting is the hardest part of making an asset. One thing I think is important for this kind of thing is thinking backwards through the steps to get to your goal. Lets say you want a part in the game that is scary. You need a bunch of assets to put in the game, which means you have to texture and model a bunch of scary looking things, and so you have to design things that are scary, which means you need to understand what makes people afraid. The best way to understand fear is to think about your own personal experience of it.
So the first step of design is understanding other people and personal experience. Ironically, while this is very simple, it seems like one of the main things big name developers get wrong. Often instead of referencing anything personal or real, they reference other games or movies. These copies often miss what made the originals great, or even if they do get everything right it still feels like a stale rehash.
One of the main things I want to capture in Overgrowth is the stillness and freedom of being in a natural environment, far away from other people. I love taking road trips and hiking, and so I drew from those experiences. There is always the risk that your audience will not have shared your experiences. It is important to get down to the compelling details that made your personal experience memorable to you in the first place. That kind of authenticity is interesting to almost everybody.
Obviously it is not practical or even possible to experience everything personally that you may want to put in a game, but research is a pretty good substitute. If you want to make a game about combat, maybe you can read memoirs of people who fought in wars, or talk to people who are war veterans about their experience. I read a lot. Museums are great. Also, Google image search is an invaluable resource.
Next time I will talk about how to take these vague ideas and turn them into a concrete concept. What kind of personal experiences have you guys drawn on when making your own projects?
Terrain edge fading
While the terrain in Overgrowth is quite large, the view distance can be even larger. On a clear day, you can see all the way from one corner of the terrain to the opposite corner. While this allows us to show impressive vistas (like those in our alpha trailer), it also means that you can often see the edge of the terrain. We found that the sharp edge of the terrain really hurt the impression of being in a real place. Here are some examples of the edge of the terrain:
There are many options to solve this problem. Some common solutions are to make each level into a small island, make the terrain much bigger, obstruct the edges with trees or mountains, or funnel the player through a series of canyons and set pieces. However, all of these options would either restrict our creative control, or greatly add to our development time and system requirements. For now we decided to automatically fade out the terrain around the edges to create a smoother transition to the cube map terrain. Here is an example of one the faded textures:
There is still a noticeable 'end' to the terrain in each area, but it is much more subtle than the hard line we originally had. Here are the same views as above, with the new faded edges:
I used a much more extreme version of this effect in Lugaru to give the effect of a long view-distance. I created a skybox of a very high-detail terrain, and then instead of using the OpenGL colored fog effect, I created a 'transparent fog' effect by manually changing the terrain vertex alpha values every frame. Here is what Lugaru 1 looks like with normal fog and with transparent fog.
I think this was a very successful effect! It gave the smooth framerate of a close view-distance without feeling claustrophobic or oppressive. Since the game was based around close-quarters fighting, I don't think it really mattered that you couldn't actually see for miles.