I've been developing a “small” engine and an fps game along side it. The game is to insure I'm adding everything I can think of that would be needed to complete the fps game into the engine. The thought had crossed my mind that I might want to open source this fps project as an example of how to use the engine. However, some of the assets I have been using were purchased from places like cgtrader/turbosquid/etc, and are not meant to have the source files (.fbx in this case) distributed freely (obviously as they're how these artists make a living).

Therefore, I am considering my options as to how I might go about open-sourcing the project, and was wondering what others thoughts might be on the below, or if there's an option I might not have considered?

1.) Contract content creators to develop assets specifically for this project knowing they will be releasing the source and that it will be under an open source license - this will be costly to achieve the same level of detail I've found in some of the models from the above mentioned sites (guns for example, you can get a highly detailed version of a glock with full pbr materials for like $30 or less). You might think, “well it's just a demo who cares”, because that's crossed my mind as well, but if the first thing a person sees from the project is something aesthetically pleasing, THEN they see that it's also something they can download and hack around on…they're more likely to check out the project. So while the quality of the assets might not directly correlate to the capabilities of the engine, first impressions are important

2.) Compile the assets into a proprietary format which works only with the engine (this will need to be done down the road for building production releases, but so far I have not implemented, the engine strictly loads .fbx files). However, I'm not sure if this would be appropriate or not, as you would still be distributing the assets freely, just not in a format that could be loaded into blender/maya/etc.

3.) Release a proprietary licensed compiled “Commercial” version and sell for like $1 or something (this wouldn't be a full game, just a vertical slice, thinking about turning it into a revamped e1m1 of doom, something that small). This way people could watch a video of the demo, see the full source (minus the assets), and then if they wanted to purchase the demo, they could have the assets required to compile from source (or use their own assets, but that would eliminate the instant gratification in seeing it work before diving into how it works (you know what I'm talking about, that feeling you get when you find something interesting on github and are able to download it, compile it, and it just works…that feeling))

After having typed this all up, I'm kind of leaning toward option 3. What are your thoughts?

Thaumaturge said:

Have you looked around for open-source 3D models? I don't know how much there is to be found in that space, but perhaps, with some digging, you might find some such that work to your purposes.

I've been googling and finding some random sites, but have found nothing of the quality I'm after. Open to any suggestions of any asset sites that might have what I'm after if you know of any.

Thaumaturge said:

Now, you mention that price is an obstacle--and believe me, I very much sympathise. That said, since this is for an engine demo, it seems to me that you might be able to keep the scope small and thus perhaps reduce the number of assets required.

That is potentially possible. I really only need an arm rig, a pistol model, and a shoot animation, along with one basic enemy type. The environment I model myself, and finding CC textures or creating them from images is easy enough. Character modelling and animation on the other hand, while I could do it, what I would output would take forever and be nowhere near the quality that someone who is training in creating such assets could produce.

Thaumaturge said:

The “commercial demo” option might work… but it seems a little over-convoluted for this situation, to my mind. I suspect that the more hoops you put in front of your potential audience/developers, the fewer are likely to try your demo.

The fact that you say it seems over-convoluted is enough for me to write that option off. I really want this to be as easy as possible to download and start working with. I even went as far as to learn cmake enough to setup a “Super Build” which pulls in all external dependencies at their correct versions (for the engine). So all you have to do is download the source, run cmake, build, done.