Action RPG Framework

BYOND is not a general purpose programming language, it's specifically for making games. This makes it possible for BYOND to make game development easier than C++ and Java could because those languages are general purpose - they don't provide you with maps, networking, movement, and procs like view() or obounds(). When it comes to game development, BYOND *is* general purpose. You can use it to make action games, RPGs, platformers, board games, card games, racing games, and realtime strategy games. This is a good thing but it also limits how easy BYOND can make game development.

It's not because BYOND is lacking. Look through the feature request forum - there are lots of good features being requested that'd improve BYOND's capabilities or provide conveniences to developers, but very few of the features would really make it easy to create a game. Even if all of those features were added you're still going to have to write 2,000 lines of code to make a game. That's just the cost of being flexible enough to produce such a variety of games.

If you want to have a development environment that truly easy to use, BYOND can't take you 100% of the way there. It's a solid foundation but it only takes you 70-80% of the way there. As you might have guessed already, frameworks are the way to bridge that remaining 20-30%. By making a framework for a specific type of game you can provide features that BYOND would never provide itself. For example, to make it easy to create an action RPG with BYOND you'd need to provide people with a health var and on-screen health meters but these aren't things BYOND would provide by default.

Here are some benefits to having and using frameworks:

1. Frameworks are appealing to outside users and current users. Showing non-users an example of a BYOND game lets them see what BYOND is capable of but compared to other tools, that'll never be that impressive. BYOND will attract people by being easy to use and frameworks provide a way to show non-BYOND users how easy it is to make a game. The current BYOND community has also responded positively to frameworks so there's no risk of alienating current users (except Falacy).

2. Frameworks let you accomplish more with less time, effort, and experience. They're not just something that inexperienced users can use because they can't make things from scratch. They'd also let more advanced users create games faster. Also, the more easily you can do things the more a tutorial can cover.

3. Frameworks provide topics for tutorials. ZBT is considered to be a good tutorial but it takes 6,500 words to cover some very basic things and, in the end, you've only got a verb panel and a couple of verbs. If you write a tutorial about how to use a framework it too will cover more in less time. The reader will get a more rewarding result in less time. If BYOND game development is supposed to be easy, tutorials that cover simple topics shouldn't be giant walls of text that only show you how to make a verb panel with two verbs.

4. Frameworks make it easier to develop resources. It's hard for people to write tutorials that cover the basics of how to use DM because it's so open ended. Writing a tutorial to show how to create a board game using a Board Game Framework would be pretty easy. It's also easier to make demos that can be integrated with other projects. If you make a demo in DM now, you typically have to build a game around the demo to make it work. If you write a demo that's made for a framework, it'll be able to integrate with any game that uses the same framework more easily.

5. Similarly, they also make it easier to help users. If someone posts a question in the Developer Help forum you might need a lot of background information about their project. If you know what framework they're using it can greatly simplify things. By being familiar with the framework you know a lot more about their project than you would otherwise. It also elevates the level of questions that people can ask. Currently if someone asked "how can I make a building in my RTS that shoots at enemies?" that's an awfully complicated question that nobody could easily answer now. If the developer was using the RTS framework the answer might just be a few lines of code.

6. Frameworks are something the community can work on. The BYOND staff has a lot on their plate and there's really no way for us to help. Frameworks would be projects that any user can contribute to.

7. Frameworks can drive BYOND feature requests. It's one thing to say "I'd like for BYOND to have this feature because I'd use it" and it's another thing to say "I'd like for BYOND to have this feature because it'd let us add this other feature to a framework that 100 existing projects are already using".

8. The presence of frameworks as part of the BYOND package itself gives BYOND a way to guide game developers. People who have no game development experience would have a way to get started and wouldn't be sucked into this culture where you make a base icon, then a title screen, then post on the Classified Ads forum asking for a turfer, GFXer, mapper, and coder to be part of your team. Also, with a way to literally make a playable game of good quality in under five minutes you can create actual standards for BYOND games.

The only catch is that the frameworks can't just be another resource that people may or may not ever find - they'd have to be treated as part of BYOND. It'd take some work but it'd make a huge difference. Here's all it'd take:

1. A menu that's shown when creating a new environment in Dream Maker that asks the user what type of game they'd like to make. Picking a genre creates a project and automatically includes the appropriate framework. There may also be options to include a sample game that would add a few files to the new project - this would let people compile and run a game to see what the framework provides. There'd be an option to create a blank environment if you don't want to use any framework.

2. Reference pages for vars and procs defined by frameworks. To the developer there shouldn't be any difference between using a proc that's part of a framework and using a proc that's part of DM.

3. Dream Maker might need a way to download content from the hub. Dream Maker could come with some version of the framework when you download and install it, but since BYOND and the frameworks may be updated independently, it would be nice to have Dream Maker able to download the latest version.

4. The staff might have to come up with the list of genres and spend some time reviewing the frameworks before including them in BYOND. If the community is developing these resources, the staff could get a lot of help from the community with this work.

There are many things you could change about BYOND that'd have a positive effect, but simply having a positive effect isn't always enough. Some things would only have a minor effect, some might take a long time to develop, and most wouldn't do much to make it easier to create games with BYOND. The Flash client for example may have a high impact but it also has a long development time and doesn't make it easier to develop games. Other features that would increase BYOND's capabilities (like adding a 3D display mode) are similar - large impact, long development time, no easier to develop games. The changes described here would have a huge impact, small development cost (by the staff), and would simplify game development more than any other feature I can think of.

Thoughts?