Sprite Kit is a "no engine solution".
The correct term for it is framework: Sprite Kit is not an "Engine". You hear that as a common term around game dev circles: "what engine are you using?". Sprite Kit is a "no engine solution". In this article I lay out my reasons for choosing Sprite Kit for my professional studio's development, and what this distinction means in practice.
Editing a Sprite Kit Scene in Xcode |
In game dev you try to separate code from assets, functionality from data, as you do in regular programming or good software engineering. A key difference is the degree of separation and what role your game plays on the device when you use a framework rather than an engine.
- With a framework you are actually writing a computer program, in essence, and that program is an executable, which runs on the platform making calls via a framework API to load assets and data. There may be tools to create data like levels and maps into a form that the framework API calls can consume. But those tools only exist as part of a set of programmer's productivity helpers, and typically its possible to edit that data using other tools if needed.
- Essentially in the engine case, your game is NOT a program, its all data. The engine is in charge, and you provide your game as a package of data for the engine to operate on. The engine has a combined editor and previewer that encapsulates the entirety of your creative process, capturing your scripts, assets and other inputs into a package and giving an almost-real-time view of the game.
What is it that we get from game engine makers, like GameMaker, Adventure Game Studio, Unity3D, Unreal Engine or CryEngine? They ship pre-packaged functionality, such as "place that quad over there with this texture on it", and "apply that animation to this object node". You provide a package of data, which includes visual assets such as images & models, and logic assets such as scripts, that the game engine loads and at various times to render and evaluate.
Framework eg Sprite Kit vs an Engine |
- With an engine
- some incredible things are easy, and
- easy things are (often) strangely hard.
- publishing/shipping is way harder than it should be
- With a framework API
- everything is possible, and
- easy things are (often) a bit hard
- shipping is what they do best
These attributes mean that engines are great for prototyping (as long as your prototype is a game that the engine can easily make) but things get hard quickly when you try to exert your will and design intent over the game; and also when you start to get closer to publishing and want to integrate platform API's for things like In-App Purchases.
Having said that engines like Unity are incredibly powerful and talented 3D programmers have produced a lot of common 3D rendering functionality into a tightly honed and well-documented beast which you can call on to get your game up and running quickly. Of course you don't get this for free: a professional studio has to pay $1500 USD for each of Unity Pro, Unity iPhone and Unity Android - so a total bill of $4500 USD (or around $6000 AUD) per seat. And then you pay that again when Unity goes up a major version number. Its worth every penny - if you need that 3D functionality.
The essential trade-off is that particular engines are well-suited to producing a particular type of game. In fact most engines already have a type of game in mind. Unity is ideally suited to games where a 3D avatar in either 1st person or 3rd person is running around a world, followed by a camera.
You can of course fix that camera, get rid of the avatar, and place a 2D orthographic scene in the view to thereby construct a 2D game. This is kind of like using a Mack truck as a garden shed: you've got all this powerful 3D architecture under the hood, and your fixing most of it in place so that you just use the 2D side of it. But as mobile devices get bigger and more powerful, and include 3D rendering hardware it turns out you can get away with doing this.
As a programmer however, you can do away with all this extra baggage and enjoy the following benefits:
- faster load times (no 3D engine to load)
- zero cost per seat development
- less licensing hassles
- traction with Apple
- immediate access to Platform API's
- App Slicing
- In-App Purchases
- open source community
- 3rd party stuff often $0
What are the drawbacks:
- You need a programmer
- Non-tech team member friction
- Less off-the-shelf "store" assets
- Code is not cross-platform
One of the difficulties with Sprite Kit is that many artists, animators and other creative professionals who are not programmers do not have experience with it yet, and thus find it hard to directly contribute their assets to it, and work with the tool set.
However I think this will change as Apple has a huge commitment to making coding with Swift and Xcode more open and available to the wider creative community.
The Sprite Kit route was a no-brainer for me as I have a career as a software engineer behind me in my game development. Also I had my previous experience with Cocos2D which has a very similar API. I don't need the features that Unity excels in such as 3D. As a founder of a studio I don't want our IP saddled permanently with the licensing and lock-in hassles of being welded to Unity3D - which is a proprietary technology. Unity also likes to monitor your games so that it can tell if you're adhering to the licensing conditions and it requires you to notify your players of this. Basically for me its about having control over my game and my IP.
Swift, and pretty much all the Xcode & Apple toolchain is Open Source and Xcode itself is at the end of the day just an IDE, so there is nothing stopping me editing my code in JetBrains AppCode or other tools. I could edit my assets in other tools such as for example Texture Packer, and TileEd.
Want to give Sprite Kit a serious try for your studio's next game? I strongly suggest going and buying Ray Wenderlich's PDF book "2D iOS & tvOS Games by Tutorials". They update their stuff and you get the updates for free, and despite the title its not just a starter book, it teaches you 90% of what you need to get going, and gives a huge bunch of sample code as well.
Your mileage may vary. But have a look at the pros and cons, and especially try to think of the longer game if you're a professional studio, or consider that you might want to be a buy out option so that you can at some point exit your studio. Due diligence and buy-out exits are much easier when you have the fewest possible licensing hassles linked to your IP.
I hope this article helps, and if you share some of the same aspects that I do when it comes to game development you can be clearer about your choice of engine or framework.
Thanks for reading!
No comments:
Post a Comment