Software/Device Development and Design

Unity and Unreal Compete: Indies Win!

Big news from game engine developers at GDC this past week:

First, Epic dropped the up-front cost of Unreal Engine 4 to "free for everyone", with only a 5% royalty beyond each project's first $3000 revenue (per quarter).

Then, the following day, Unity released their long-teased Unity3D 5, along with an indie bonus of their own: Nearly all previously Pro-only features are now available for free to non-Pro users (excluding the snazzy Pro-only editor skin, darn!).

Also exciting is Valve's unveiling of Source 2, also free for developers. Although, I admit I know very little about Source, so I can't really comment any further on it.

I've had an eye on both Unity3D and Unreal Engine for awhile now, I even dabble in Unity3D a bit, and I find it interesting the directions both have taken. The two engines come from opposite backgrounds: Unity3D from the indie scene, and UE from the AAA industry (Although initially, Epic itself originated from the old DOS shareware scene - arguably the original indie game scene). But lately, they've been encroaching on each other's territory. Unity has been touting one AAA-targeted feature after another. Meanwhile, Epic has reversed the "expensive, exclusive and difficult" image from their UE3 days, and drastically reduced both technical and financial barriers of entry - a move aimed, in large part, straight at indies.

The wall between indie and AAA development tools has collapsed, attacked from both sides, with Unity and Epic being two of the biggest demolitionists (among others as well). And indies are quite possibly the biggest beneficiaries.

Here's what the two engines now offer to indies:

  • Anyone can get started. (Both Unity and Unreal)
  • Easy to use game editing environment, with rapid prototyping and development turnarounds, and loads of useful tools. (Both Unity and Unreal)
  • Engine with AAA-quality and speed. (Both Unity and Unreal)
  • Target all the most popular PC, console and mobile platforms... (Both Unity and Unreal)
  • ...and then some (Unity)
  • Develop and release without investing one cent on tools or engine. (Both Unity and Unreal)
  • Official online store for buying/selling assets. (Both Unity and Unreal)
  • Ease and safety of [an older version of] C# and other CLR languages (Unity)
  • Full power of native C-linkage languages (implying a potential for possibly using D!) (Unreal)
  • Full, unrestricted access to the entire build system and engine source. (Unreal)
  • Partial ability to develop on Linux (Unity via Wine, albeit unsupported and occasionally problematic[1]. Unreal via an in-progress community-driven Epic-sanctioned effort[2].)

Even the few differences above aren't so different after all: Unity is perfectly capable of interop with native C-linkage languages (it just requires taking extra trips manually across the managed/unmanaged barrier), and there's no reason a managed scripting system can't be used on top of Unreal Engine (you're just "on your own" with that, AFAIK).

Of course, Unity and Unreal aren't even the only options in town. There's still Source 2 as mentioned earlier, plus CryEngine, MonoGame (a resurrection of XNA), and even others.

It's a good time to be an indie: Engines compete, we win.


[1] I'll have to save my adventures using Unity3D's editor on Linux for a separate post. In brief: It gets the job done, although not spectacularly well. Certainly not well enough to obviate the need for a native Linux editor.

[2] At the moment, I have no idea what the exact state of Unreal Engine's editor is on Linux. I know it exists, my understanding is that it's usable, but I know nothing about how usable. Could be flawless, could be comparable with Unity under Wine, could be less, I don't know. But I am definitely interested.

Read more


Unity3D vs. Linux Developers: Profit is Fine, But Ignorance is Better?

A little background:

Unity3D has quickly become one of the most significant game engines and development tools in the world, fueled in no small part by being insanely cross-platform. But somewhat contradictory to their belief in cross-platform, their actual developer toolset only works on Windows and Mac, with what's become a rather noteworthy disinterest in supporting Linux-based developers. This, despite the fact that the #1 voted feature request on their own feature request board, by a wide margin, has long been a Linux-based editor. Naturally, this has been quite controversial with many predictable arguments on both sides.

As a side note, I can personally attest that Wine is not a suitable long-term solution here. I do use Unity under Wine, and I can get work done with it, but there are definite problems (which I'll have to save for a separate post).

With that background out of the way, the rest of this post is mainly in response to this:

"Now, let's be honest, of those 9,175 votes, how many of them would actually be developers who would buy [Unity3D Pro] and develop on Linux? Answer: probably far less than the actual number of votes on it."
    - liamdawe @ Gaming on Linux: Unity Confirms They Have No Plans For A Linux Editor

Let's do a little basic gradeschool arithmetic:

In the time since that original post, the votes have increased to 16535.

Up to 10 votes are allowed per user, so let's be maximally conservative and pretend that everyone voting on it used all 10 votes (I didn't, but let's pretend I did). So that's 1653 people.

As for the number of people who would actually buy the Pro version: First, let's make the questionable assumption that the people who have already voted are the only ones who would actually buy it. Second, would you say 10% qualifies as "far less" than the number of people? Hmm, yea, I'd say 10% qualifies as "far less" than 100%. So that makes 165 buyers of Pro.

Let's completely ignore the fact that using Pro for two of the most popular platforms (Android and iOS) carries an additional per platform cost equal to that of the basic "Pro" itself.

There are two pricing options: Flat $1500 until the next major release (at which point an upgrade free is required), or monthly fees that total $1800 every two years ($75 * 24) which is roughly in the general ballpark of the lifespan of each major release. So let's go with the lesser: $1500, and completely ignore upgrade frees.

Unity also profits off their Asset store, even from non-Pro users, and soon they will also offer other value-extras to non-Pro users (hosted build servers and analytic tools). Let's completely ignore all of that.

That makes an estimated minimum baseline of 165 * $1500 == $247,500. Nearly a quarter million dollars US just for a conservative baseline estimate on Linux developers alone. (Don't forget, software developers are much more likely to be Linux users than average computer users are, it is arguably a programmer's OS after all.)

Oh, but wait, there's costs:

Unity Editor is built primarily on what? Unity Engine and Mono. Both of those already run on Linux, with full support. And most of Unity's C++ code (yes, it has C++ code) is in the engine (again, already fully supported on Linux), not the editor. And Unity Editor already runs on one Posix system: Mac.

Additionally, what most people who have never done cross platform development (I have) don't realize is that unless you're doing things really, really, REALLY badly, the vast majority of even a natively compiled (let alone CLR/Mono) program is completely OS-independent.

There's also the well-known developer's rule of "0, 1, infinity" (if you don't know it, look it up). Applied to cross-platform development, it means that adding your first extra platform (already done - Mac) costs far more than adding your second, third or fourth extra platform. (Ok, yes, I'm aware that "0, 1, infinity" technically refers more to not imposing arbitrary hard numeric limits, but the practical reasons behind that are rooted in exactly the same principle I'm applying here.)

All that combines to mean that the real costs (as opposed to imaginary scenarios by peanut gallery members who know exactly nothing about real-world cross-platform development) of creating and supporting a Linux editor is minimal.

There is money on the table, and Unity is throwing it away. They are giving that money, userbase and mindshare to their #1 competitor: Epic, a company who has historically been somewhat MS-centric and yet has already become vastly more friendly to Linux developers than Unity.

Read more