f13.net

Silicon Knights sues Epic. Drops UE3 for Too Human.

Gamers ask: "What the hell is Too Human?"

Silicon Knights (quick refresher: The people who made Eternal Darkness: Sanity's Requiem for the Gamecube, as well as Metal Gear Solid: Twin Snakes) is suing Epic games, claiming that their Unreal 3 engine isn't up to snuff for game development and that it isn't delivering what Epic promised.

Oh yeah, they're obviously dropping the U3 engine for Too Human, a planned trilogy with a $10 million per-game budget that has been in development since 1996.

Vaporware much?

Props to Gamasutra, who has a copy of the lawsuit and some (pretty typical lawyerish) reactions from SK.

Comment by Schild to follow (hopefully, lest I look like a jackass).

When Nintendo lost SK it looked bad, in retrospect it don't look so bad any more. Too Human has been in development for so long, *now* they are dropping the underlying engine? Wow.

That said, the lawsuit does make some pretty specific allegations and if they can get other developers to corroborate it might go somewhere.

Not indicative of anything, but...

From Neogaf

If I look those up, are those all on UE3?

Yes.

Damn, the Neogaf thread, while not worth wading through 17 pages is full of hilarity.

(http://i7.tinypic.com/4p2gu80.jpg)

MS is certainly in an interesting situation as a result of this.

On one hand, GoW and it's mad mad moneys.

On the other, a (3 game contract = 'for all practical purposes') internal dev that's mad pissed and possibly rightfully so.

Good times.

Not coming in with an "official" position on this, but a couple of things to discuss/note:

--A game engine provides functionality, not skill.

--It's pretty common for a game engine company to use their own technology to make a game, and have evolutions of that game code that develop through the process of using said code. That's a good thing, not a bad one. Commonly, they may refactor the lessons learned into future products, but it's not a requirement.

--No comment on the contractual clauses discussing releasing updates specific to a hardware platform within a set time of that hardware platform being released, although I personally don't think that was a smart clause to allow.

--If you look very carefully, and read between the lines, SK is implying that they are re-engineering the engine and want to use it royalty free. They are claiming that they should both not be responsible for paying license/royalty fees, and can continue using the engine with the goal of "making their own engine".

This could potentially open up the age old copyright argument and common licensing agreements that exist in the industry today, and if they win this portion (gaining the right to basically re-write their own engine using Unreal as a template), it will not bode well for game development in the long term. In my eyes, this is almost the exact opposite of the "patent anything and protect it forever" misuse that we are seeing from patent holding companies regarding things like orbit cams, etc.

Disclaimer: it's in my sig, but this is Stephen Zepp the internet dork's opinion, not Stephen Zepp the GarageGames employee's position, and should be taken as such.

Thanks for reminding me, as I was getting some funny ideas as to how GG operates.  :-P

Heh -- needs one more line added to the list:

All other games currently in production except for Madden - delayed

Heh..not sure how to take that  :evil:

GG is a political action committee. :-P

Whoa, hey. These are all just announced to be delayed?

Heh..still not sure how to take that, but I firmly believe there is a deeper implication to this suit than contractual disagreement. An analogy that may suggest what they are trying to do: "The car sales guy said this vehicle would go 0-60 in 4.9 seconds, but I took it to Denver and it didn't do that, therefore I can add some stuff to the car, and then turn around and have rights to mass produce it in other colors and sell the result in direct competition to the original producers".

If that passes in the suit and SK is granted their claim for recompensation, it will put middle ware game engines out of business, which will set the industry back decades, and may be unrecoverable. Call that chicken little screaming if you like, but it's drawn my attention in a big way.

I don't know if it would kill middleware off. I'm not legally trained, so I don't know exactly how well they hold up in court, but I expect that, if they won the right to build their own engine based off of the knowledge extracted from examining and working with Epic's tech, you'd instead see something else suddenly popping into all middleware contracts: A noncompete clause.

Essentially, if you license our middleware, you agree to not create or market your own for-profit game engine for the next X years.

Also not great for the industry, especially for dev houses that have homebrew-engine games coexisting with licensed-engine games (c.f. Bioware, self-developed stuff for NWN, licensed stuff in Bioware Austin's Project, and I believe some licensed stuff in Bioware Pandemic's projects). Some of it might be circumventable by balkanizing your business units into separate corps wholly or primarily owned by a holding corp (again, not sure of the law here). Or it might just balkanize the industry into pockets of dev houses, each a one-engine-tech shop.

Let's look at it piece by piece, in accordance with the GamaSutra claims:


Inflammatory at best, impossible to prove as well. while simultaneously sabotaging efforts--another way to read that is "we couldn't figure out how to use the API, but they did--therefore they were intentionally sabatoging us".


How do you define "anything resembling working XBox 360 code"? Again,  this points directly to "we couldn't figure out how to make the API do what we wanted, so it's their fault".


Epic game developers are not going to be people answering support questions. Period. "racing to complete and market its own games" is a ridiculous reason to try to provide for why SK was/was not getting the support they felt they were owed due to the contract.


Absolutely off the wall. I ran the support division at GG for two years, and support does not mean writing or troubleshooting game level code for anyone. That's like suing a hammer production company because the hammer can't cut holes in drywall. If you don't have the skill to build a house, you can't blame the architect for not explaining how to lay a foundation.


They are complaining because Epic provided post-mortem demonstration code of a successful game to demonstrate to future customers reference implementations, but didn't provide this value-add in time for SK to make the game that SK couldn't do on their own.

What that means is that game engine companies should never provide example code, because someone can say "hey, if you would have provided that code last year, we would have made millions by copying it, therefore it's your fault we didn't make millions".



or, if you read between the lines,

"Hey, we learned nothing at all from an engine that has been used to make more games than any other in the business, so we're going to go ahead and, starting completely from scratch with no input whatsoever from this example that has had zero success, ever, accomplish something that took you almost a decade and more dollars than we could ever make in our company's lifetime, in time to make our game. Oh, and it's ours.


"Ahh, well, ok so maybe we aren't completely independent of your code, but it's the thought that counts--and, uhh, by the time we are independent, we'll be, err, independent! And yeah, even if we weren't independent of your code (*cough*), we're, errr, yah, enhancing, so it's ours anyway!".


Anyway, there are several other points in the GamaSutra article that could be responded to, but it all boils down to "we tried to do too much, failed, and it's your fault--and we can prove it because you were able to do better!!!".

You mean something like: (GG EULA snippet)


I can guarantee that Unreal has a similar clause.



Effectively taking us back to requiring each game developer to build their own engine, from scratch, thereby losing the benefits of middle ware in the first place.

Hard to say without further action or access to the original contract. It sounds like SK is trying to cover their asses against a counter-suit with the later claims or some sort of development cease-and-desist. Basically they want a loaner, because without one they can't meet milestones and would effectively be out of business. That's my take anyway. If they have to return the defective product what do they show the schedule nazis at the next meeting?

Without knowing what the support contract was and other details who knows? It could be SK sucks, or it could be the engine really did suck ass. I suspect it will get bogged down in what "workable" etc means. Which is why corroboration could be the linchpin. If a bunch of other devs line up and say that they too found it unworkable and the support not up to contract then it could really be something.

Valid points I guess, but it comes down to proper use (and of course proper marketing/sales, from the engine company side) of a product. I'm obviously not a lawyer, and while I certainly don't want to be fighting for Unreal (direct competitor in theory, if not market segment), to me the only thing I can see them winning on is proving failure of warranty of merchantability, which suggests that they have to prove that Epic told them Unreal had a big red "Make My Game" button, and it didn't.

Hot topic for me so heading to cool off, but the entire lawsuit to me is just fundamentally wrong on every level.

Almost, but not quite. When I said "create or market", I should've included the clause "even for entirely internal use". I.e. Thou shalt not write thine own engine in-house for a game you intend to sell (or, obviously, for sale itself). Your license only provides for "derivative works" - although that sounds like an apt description of what SK is creating.

Further, I'm not sure whether UE3 is supplied as source or prebuilt modules. Their comment about providing "working 360 code", among other things, leads me to believe that it's provided as source. The reverse-engineering clause doesn't really make sense when you can just peek at the source instead of actually having to do work to reverse-engineer the algos.


I don't think so, it would just make the formal business arrangements more complex and multi-game houses more expensive to set up. You'd effectively form a corporation for each engine license (or group of licenses, as Bioware Austin's case would necessitate). These would then be owned and managed by a holding corporation that holds the rights to the company brand and "licenses" the brand to the subordinate corporations. Operating income from the subordinates is then repatriated to the holding corp in the form of dividends.

In the alternate case, going back to a one-engine-tech shop also doesn't entirely eradicate the benefits of middleware, as you would probably still be able to license various pieces to chuck together, but it does seriously constrain the industry in terms of flexibility.


Yeah, I agree, the lawsuit, as described by Gamasutra, seems to more or less be bullshit cooked up to throw off contractual obligations to not rip off Epic's code.

One final thought:

I do realize/recognize that what appears to be their fundamental argument is one that was used against Microsoft back the anti-monpoly suits--"(they) had insider info and advantages they didn't provide others", and while I see it as potentially arguable, I don't in any way see it as valid. the argument worked (I think?) against MS because it was a documentation issue--MS didn't document everything that was in the OS, therefore they could be "sneaky" and use the more optimized stuff to produce better apps.

Where that breaks down for me is that Unreal is not a monopoly, and it's not a required asset for successful game development.

I need to stop posting!

However,  while I don't have legal arguments to support, this still screams at me copyright/licensing violation somewhere.


Unknown either (I don't know enough about Unreal's licensing options, but I'm reasonably certain it doesn't provide raw source code for most licensees). This exact discussion came up at work today though, and the example discussed was something to the equivalent of:


vs


How do you measure if the second code block is a copyright violation of the first? It's arguable that 'the code is 60% different!!! (number made up)", or it's arguable that "the algorithm is exactly the same". The example can be taken to whatever extreme you like.

The whole thing sounds fishy.  The complaints against Epic would get them in a lot of hot water with several companies.   The delay particularly should of heavily damaged several companies to the point that they would seek whatever damages they could get.  They could all be true but why is SK the one to actually go for it?  We don't know enough but the surface of things sure makes it seem like a desperation play by SK.

Anyone have public knowledge of what it costs to license the engine? Are there different types of that engine you can license?

I remember reading on a website that to email Epic Games, they will work with you in your attempt to license the engine.

An assessment on another site is > $750,000. Which is to say, after 10 years of SK's work on Too Human and the amount of money they've thrown at the project, it's a drop in the bucket and they need better programmers! BUT! The number of delays on UE3 based games is... well... probably more than games based on established engines. Final BUT!

UE3 is bleeding edge. You get what you pay for. In this case, you paid for buggy unfinished, unestablished engine-ware. Whoops.

Yes, I know I just made up the word Engine-Ware. Deal with it.  :raspberry:

Uhh...Mass Effect is not delayed unless it happened today. It's coming out in November and that's the first release date they've announced. Alot of people were speculating it'd be this summer but that was just speculation and Bioware had nothnig to do with it.

When you say Bioshock is delayed do you mean it was delayed already or do you mean it isn't coming out next month as planned?

It sounds like Epic:
1) Was very late delivering the engine.
2) Delivered a badly done hack job.
3) Delivered little to no documentation since they were late and in a hurry.
3) Was too busy working on their own games to provide support.
4) Had to rewrite huge swaths of the engine to actually get Gears of War to work.
5) Didn't provide their licensees with the needed updates until much later.

If any one of those points was in an a contract or service level agreement the lawsuit is warrented and they will probably win.  I've even heard talk that Epic used the money they got from U3 licenses to work on Gears of War instead of finishing and supporting the engine.

I agree with Stephen Zepp that it's intellectually dishonest for SK to claim that slowly "replacing" each bit of the U3 engine is anything like writing an engine from scratch. 

(I say this as a programmer who has helped build a game from scratch (CoH) and is currently helping build a game from an Unreal 3 license (Global Agenda).)

Well, poor programmers perhaps, but you should still get what you pay for. A few months delay is a lot when a better part of you dev team depends on the engine being delivered on time in order to be productive, we're certainly talking a lot of money lost here, perhaps more than the actual cost of the licenses.

Some of the points in the contract dispute may actually be valid--but when you take the suit as a whole, and look at what they want as compensation, it demonstrates just how ridiculous the entire thing is.

They not only want to be able to re-engineer the entire game engine and claim it as their own, but they also want Epic to lose all the profits from GoW.

Let's look at the UE3 website! Linkies! (http://www.unrealtechnology.com/html/licensing/faq.shtml)


Emphasis in italics mine.

I agree that your example is ambiguous, at least semantically, but that is in part because it's ridiculously simplified. Thing is, there's tons of ways to implement a given algorithm in code, and multiple ways to arrive at that algorithm in the first place. Indepedent development or reverse-engineering I support; however, as with SK's case, "rolling your own" by slowly "replacing" parts of a licensed product is rather unambiguous theft rather than reverse-engineering.

Intellectual property arguments, eugh.

The patent office tends to disagree with you from what we're seeing.

For example, there is currently a patent on "orbit cam", which, if successfully defended, means that every single game on the market that has any form of camera that orbits a point would be  required to pay a licensing fee. From what little information I am aware of, it's algorithm independent--simply having an orbit of a camera about a point, regardless of your implementation technique, is violation of the patent.

Agreed, IP arguments, ugh. We also have EULA enforcement issues there too--which are also a huge ball of wax.

I'm waiting for that case to wrap up so I can decide whether to pursue litigation of my hitponits (life, health, corporeal physical stability measure reserve) patent.

I saw the same sort of thing regarding 'ghost cars' in racing sims.  Not at all what I thought a patent was for.  Very odd. 

Software patents, especially algorithm patents, are bullshit. Period, end of story. Patents that don't even cover a specific implementation, but rather a concept in general, are even worse. The USPTO needs to get burned down and rebuilt from the ground up. Same thing with the copyright and trademark office.

... I'll stop before I have to send my own posts to Politics. :|

I agree that is it's bizarre to claim that an engine is your own work once you've tweaked/replaced the algorithms since that's really just the easy part of creating an engine, save for a few critical parts. The hard part of rolling your own engine is mostly the design, there's no such thing as a flawless design, or at least I've never heard of anyone accomplishing it. From my experience the quality and thought put into the design is what will save you a shitton of work and determine whatever you're building on top of the engine will resemble a hack keeping the shit together or pieces of a puzzle perfectly fitting together in order to create a piece of art. I can only guess that most likely the design of the unreal engine has had a lot more thought behind than most people would spend creating an engine while also being the result of experienced people with years of creating engines behind them. However, at the end of the day, if it's poorly documented, the implementation is of low quality ( buggy ) due to poor Q&A/rushed, and shipped way to late, it's still shit.

BTW, I'll probably regret posting this when I wake up because even if I feel I've put a great post unto this thread increasing my epeen I'm drunk as a fuck and having problems hitting the Post button. I'm happy that there's a spell check to save me from posting jibberish as well.

Am I the only one that WANTS the industry to get set back a decade or more? Games were WAY better 10 years ago. Now all we have is Microsoft controlled DirectX clones of one another. I haven't seen an origional idea for a game since the late 90s. I hate going to best buy and seeing rows and rows of games that all run the same engine. They are all the same flippen game for Gods sake. Someone go and invent something new! And maybe do something in Glide so the game can be ported somewhere other than windows. I havent seen a linux port of a game in nearly 10 years either.

You haven't been looking then.

I don't think that's what he meant.  It's not like Epic is handing out licenses on a street corner, and game companies are saying "well, now that we've got it, we may as well do a game that's basically a copy of Unreal".  I believe it's generally after they have the basics nailed down (e.g. "This is going to be a futuristic story based FPS") that they start looking at their options for engines ("How can we do this with a reasonable budget").  Since using middleware is usually seen as a way to lower development costs (I think, feel free to correct me), it seems like it would be a detrimental blow to original games as much as it would be to knock-offs, because it would be one less option open to a company risking a sizeable chunk of money on an untested idea.  When someone's got some crazy original idea, they can either spend their budget trying to make it fun or they can spend their budget trying to write basic components of an engine. The Unreal Engine doesn't HAVE to be used for Unreal knock-offs; the UE2 was responsible for a ton of games, including Magic: The Gathering (Battlegrounds), Vanguard, and Ragnarok Online 2.  The original UE was used for a pinball game.

This is only somewhat green, but you haven't seen many board games published in Swahili, either--both have good reasons.

Kail's pretty close to what my point was about--and after thinking this through for several days, I think if nothing else happens, this suit will change the way middleware works in the equation (if it continues to work at all).

What's become clear to me more than anything else is that the concept of a "game engine" is something that 90% of the customer base view as completely different from what the engine developers think--studios, indies, whatever the customer is doesn't want a game engine, they want the things that turn a game engine into a game already done for them.

Back to Charlie's point, and why I think it has some merit, it's my opinion that you see so many shooters from Unreal engine licensees is because that's the easiest path--the tools and systems that are "more than game engine, not quite full game" are already done, and therefore the least path of resistance to publishing.

ok, maybe 10 years is an exageration... but you know what I mean. They really are few and far between.


I'll have to call you on this one.
http://www.asymptopia.org/staticpages/index.php?page=swa-eng-01&mode=print
and, ironically, it was done on Linux
I'm willing to bet nearly every boardgame there is has been converted to Swahili.
Why? Because boardgames don't require DirectX.

Heh...ok, it was a poor attempt at an analogy, from a pure business perspective it's simply not feasible yet to maintain linux ports for most games. I hate it myself (old school linux guy), but when it costs more to port/prep for any type of linux than you can ever make in revenue, it just doesn't work yet as a business model.

I would agree with that. The Cedega emulation continues to get better, though, so that's an option for many games.

No they weren't. Either you hadn't played as many games 10 years ago or what you like has fallen out of fashion (e.g. turn-based combat) for mainstream publishers.


I miss seeing AAA title games in AmigaOS as well, and can only hope the market realises the untapped potential of AmigaOS players.

And haven't you heard of consoles or indie game development, where a lot more slightly different stuff appears?

Having read the linked article and Mr Zepp's examination of it, all I can say is that SK has (imo) a strong circumstantial and conflict-of-interest argument against Epic. Put something like that in front of a judge who might not understand the technical complexities of middleware and SK has a strong chance of winning (not an expert legal opinion).

Why? A judge may (and should) hold them to the contract that gets signed between Epic and that third party and if they find that Epic was holding up sending out new versions of the engine / engine support until after GoW had development time spent on it, which allegedly they did, it might breach a contractual clause. It's all well and good for this discussion to put the blame on SK, but if Epic has been sloppy in dealing with their engine licensees or more focussed on their game development, then they've left themselves open for this kind of lawsuit.

Yes, SK might be protecting themselves by going after Epic, but that doesn't make them automatically wrong.

Can a company serve two masters, in both releasing games AND selling licenses to for middleware use AND keep the necessary quality and protections up on both? Yeah, they probably could, but they'd have to be very, very careful about it.

Will someone call me first so I can get out of the building?

I have to agree with Yoru

Shacknews does some journalism (shocking, I know):

http://www.shacknews.com/featuredarticle.x?id=468

I contacted a number of people with the UE3 engine. They're under NDA about many of the things they can say about it. Basically, anyone who has anything interesting or damning to say isn't going to say a damn thing until they see how the lawsuit proceeds.

This made me smirk.

lol why?

I just don't think of Linux as old despite how long it (not GNU) has been floating around.  In my mind I somehow just see abberant GUIs, unmaintaned programs, missing device drivers, etc.  Logically, I know that there are enterprise-grade implementations of Linux... but those make me smirk as well.

Well that's because Linux aint old. At all. *nix is. BSD is. All the other lost OS candidates are (BeOS, Warp, Planet 9) but of the modern surviving OS's Linux doesn't predate OS X (technically though as NextStep, released in 1989, it clearly predates Apple OS X), and would have a hard time claiming 'first post!' against Windows NT. (NT 3 was released early 1992, Linux would be hard pressed to claim release in 1991.)

*nix is old, but our favorite penguin is not really any older than the XP/Vista, OS X crowd it walks with. Which is good, it's aging well with them, although not commercially as viable for gaming, but then neither is OS X.

You made it look like I didn't do any research... which I didn't, heh.  I kinda think of XP as old but I guess the deal is where you draw the line.  My thinking might make more sense if I used the word "mature".  I'd be willing to pass over "old-school XP guy" quicker than "old-school Linux"... nevermind that Linux is just the kernel and you should say Redhat or SuSE or something instead.

I don't like Linux.

Let me expand on this, based on developing software for way to many years now:
Developer, in charge of foozle integration, has a 'Not Invented Here' moment, and decides they can do this bit better.
Developer (or the team lead by Lead) ends up wandering around in the woods far longer than they thought, fatally delaying the project.
When pressed for a reason, Developer (or Lead) repeats the initial NIH reasoning to Business Person, who only really hears 'The Licensed Engine was Crap'.
Hysteria and Blame ensues, as people try to save their jobs by pointing fingers at Licensed Engine in a desperate attempt to Not Loose Their Jobs.

Everyone Dance.

*does a necro dance just for you!*

http://www.youtube.com/watch?v=rT90keJ51bY