Slashdot Mirror
Carmack: Lord of the Games
seer writes: "This article on Red Herring is a nice look at the interworkings of id software, most specifically their famous employee John Carmack. It delves deeply into the fact that id has stayed a very small company and dabbles with other topics such as Carmack's tendency to stay away from Microsoft 'standards' and the whole DooM ]I[ debacle. An interesting read."
There was a point in id software's evolution where John Carmack almost adopted Java as the "scripting language" for Quake 3. This got abandoned REAL quick when it wasn't fast enough.
Java is all about interoperability, then ease of design, then speed.
Id software's game engines are all about speed, then interoperability, then ease of use. All the same, they STILL tend to be pretty easy to use, since they aren't motivated by business decisions as much as they are by making a really, really cool game engine [while this has positive business ramifications, obviously, that's not why they do it].
Quake 3 is a SUPERB game engine on all platforms. I can write my own game as a Quake 3 mod, and without any recompiling, have it instantly work, at high framerate and with no bugs or glitches, on three different platforms. Show me how Java can do that.
Id software's game engines ALREADY surpass Java. It's not going to get worse from here; it's going to get better.
-Hentai [in vita non pacem est]
You may be missing the bit that says "Mr. Carmack, a multimillionaire."
Once you're able to buy just about anything you want, the ability to buy more stuff isn't that attractive. Beyond that, "more money" would come with an inevitable loss of control -- to take ID Software to a large company would mean hiring legions of programmers and managers.
At that point, it becomes what most large buisnesses become -- merely brand names. When that happens, it's no longer Carmack's baby, and it probably wouldn't be something he enjoys.
Ergo, ID Software stays small.
"Evil company X is threatening to restrict our rights! Let's all get together to stop--OOOH! SHINEY!!!" -- AC
To this reporter, Carmack isn't using OpenGL, an existing 3D graphics standard which Microsoft refused to adopt in favor of their (for years inferior) attempt to lock programmers in to DirectX, he's using "his own graphics technology" which is "almost like a religious thing for him".
By the time Quake 3 Arena came out, I think a lot of people realized that id had basically become a brilliant game engine company that should just cast off the illusions that they were experts in creating innovative gameplay. Because I think, while you can debate the merits of the technology behind id's products until the end of time, it became clear to many that the innovative gameplay was happening somewhere else. While the engine was brillaint, Quake 3 the game was the same old, same old. Deathmatch in a brown castle.
While many people had not-so-kind things to say about the multiplayer aspects of the origional Unreal, when Unreal Tournamnet came out, Epic was pushing the bounderies of online gameplay, while id was left in the dust, cranking out the same thing yet again.
Not to say Quake 3 was a crap game, a hell of a lot of people enjoyed it then and enjoy it to this day. I'm just saying that it was part of a downward trend at id, one that they seem to have addressed, and I commend John Carmack for that.
Its actually true. Money is not a motivator and a recent, heard it on NPR, showed that acutually having more money can in the long run ruin your happiness.
The argument goes something like this
1. In the study, most people agreed that if they only had 20% more money they would be happy (ie, money would buy happiness)
2. But as the study looked at people over a long time, they found that people in the lower and middle classes actaully felt happier than upper higher incomes.
How does this work?
1. if you are lower or middle class, then 20% more money can be more easily achieved, so more people then to achieve that goal in the long run, versus
2. With the upper income, 20% is much harder to achieve, so the long run they dont achieve their happiness as often,
3. Hence Carmach may have felt he has reached a level that he doesnt need more money, ie it doesnt make him more happy
anyway thanks for reading
Sigs are dangerous coy things
You are uninformed why John Carmack did not use Java. FlipCode had the following pertaining to Carmack's view on Java: John Carmack considered using Java in id's Q3. He discusses several pros and cons. His biggest problem with Java was (believe it or not) portability -- It seems that Linux's and Solaris' implementations of the JVM are lacking (with respect to JNI) as compared to Windows'. However, John does say at the end of this interview, "As for embedding... If I did want to go off and start fresh, I would likely try doing almost everything in Java." John Carmack considered using Java in id's games for quite some time, "... I want Quake 3: Arena running on every platform that has hardware- accelerated OpenGL and an Internet connection.... I had been working under the assumption that Java was the right way to go, but ... The programming language [used in Quake3] is interpreted ANSI C. The game will have an interpreter for a virtual RISC-like CPU." (Unreal followed a similar approach.) ...
"When it comes around to the next development cycle, I will make the Java decision again." As for embedding: "We are still working with significant chunks of an existing code base. If I did want to go off and start fresh, I would likely try doing almost everything in Java."
-- Game Developer Magazine, July 1999 issue. As more and more developers start taking advantage of all that Object-Oriented Programming (OOP) has to offer game development, extensible Java game objects may become the norm. More advanced players may even start to demand it - who wants to learn a proprietary script syntax for each new game they want to extend? If Sun delivers what they've promised, Java may (in time) become the de facto standard. Or not... 8-)
Think of the name of the company: *id* software. It's supposed to be simple software that appeals to the gamers id (i.e. cruelty, lust, greed, et al.). The id does not need, nor does it understand deep plots, or innovative gameplay. It understands revenge, the joy of domination, you know, the primitive stuff.
At least they're not making the kind of tame that their title also implies: porn.
BlackGriffen
Of course, it also has the great side effect of making those old games fully hackable and getting more people to upgrade to their latest game.
But that's the cynical side of me, I love the fact they release the source code too.
According to an IBM study, natively compiled Java isn't much faster (and can in some cases be slower) than bytecode-interpreted Java.
The more I learn about the Internet, the more amazed I am that it works at all.
The above post, not the parent, should be "+5, Insightful". Every game which carmack releases source for become unplayable in a week. Which means I am forced to upgrade to the newest game, in a manner much more convincing than any microsoft release-- because I am one of the best quake players, and I get accused of cheating and banned from servers when I don't. Meanwhile the opponents whom I dominated a week before somehow stop missing rail shots, start shooting through walls, and move too fast for even me to hit. I still usually win, because no cheat can give you the strategic insight needed to be really good, but much of the fun is taken away.
I've given Carmack more money than I've given Gates, for this very reason. Q3A isn't more fun than Q1, or at least it wasn't before the speedhack killed it. Neither is RTCW really more fun than Q3A- although I guess I'll buy it soon, right about the time people start flying in Q3A.
Open Source works great for projects which began as open source and had meaningful feedback from the beginning. For games or other applications which have finished their development cycle it is a nightmare; it combines the worst features of security through obscurity with the worst features of open source software.
IMUHO if Carmack was really a supporter of open source, Doom III/RTCW would be open NOW so that when they came out they would be unhackable. I think his rationale for releasing his source is simply that he wants to sell more cd's, while masturbating to the egoboo he gets from being worshipped on slashdot.
If I'm wrong about this, and he is really just a brilliant but naive programmer, I apologize; the reason I keep buying Quake instead of half-life or one of the other clones is because I believe the innovators like carmack should be rewarded. He gave us the fps genre, and I always get my money's worth from a game he makes within a few days of buying it, so in my mind the upgradaing is worth it. However, the open sourcing is not something I think is good for either the quake community or the open source community, and is not something I think he deserves kudos for. The day the source for a game is released while still in beta I will join y'all in celebrating his altruism.
From Carmack's point of view, cheating is a social problem, not a technological one. Well, you could have all gamestate stored on the server, but Quake wouldn't be very fun at 1 fps. The only real solution is trusted clients, and that ultimately means trusted people.
The technology exists to certify the identity of trusted players (x.509, SSL, etc). Nobody uses it. Since you're the bigtime player maybe you could tell us why.
Money not a major motivator? I wonder how honest that statement is. More importantly, where can Id be going with Carmack's other "attitudes?"
No, money isn't a major motivator for Mr. Carmack, and as for his attitudes, I've found him to be one of the most even tempered people I've ever met.
Less Talk, More Beer.
From the article: "That's not what we're doing," Mr. Carmack says. "We're doing entertainment. Saying it's art is a kind of sophistry from people who want to aggrandize our industry."
I totally disagree with this statement. I view coding (particularly coding for games) as something that straddles the gulf between work and art. It may not be 100% pure art, but it's certainly not aggrandizing to say that there is a fair amount of artistry in well conceived and written code. The first 25 years of my life was spent pursuing a variety of artistic endeavors (writing, music, visual arts) and I get nearly the exact same feeling in me when I'm writing code as when I'm composing music or drawing. There is definitely some link between those activities. I feel the same creative impulses firing when I'm programming as when I'm doing any other art form and I feel that same sense of artistic fulfillment or satisfaction when I'm finished with a project. It probably sounds a little fruit-loopy, but it's the truth. There is an element of artistry in writing code. I have no doubt about that.
--Rick
--Rick "If it isn't broken, take it apart and find out why."
So many software companies (and not just MS, mind you) are content to sell code at full price that they've written years ago and just require periodic maintenance now.
It seems to me that one way to prove beyond a doubt that you have skills it to write a program, make your money off of it over the course of a few years, and then give it away because the stuff you've written since then is actually worth more to people.
A few corrections to the article:
"My own graphics technology"
is OpenGL.
"Mr. Carmack also plays computer games in the office with his coworkers"
I played Q3 quite a bit, but not much since then. The team focus of TeamArena and Wolfenstein just isn't my favorite type of game.
"Polygon counts"
The Doom engine is not an ultra-high poly count engine, because it is built around dynamic lighting and shadowing, but it is still a large step up from our previous games. Typical scenes will have around 150,000 polygons, versus 10,000 for Q3. There will certainly be other games with higher raw polygon counts, but that is really focusing on the trees, not the forest (image quality). The large numbers that have occasionally been tossed around are the polygon counts for the high detail characters that are used in the generation of normal maps for the real time rendering. Some characters are over 500,000 polygons in their original form.
"It looks like the type of game that is so thrilling to play that gamers will do so over and over again, even though it lacks a narrative plot."
Unlike everything we have done before, the new Doom actually DOES have a real plot, and I think it is going to be presented well. I don't really expect most people to believe us at this point, but wait and see...
"The new Doom likely will require a no less powerful chip than the soon-to-be-released Nvidia GeForce3"
It is designed for full impact on a GeForce-3, but it still runs on a GeForce-1 or Radeon.
They didn't reproduce the graph of our revenues from the print version, but that was also way off base. I guess they estimated them based on our title sales, but while Doom II remains our best selling title, we have much better royalty arrangements now than we did back then, so we make more money today.
John Carmack
carmack is cool in that he's one of the last bastions of hope for non microsoft controlled game development (i.e. he uses open gl).
long live open gl (hopefully)!