Hacker News Reader

Comment by cubicle67 | Hacker News Reader

Carmack is the original. He's been writing class leading graphics engines since the days of distributing software via discs in plastic bags.

He wrote the first side-scrolling engine for PCs (long before they were called that) back when it was considered impossible.

He made people sit up and take notice of Shareware

He wrote the engines for Commander Keen (2d platformer) Wolfenstein (3d) , Doom (3d with variable height floors/walls), Quake (full 3d), Quake II (coloured lighting etc) etc etc

He's one of a very small number of people who were writing groundbreaking code right at the beginning, and are still writing groundbreaking stuff.

He's constantly made his code public, as much as possible

He loves fast cars; how does a 1000hp twin turbo w/24psi of boost Testarossa sound? (that was 11 years ago)

He founded a company that builds and flys some serious rockets. real ones.

Seriously, if that doesn't make you a coding god, what does?

View on HN

Not only that, while he's not the only talented graphics developer out there, he is one that has consistently brought something revolutionary and new to the table.

id Tech 5 is going to introduce the entire concept of megatexturing - the grand unifying theory of geometry, just as Doom 3 was the first major engine to implement a unified lighting model. This is going to have major implications for both content production and rendering, and is exciting as fuck.

Carmack's impressiveness is not only in his ability to write kickass looking engines, but also his ability to lead the pack in completely rethinking the sort of things most devs take for granted.

Nitpick: As the name implies, megatexturing solves the problem of texture memory management and rendering, allowing essentially unlimited texture detail on any object. The problem of unlimited geometry detail is still unsolved. Carmack has hinted that id Tech 6 may tackle the geometry problem using voxels instead of polygons, and id actually demoed a voxel rendering system at SIGGRAPH 2008: http://www.youtube.com/watch?v=VpEpAFGplnI

Are there any good whitepapers on megatexturing out there? I've been trying to learn about it, but the powerpoints aren't cutting it.

This is the best explanation of the technique I've found (if you need enough detail to implement it): http://silverspaceship.com/src/svt/

He uses the name "Sparse Virtual Textures" instead of megatextures so it's not very googleable. Also it's a video, not a paper, but the code is included.

Didn't Quake: Enemy Territory have megatextures for the maps? Is that iD Tech 5?

It was iD Tech 4 plus MegaTexture. The feature for iD Tech 5 is being called Virtual Texture, and it supports more than just the terrain.

Sheesh, didn't expect to be so severely down voted. I guess I chose my words poorly. I'm fully aware of Carmack's key accomplishments. He's an incredibly talented and influential guy. I didn't mean to imply that their are many people with a career much like his.

What I meant to say was that there are many great engineers who are quietly developing fun and technically impressive games.

Also, I had some deja vu and Googled up a comment I wrote a while back. It is about why I feel that Carmack has missed the mark on several major game technology trends: http://news.ycombinator.com/item?id=419606

I understand what you're trying to say. I'm even close friends with someone whose skill level I'd call Carmack-like (an assessment I made, as a professional graphics programmer, over the course of two years). Your point is a good one, but not one that anyone wants to hear. I'd say just let Carmack have his moment; doing that on an iPhone is nothing short of breathtaking.

Sure, there's others with a very high degree of skill like Carmack, but it's his consistent ability to come up with amazing stuff over the last two and a bit decades that gives him legend status.

It's not just ability and skill, it's also impact on the industry (which, obviously, requires a degree of luck and circumstance). It sounds pretty hash, but it's kinds hard to be considered a legend/god/whatever if no one's heard of you.

But you raise an interesting point. For those people how do make an impact, to what degree is it due to circumstance, and to what degree their own skill? There's loads of very highly skilled people in any given field, and yet only a very small number of them will ever make much of an impact on that field. Why is this, and what differentiates the ones who do from the ones who don't?

> Your point is a good one, but not one that anyone wants to hear.

That's why I left Reddit—the tyranny of the majority. I guess it's here now, too. Where to next?

Let me let you in on a little secret: karma doesn't matter. Most people won't look at the number next to your post. In fact, some might read your post solely because it has a low rating. It's hardly tyranny of the majority when your opinion can be seen just like anyone else's.

I know that, and you know that, but... achievement-oriented game mechanics pull in the kind of people who enjoy them, and a gathering of Achievers encourages a culture of competitive sport, rather than socialization/rational discourse.

Here's a parable: imagine that you and your friends are all playing a pen-and-paper RPG with a focus on storytelling, getting along, quite happy to collectively have interesting things happen to all the characters. Then, imagine that a new guy joins, and he's a munchkin—he plays the game to win, getting all the other players angry. He doesn't like his character to "get screwed" to tell a good story, and so on. The other players realize this, but they put up with it, hoping, for a while, that he'll eventually adapt. But he doesn't. So, eventually, the other players begin to leave—and new players join that play in much the same way the munchkin does. Eventually, the group is no longer story-focused, but is rather completely focused on collecting imaginary points and tokens.

See what I mean? Although karma doesn't matter objectively, if it's important within the "shared delusion" of the majority, it can still do damage. That something was voted high or low wasn't the basis for my statement—it was that it seems, lately, that the majority shares the delusion; that this place is a munchkin group.

"that the majority shares the delusion; that this place is a munchkin group"

witch brings us to the point: You are right and the majority is wrong. But without good explanation, you still deeply care about what other people think about you("it still can do damage").

When I played pen-and-paper RPG, the storyteller used to favor her friends-interest over anyone else(the majority), so maybe that is the real objective: Impose your way of thinking over the majority, If not, you don't play.

> But without good explanation, you still deeply care about what other people think about you("it still can do damage").

That's not quite what I meant by the phrase. It "does damage" because others in the majority accept that it does—for the same reasons that I am "in debt" when a number on a machine reads negative. I might agree or disagree with the conception of money as a fiscal instrument, but it is supported by the group in a way that prevents me from participating in the group unless I acknowledge it and treat it the same way that they do. To not acknowledge money (or karma) is, in effect, an act of self-alienation. You "reject yourself" from the group by doing it, because the game they are now playing is a competitive one—and if you do not acknowledge the table stakes, you cannot ante.

This is why it is important to do some research on social game design (or consult a game designer) before you code a game mechanic into your social webapp. Years of research into MUDs and MMOs show that different mechanics reward, and thus encourage, different types of player-styles, and thus shift the user-base in different directions.

Here's the most famous paper on the subject, just to get anyone interested started: http://www.mud.co.uk/richard/hcds.htm . It would be awesome if someone would write an "Interpreting Bartle for Web 2.0" article and submit it back here; I've been meaning to do it for months now but it's still a ways down on my to-do list.

You can always sit quietly by yourself in the corner.

Carmack is a lot like Steve Wozniak. There's no question that he's an incredible hacker, but (as I've heard second-hand) he isn't crazy about managing large teams. Carmack was his most influential in the late '90s, when a small group of developers could still make AAA titles. That isn't really true anymore, now that high-end games routinely involve a hundred or more developers.

Carmack's team is still large enough to develop a top-notch engine that can be used in AAA titles put together by some other shop. The idTech engines have not been slacking in innovation, and the biggest thing limiting their use outside of id recently (lack of scalability down to low-end hardware) is clearly not a problem anymore.

They're clearly making more of an effort with Rage than they did with previous engines. They were spoiled to a degree by how eagerly 3rd-parties licensed the Doom, Quake, and Quake II engines. Id developed a nasty reputation for dumping code on licensees with little support. In fact, one of the main ways Epic stole the licensed engine market from them was by providing much more aggressive marketing and support. Id's attitude was always "if we build it, they will come".

Even high end games only involve a few dozen developers. You would get nothing done with a hundred developers trying to do anything on a single project.

The credit roll at the end of high-end games is filled with level designers, artists, qa, management, designers, etc. The actually developer section is relatively small.

That's why I used the word 'developers'. Even twenty to thirty programmers is still a lot to manage, particularly if you have ambitions to do any of the programming yourself.

He also did the first port of X11 to Mac OS X, see Xquartz(1).

Reply on news.ycombinator.com