Except I find that Mozilla IS as featureful AND MORE stable than its proprietary competition, and I'm using it right now. I don't disagree with your final point,
"...as long as OSS is not as featureful OR as stable as its propietary competition, no one, but NO ONE will use it."
Although I'm not sure it applies to Mozilla anymore.
The cause of vaporware is in the marketing department, not the engineering department.
Far be it from me to defend any marketing department anywhere, but to be fair, marketing isn't the only source of project bloat. Anyone who's A) overambitious and B) in a position of power can create the sort of disasterous conditions you've described. It could be the marketing department, or any set of suits making decisions from upon high, or even a single megalomaniacal project director.
Clever design + bad dev team = Deer Hunter, so there is an argument to be made for both sides.
To be fair, I recall reading somewhere that the Deer Hunter team only had about three months to make the entire game. Rather than a "bad dev team", that may simply have been rough circumstances.
There is no substitute for a good game design doc.
I agree with your points here. I talked with a few of my co-workers about this thread, and we arrived at the tentative conclusion that J.C. is talking about sort of pie-in-the-sky speculative design work rather than concrete, plan-game-elements-out design work.
Likewise, I think he's probably coming off harsher than he means with regard to design documents. I could be wrong, but I think he's mainly concerned with the inflexibility that may accompany a design doc -- a design must be a living, adaptive thing that grows to meet the needs of the project. Unfortunately, it's easy to get attached to work you've done and cling to ideas that may no longer be appropriate.
I sincerely hope that Carmack doesn't just feel this way because he doesn't want some damn designer telling him what his priorities should be. That would be pretty disappointing, because that interdependency is part of the fundamental teamwork necessary to ship a game. Design sets priorities for the Art and Engineering groups, and likewise, they set priorities for each other and for the Design group. That's just the way it is.
Damn. I had to open my big mouth. Now I'll never get a job at id.:)
I'd like to know exactly how J.C. defines "abstract creative design" and "strategic creativity". I'm a game designer, and most projects I've seen suffered from a *lack* of a clear, comprehensive design up front. I don't know if that's what he means or not.
Well... where I work, that's called an "A bug". We can't allow situations where the player can render the game impossible to finish.
Imagine this: You collapse said ceiling, then save your game. Now... you didn't know that this would block your only path out, so it's not really your *fault*. But now that you've saved the game, and you've made it impossible to finish, you'll have to start over (or, more likely, throw the game in the trash).
Along with bad crash bugs, this is the worst sort of game bug.
Actually, a PC/PS2 game called Red Faction featured something it called "Geo-Mod" technology, which was basically real-time boolean subtraction to simulate holes in walls, etc.
It wasn't as interesting as it could have been, from a game-play perspective. The problem is that we game developers are used to making environments *very* static -- often based on BSP-trees and such. This saves the engineers a lot of headaches, but guarantees rock-solid, unchanging environments.
What you're talking about, with full structural physics and everything... it's going to be a while. I'd love to see an area where, if you destroy enough pillars/supports, the floor above you comes crashing down. Unfortunately, that's a sh|tload of math, and the environment would have to be created in such a way that it doesn't close off any critical paths when collapsed, etc etc etc.
It's a tough problem. But consider this: We're just now starting to get shadows that move when you move lights around! (That's what Carmack's working on right now...)
As Stiletto already noted, Direct3D and OpenGL don't really differ in performance anymore (if they ever did in the first place).
Also, Quake 2 wasn't really Glide *per se*. It was OpenGL that interfaced a custom 3DFX 'MiniGL" driver. Nevertheless, it was coded to use the OpenGL API. (You can still run Quake 2 on an Nvidia card in hardware mode, which obviously wouldn't be possible if the game were truly Glide-based.)
Bay Area radio host Ronn Owens got a bunch of copies of this email, and decided to have Mr. Ansary on his show. The interview was pretty insightful, and is available on-line:
Okay, it's a minor point, but Unreal featured volumetric fog before Quake 3. (And, for that matter, Unreal's volume fog wasn't restricted to ground fog.)
If there's one thing I learned in art school, it's that anything can be art if placed in the right context. As such, "Is it art?" isn't really a very useful question.
I think we're much better off asking if a thing is good art, or interesting art, or provocative art, or inspiring art.
I downloaded Aaron and got it to make a few pictures... it was sort of interesting, all in the same style, with the same kinds of marks. They all seem to be variations on a theme: human characters within interior spaces. My suggestion to the engineer(s) would be to focus some attention on the full range of human facial expressions. The facial expressions currently seem a little arbitrary; we're so sophisticated in the way we read faces that it's easy for rendered faces to seem bland and expressionless.
Also, it would be nice to see some simulation of lighting in a scene -- not fancy or even correct lighting, but something to capitalize on our response to moody lighting on figures and architecture.
If you're not squeamish about the chemically-modified lifestyle, melatonin can really help.
I take 1mg sublingual melatonin each night. It's not to help me get to sleep (never had a problem there) but it's *critical* for me to get up. When I don't take the melatonin, I have a terribly difficult time getting up, waking up, and generally starting the day. Melatonin helps me maintain that internal alarm clock and wake up (on my own!) at the same time every day.
If you choose to try melatonin, I suggest:
1. Sublingual! I've had poor results with the other kind.
2. Start out at 500 micrograms (.5 mg) if possible, and push the dosage up if you need to.
Yes, you're right; there's not nearly enough innovation in games today.
The problem is that innovative games don't consistently outsell "me-too" games. Software publishers are very cautious folk, with good reason -- in any given year, 10% of the PC games make 90% of the profit.
Consumers vote with their wallets. If publishers aren't rewarded for supporting innovative development houses, then innovative games won't get made.
C'mon, JK, you know better than that. Please leave spoilers out of the front page view.
It's particularly relevant in the case of Unbreakable, because part of what makes the movie work is the slow revelation of the film's true themes.
People for whom the label "Superhero Movie" is a turnoff... you might consider seeing it nevertheless. It's certainly not like any other Superhero Movie I've ever seen.
We've found that for the PS2, it's actually faster to just go ahead and draw all the backfaces than it is to cull them -- mainly because the PS2's strength is in its poly-pushing, and its main processor is relatively weak. (Another significant difference between the PS2 and a PC.)
Re:Library of Congress
on
Deja For Sale
·
· Score: 1
It's the right idea, but the Library of Congress has taken a decidedly anti-Internet stance. I recall reading something recently wherein the LoC's head guy said he won't be placing any of the Library's contents on the Net.
Thus, they seem like they'd be the last institution interested in hosting Deja...
BSP-based games like Quake, Unreal, etc. have very special requirements for their level data -- generally they need levels to be composed of convex polyhedra. That's why most such games have their own proprietary level editor. The critical feature of these editors is that they prevent designers from ever creating concave polyhedra in the first place -- thus, they're always guaranteed clean data when it comes time to process the level. If discreet wants up-and-coming 3d titles to make use of gMax, they'd be in a better position if they could either prevent concavity or somehow translate the level data before it reaches the game's processing tools.
(I know, Unreal papa Tim Sweeny has a Max->Unreal translator inside Unrealed, but I have yet to meet a programmer who can figure out how he did it:)
Convexity aside, I've had many arguments with other designers and artists over the use of 3ds Max as a level editing tool. The few times I've seen a designer work in Max, I was astonished at how slow the app became after loading even a medium-sized level. That, combined with an abysmal snap-to-grid system, would make it pretty difficult to convince me to use it...
*Very, very small - I carry it around in my pants pocket. *High-resolution - max 1800x1200. *Decent price - I got mine on the net for $550.
Cons:
*Only stores pics internally as JPEGs. This means that, even at highest quality, you still get a few compression artifacts. *No built-in USB. USB transfer package costs extra. *No optical zoom - no room for a zoom lens on this tiny thing.
I was torn between the Sony Cyberblahblah and the MX-2700. The Sony picture quality was higher, but the small form factor, convenience, and lower price of the 2700 won me over. I've been very happy with it.
That's right, one of the biggest innovations he mentioned is the trashing of precomputed lighting. Every FPS I can think of since Doom has featured a very static lighting model, usually through lightmaps. Carmack said that in the next iteration of his engine, all lighting (and shadows) will be generated in real-time. From a game-geek perspective, this is a huge step forward. Assuming he can pull it off, imagine the simplest scene with dynamic lighting... a plain room with a chair and a bare lightbulb overhead... as the lightbulb swings back and forth, the shadow of the chair crawls across the floor and up the wall a bit...
Slashdot Mirror
Except I find that Mozilla IS as featureful AND MORE stable than its proprietary competition, and I'm using it right now. I don't disagree with your final point,
"...as long as OSS is not as featureful OR as stable as its propietary competition, no one, but NO ONE will use it."
Although I'm not sure it applies to Mozilla anymore.
The best essay I've ever read on (narrative) game design theory is Crimes Against Mimesis by Roger Giner-Sorolla.
Superb stuff.
The cause of vaporware is in the marketing department, not the engineering department.
Far be it from me to defend any marketing department anywhere, but to be fair, marketing isn't the only source of project bloat. Anyone who's A) overambitious and B) in a position of power can create the sort of disasterous conditions you've described. It could be the marketing department, or any set of suits making decisions from upon high, or even a single megalomaniacal project director.
He took the words right out of my mouth.
Clever design + bad dev team = Deer Hunter, so there is an argument to be made for both sides.
To be fair, I recall reading somewhere that the Deer Hunter team only had about three months to make the entire game. Rather than a "bad dev team", that may simply have been rough circumstances.
There is no substitute for a good game design doc.
:)
I agree with your points here. I talked with a few of my co-workers about this thread, and we arrived at the tentative conclusion that J.C. is talking about sort of pie-in-the-sky speculative design work rather than concrete, plan-game-elements-out design work.
Likewise, I think he's probably coming off harsher than he means with regard to design documents. I could be wrong, but I think he's mainly concerned with the inflexibility that may accompany a design doc -- a design must be a living, adaptive thing that grows to meet the needs of the project. Unfortunately, it's easy to get attached to work you've done and cling to ideas that may no longer be appropriate.
I sincerely hope that Carmack doesn't just feel this way because he doesn't want some damn designer telling him what his priorities should be. That would be pretty disappointing, because that interdependency is part of the fundamental teamwork necessary to ship a game. Design sets priorities for the Art and Engineering groups, and likewise, they set priorities for each other and for the Design group. That's just the way it is.
Damn. I had to open my big mouth. Now I'll never get a job at id.
I'd like to know exactly how J.C. defines "abstract creative design" and "strategic creativity". I'm a game designer, and most projects I've seen suffered from a *lack* of a clear, comprehensive design up front. I don't know if that's what he means or not.
Well... where I work, that's called an "A bug". We can't allow situations where the player can render the game impossible to finish.
Imagine this: You collapse said ceiling, then save your game. Now... you didn't know that this would block your only path out, so it's not really your *fault*. But now that you've saved the game, and you've made it impossible to finish, you'll have to start over (or, more likely, throw the game in the trash).
Along with bad crash bugs, this is the worst sort of game bug.
Actually, a PC/PS2 game called Red Faction featured something it called "Geo-Mod" technology, which was basically real-time boolean subtraction to simulate holes in walls, etc.
It wasn't as interesting as it could have been, from a game-play perspective. The problem is that we game developers are used to making environments *very* static -- often based on BSP-trees and such. This saves the engineers a lot of headaches, but guarantees rock-solid, unchanging environments.
What you're talking about, with full structural physics and everything... it's going to be a while. I'd love to see an area where, if you destroy enough pillars/supports, the floor above you comes crashing down. Unfortunately, that's a sh|tload of math, and the environment would have to be created in such a way that it doesn't close off any critical paths when collapsed, etc etc etc.
It's a tough problem. But consider this: We're just now starting to get shadows that move when you move lights around! (That's what Carmack's working on right now...)
(pushes glasses up nose)
As Stiletto already noted, Direct3D and OpenGL don't really differ in performance anymore (if they ever did in the first place).
Also, Quake 2 wasn't really Glide *per se*. It was OpenGL that interfaced a custom 3DFX 'MiniGL" driver. Nevertheless, it was coded to use the OpenGL API. (You can still run Quake 2 on an Nvidia card in hardware mode, which obviously wouldn't be possible if the game were truly Glide-based.)
Its already on the net.
But the quality is much lower than
you get in theatre.
That's the worst haiku I've ever read.
Bay Area radio host Ronn Owens got a bunch of copies of this email, and decided to have Mr. Ansary on his show. The interview was pretty insightful, and is available on-line:
http://www.sftalk.com/tamim/tamim.ram
Okay, it's a minor point, but Unreal featured volumetric fog before Quake 3. (And, for that matter, Unreal's volume fog wasn't restricted to ground fog.)
If there's one thing I learned in art school, it's that anything can be art if placed in the right context. As such, "Is it art?" isn't really a very useful question.
I think we're much better off asking if a thing is good art, or interesting art, or provocative art, or inspiring art.
I downloaded Aaron and got it to make a few pictures... it was sort of interesting, all in the same style, with the same kinds of marks. They all seem to be variations on a theme: human characters within interior spaces. My suggestion to the engineer(s) would be to focus some attention on the full range of human facial expressions. The facial expressions currently seem a little arbitrary; we're so sophisticated in the way we read faces that it's easy for rendered faces to seem bland and expressionless.
Also, it would be nice to see some simulation of lighting in a scene -- not fancy or even correct lighting, but something to capitalize on our response to moody lighting on figures and architecture.
If you're not squeamish about the chemically-modified lifestyle, melatonin can really help.
I take 1mg sublingual melatonin each night. It's not to help me get to sleep (never had a problem there) but it's *critical* for me to get up. When I don't take the melatonin, I have a terribly difficult time getting up, waking up, and generally starting the day. Melatonin helps me maintain that internal alarm clock and wake up (on my own!) at the same time every day.
If you choose to try melatonin, I suggest:
1. Sublingual! I've had poor results with the other kind.
2. Start out at 500 micrograms (.5 mg) if possible, and push the dosage up if you need to.
...I look forward to seeing Metallica embrace RIAA's new plans.
Yes, you're right; there's not nearly enough innovation in games today.
The problem is that innovative games don't consistently outsell "me-too" games. Software publishers are very cautious folk, with good reason -- in any given year, 10% of the PC games make 90% of the profit.
Consumers vote with their wallets. If publishers aren't rewarded for supporting innovative development houses, then innovative games won't get made.
C'mon, JK, you know better than that. Please leave spoilers out of the front page view.
It's particularly relevant in the case of Unbreakable, because part of what makes the movie work is the slow revelation of the film's true themes.
People for whom the label "Superhero Movie" is a turnoff... you might consider seeing it nevertheless. It's certainly not like any other Superhero Movie I've ever seen.
We've found that for the PS2, it's actually faster to just go ahead and draw all the backfaces than it is to cull them -- mainly because the PS2's strength is in its poly-pushing, and its main processor is relatively weak. (Another significant difference between the PS2 and a PC.)
It's the right idea, but the Library of Congress has taken a decidedly anti-Internet stance. I recall reading something recently wherein the LoC's head guy said he won't be placing any of the Library's contents on the Net.
Thus, they seem like they'd be the last institution interested in hosting Deja...
This is a critical point; can someone moderate it up (even more)?
Really great points. Could someone moderate that message up?
BSP-based games like Quake, Unreal, etc. have very special requirements for their level data -- generally they need levels to be composed of convex polyhedra. That's why most such games have their own proprietary level editor. The critical feature of these editors is that they prevent designers from ever creating concave polyhedra in the first place -- thus, they're always guaranteed clean data when it comes time to process the level. If discreet wants up-and-coming 3d titles to make use of gMax, they'd be in a better position if they could either prevent concavity or somehow translate the level data before it reaches the game's processing tools.
:)
(I know, Unreal papa Tim Sweeny has a Max->Unreal translator inside Unrealed, but I have yet to meet a programmer who can figure out how he did it
Convexity aside, I've had many arguments with other designers and artists over the use of 3ds Max as a level editing tool. The few times I've seen a designer work in Max, I was astonished at how slow the app became after loading even a medium-sized level. That, combined with an abysmal snap-to-grid system, would make it pretty difficult to convince me to use it...
erik
Pros:
*Very, very small - I carry it around in my pants pocket.
*High-resolution - max 1800x1200.
*Decent price - I got mine on the net for $550.
Cons:
*Only stores pics internally as JPEGs. This means that, even at highest quality, you still get a few compression artifacts.
*No built-in USB. USB transfer package costs extra.
*No optical zoom - no room for a zoom lens on this tiny thing.
I was torn between the Sony Cyberblahblah and the MX-2700. The Sony picture quality was higher, but the small form factor, convenience, and lower price of the 2700 won me over. I've been very happy with it.
erik
That's right, one of the biggest innovations he mentioned is the trashing of precomputed lighting. Every FPS I can think of since Doom has featured a very static lighting model, usually through lightmaps.
Carmack said that in the next iteration of his engine, all lighting (and shadows) will be generated in real-time. From a game-geek perspective, this is a huge step forward.
Assuming he can pull it off, imagine the simplest scene with dynamic lighting... a plain room with a chair and a bare lightbulb overhead... as the lightbulb swings back and forth, the shadow of the chair crawls across the floor and up the wall a bit...
It's going to be pretty sweet.
erik