And that's to make it a bloody desktop operating system. The instant you do that, it's just popular software being exploited by big corporations. So what if it's GPL'd? Microsoft will just release Winux and then it won't be your OS. So everyone will have to start over with a new community OS.
I really don't understand why so many people want Linux to gain mainstream acceptance. Its biggest appeal was always that it was a hard core hacker OS, invented by and for the community, with none of the garbage required by the populace at large.
Let it stay that way, and let the popular OS's do the pandering to the masses for as long as they can stay in business.
No one is threatening anyone. Xbox attracts developers by demonstrating an incredible lack of spine and kissing their own hiney. I'm not knocking MS for this, mind you; it's drawn a lot of developers into the fold. The idea of a developer-friendly platform is attractive both to old console developers and PC developers tired of the inhomogeniety of their computing base.
For the technologically savvy, $17 is chump change. I sometimes spend more than that going to dinner. Pirates are, in general, not freedom-leaving heroes against repressive corporations, but a bunch of cheap wanna-bes.
It really kind of sucks. Music trading used to be a pretty cool thing (and still is outside the mainstream, although the name itself has gotten tainted). Tapers would bring taped concerts from the Dead or Phish or whatever into circulation within a certain community --- the Internet revolutionized that. But now it has to go underground again.
Too bad really; even though it was sometimes illegal to tape bands, it was a legality few people in music really cared about, since most tapers bought the band's regular albums anyway, as well as paying to go to their concerts.
Committee0based standards aren't the way to do graphics. The industry changes too fast. I don't know if you've ever dealt with a standards committee, but it's a nightware. Having a single core research/industry force define an API has its advantages. SGI used to be that force; now Microsoft is. Maybe NVidia or someone else will be the force in the future.
This way, DirectX or OpenGL or whatever is the standard, and card manufacturers implement it if they want consumers to buy their card. In five years, when card technologies change or maybe when we use O2-like architectures in PCs, or whatever, we will have a different standard to define what is good. Then, whether or not the pixel shading language used in DirectX is good or not is irrelevant. Of course, this does have a downside in that certain games might no longer function on future computers, but this is a common risk.
This does mean game/graphics people have to work harder, but that's why we get the big money, right? Er, wait a minute . . . . Doh!
This is a reply in two parts, first, to some specific comments, and then to stuff I should have said in the first place.
"And of course the DX3, DX5, DX7, DX8 interfaces never changed eh?:)
DirectX is inclusive. If it was supported by DX3, it'll still work. Code written to support OpenGL extensions, however, often breaks. That's the point of an extension; otherwise, it'd be in the standard. Some extensions are likely to remain in place, others, however, are vendor-specific and might just end up vapor.
NVidia loves DirectX more --- well, most of my friends there were pretty pissed at Microsoft for delaying DirectX repeatedly, but that's another issue. NVidia likes whatever the public buys.
Right now, games are using more DirectX, meaning that those APIs are more important, at least outside of the Quadro market.
As for Alias, of course they love OpenGL, Alias is an SGI product (SGI owns OpenGL's trademark, remember?). But then again, Alias isn't a game modelling program; OpenGL is pretty good for architectural and CAD applications, probably better than Direct3D.
The ultimate problem is that SGI is no longer the stable foundation that they once were. They were integral IMO to the advancement on OpenGL. No one has a similar commitment. NVidia certainly won't advocate OpenGL for the sake of OpenGL. They'll support it and support it well as long as there is demand, but I don't see them pushing the standard.
DirectX, on the other hand, advances because of the heavy-handed advocacy of Microsft and because there is a research division full of top graphics people there (Glassner, Hoppe, Cohen, etc).
Without one hardcore research/industry advocate pushing the technology, OpenGL will continue decrepifying. Different companies will create different non-compatible extensions for new effects, while the standard itself becomes bogged down in standards committee arguments, leaving it lagging many years behind the state of the art.
Note, I'm not saying that there's anything critically wrong with the OpenGL API itself; until DX 7 rolled around, I wouldn't approach DX without a gas mask and a really long pointy stick. But Microsoft has responded well to the community, vastly improving the interface, something that we can expect to continue in future versions.
As a side note, though, this entire article is sort of a red herring, although few people have mentioned it. 3D is not necessary for good games, at least not certain types of games, many of which are popular with Linux users. 3D can,in fact, be bad for RTS games, where it can be difficult to get a grasp on the tactics of a battle when the viewpoint can spin around, zoom, and where mountains can obscur some of your forces (consider Warcraft III's locking of the viewpoint, for instance, even though meshes are used for mobiles). Plus no card can render mobiles with sufficient tentacles, hair, gnats, floating eyes, or whatever --- and we might really want these things, consider the difficulties of rendering lots of beholders from Baldurs Gate II or many of the Zerg in Starcraft in 3D. In isometric games, sprites can do this trivially, since it's just a picture. After all, graphics is just a tool used to better a game;it is only a tool, however, not the game itself, and it can get in the way of the quality of the game.
More importantly, these "old-style" games can be coded in the true Linux tradition for free. Leave commercial games for commercial OS's. The only reason people don't code games like this for free for Linux is the worry that the code will break --- but that's changing as driver support becomes more standardized.
Of course, there are several companies make DirectX compliant implementations for Linux, so perhaps it's irrelevant anyway.
Using OpenGL extensions is a risky business --- they're not all necessarily the same and they're often not supported on different platforms. Further, because they're extensions, future support is questionable, or if supported, the interfaces might change. Some extensions, like mult-texturing to GL 1.2, will likely remain the same, but that's because it's an ARB-reviewed extension.
Further, OpenGL is typically supported as an abstraction on top of the DX driver code anyway nowadays (since it's pretty much a subset of DX functionality).
Anyway, who knows what the future of OpenGL is? It doesn't have anywhere near the support it used to. Once, it was nearly standard, now it's becoming more niche.
Frankly, DirectX, especially v8, it better than OpenGL, and not just because of vertex and pixel shaders. There's a lot of other free code that goes with DX8 in the extension libraries --- progressive meshes, skinnable meshes, automatic code for computing bounding boxes, etc. One might still choose to write their own optimized versions of these routines, but at the very least, using D3DX8 gives you a faster prototype.
As for "crippling" your software to fit hardware specs, well, you always have to do that, whether you use OpenGL or DirectX or any other API. After all, you have to query for extension support in OpenGL if you want to use extensions. Otherwise, your code is broken. The same is true of DX.
You would especially have to query for this is you were using NVIDIA's OpenGL SDK.
So, you can sacrifice simple Linux support and use DX, or you can prolong your development using OpenGL and get your Windows project out later (note, I'm assuming that one's writing a 3d-intensive appilcation, OpenGL is probably better for 2d and 2 and half d due to cross-platform portability issues).
Note, I'm not saying people are stupid for using OpenGL. It can be done. Any reasonable card will provide the same functionality, although it typically isn't standardized across cards and might be less supported. If you're writing the engine close to the hardware, OpenGL might be a good choice (of course, Dx has always been more flexible in making assumptions involving synchronicity and state, so this is only true if one really wants Linux).
Nah, they haven't figured out that Vietnam was really caused by John Wayne movies and that WWII was really caused by TS Eliot's The Wasteland; why would people suddenly be able to string together logical thoughts any faster now, and link Waco and Columbine to the extraordinarily obvious causal source video games?
After all, it's obvious that being a heroic Britsh marine sneaking into a Nazi base to defeat the forces of evil promotes Nazism. Sometimes, I just don't get the stupidity of the American people.
Next thing you know, people won't be able to figure out the Wayne Newton -> George Bush ->
Barbara Streisand -> mind control with microwave beams connection. Just remember to wear your tin foil hat!
Oh yeah, for readers other the poster, you'll gain a great deal of understanding of trolls if you read books on the symptoms of paranoid schizophrenia. Maybe these connections will make sense to you too then.
While they do develop in OpenGL, making the point moot, Direct3D and OpenGl are quite similar. In any event, most of the real graphics code is specific to neither DirectX nor OpenGL, but involves higher-level algorithms such as portal,
terrain management, and animation-related code.
This is not to say that porting is trivial --- it is just that porting from DirectX to OpenGl is a lot like porting from Win32 to Linux, except easier. It's mostly low-level boilerplate code in both cases. Further, it doesn't require someone who groks graphics. D3D and OpenGL can easily be learned sufficiently to do it in no time at all by a decent programmer --- one doesn't really need to understand what the code is doing at an algorithmic level.
It's hardly lack of creativity. Most players (and readers of fantasy) like a particular kind of world --- they like soemthing that's orcish, elvish, dwarvish, etc. Comfortable ideas make the game/story simple to become immsersed in, especially if you like that kind of thing.
Look at Everquest vs Asheron's Call. AC took the route of using new monsters, EQ is very classic; for many people who played EQ, the difference was undesirable (although for others, it was a plus).
The point is, classic references are not examples of lack of creativity, it's what you _do_ with those examples that matter.
Honestly, though, Ernst Adam's various editorials on the topic of game design are less than edifying in general. Perhaps he should look for more creative and deep solutions to problems himself, rather than write articles that only wanna-be game designers will interesting.
As was stated by the previous comment to this one, this already is happening to some extent. However, be aware that licensing an engine doesn't necessarily cut costs. You have to pay the licensing fee, and then you have to pay programmer costs to get used to the engine. Getting used to an engine can take quite a while. From a business perspective, licensing can still be worth it as you just focus on content, rather than needing to worry about programming your own engine.
As a broader question, though, as regards the poster: most people are really talking about hardcore games here. Hardcore games need to be cutting edge. Most game programmers read Siggraph, Eurographics, and the other computer graphics conferences/journals religiously. As soon as a new idea comes out, it is adopted if feasible. Look how fast portal culling became the rage; while the original idea dates back a bit (Teller and Funkhouser are my first recollection in 92, but there was a predecessor, just don't want to look it up), portal culling as it currently is didn't become feasible until Portals and Mirrors in Siggraph 95, I think it was. Pretty much as soon as that paper was published, it was immediately adopted into new game designs. But doing so required a substantial overhaul of the engine. While theoretically portal cullers are similar to BSP tree renderers, they are not identical.
However, it is not unfeasible that a new vein of games might evolve. Text Muds, for example, were popular long after text games were passe, although part of that was due to the internet aspect of Muds, when few games were on the internet. But people like to hack games in their spare time, and some of those games are quite fun.
However, someone still needs to write an engine that is cutting edge; many people demand it. Whether the game creators license it or use it directly is another issue. In any event, it is unlikely that such engines will be free, although slightly more out-of-date ones might well be.
Ultimately, most game developers/programmers just want to get paid reasonable salaries for doing what they enjoy. Any business model that safely supports that is good to go.
Actually, the best selling games are quite easy. Who Wants to Be a Millionaire and
Barbie Pet Rescue were two examples of top 10 games in 2000 based on units sold: as a bonus, these games also had some of the lowest development costs.
Many other similar games sold quite well compared to their cost of development.
Hardcore games, on the other hand, have a history of being cutting edge and costing a lot to develop.
In the future, we're likely to see more titles like Barbie Pet Rescue in the top 10 sellers. This isn't bad, they're just completely different markets. Although it bugs game reviewers;) .
Several comments need to be made here.
First, the game industry is not the same as the OS industry. Regular consumers buy games, not corporations. For a corporation, paying thousands for customer support is irrelevant. For a regular consumer, paying _any money at all_ for support is often unacceptable.
Second, the profit margin on games is extremely small. You say Blizzard and Sony make money; well, goodie for them. They're two companies. Blizzard also happens to have produced that vast majority of best selling games in recent history.
Reality is that only 1 in 10 games makes money. Most companies are created and disappear so fast, you never even know they existed; even those that make a good title typically make barely any profit and then turn into one hit wonders. Further, game programmers aren't the most highly paid programmers as is. Most do it because they like games; otherwise, they'd take a job with fewer hours per week and a substantial salary increase. If you cut the salaries even further, many people will just say it's too much stress for too little remuneration, get a forty hour a week job, and write tetris or asteroids clones in their spare time.
Admittedly, this is under the current publisher model, where money is lost to middlemen. One could imagine downloading software for fees, for instance, which has been done for small games like tetris clones and worked well there, but seems unlikely to work well in games with substantial development costs. In any event, developing cutting edge games costs lots of money. Before you go advocating open/free source games, actually go out there and write a cutting edge game: create a portal culler, a terrain engine, custom art and sound, solid game design and make it all efficient. It's a lot of work, which implies that it costs a lot to get it built. How are you going to offset these costs and still make a profit?
That's really the thing that annoys me. I hear so many people advocate open/free source software for games, but so few give reasonable suggestions for how companies are to remain in the black. I'm not saying that games can't be eventually open sourced --- once a game reaches "price-saver mode" about a year after release and goes down to $20-30/copy, this is reasonable as the bulk of sales will have already been made and open sourcing it might jump start mod development; even if you don't make money off incidental sales from mod developers and players, at least you get karma with the hardcore community ( a la Id).
Of course, there are free game projects, and for those, open source/free source works fine. Some of those games are actually quite fun, although I've yet to see one that didn't lag technologically by a substantial amount. I usually work on free projects myself in my spare time. But then again, I don't need to stay in the black since I'm doing it for free.
Of course, maybe some games could be subscription-based like Everquest. Well, that's great for games like EQ, but perhaps not so great for single-player games. Even for games like EQ and Asheron's Call, it isn't necessarily so good. If everything is open source or free source, then people can make their own servers and not pay the subscription fees. I suppose you could still sell CDs with new content and hope that more people than not bought it rather than ripped it, but it seems risky.
Also, at the risk of making this post even more disorganized;), I want to address the piracy issue. Yes, games are pirated and some game companies still make profits. So what? The Finns have been estimated to have a 90% piracy rate on software, but this doesn't put MS out of business? The point is, game piracy is fairly low in most places, largely because most people aren't cheapskates and prefer not to do illegal things. If you make payment an honor system issue, however, piracy rates will go up by an extreme amount --- just basic psychology here, people avoid a stick, even if it's very unlikely they'll get hit, but they don't often do things just for the sake of doing the right thing, particularly if it isn't something _obviously_ ethical like tithing. This is just basic psychology.
People need to start thinking about these issues instead of what's good for them in the immediate future. Pirates need to stop being so bloody cheap and lazy. Come on, it's $50 and a walk to a store or a trip to amazon.com, is that too much to ask?
Of course, some people might not mind playing games that are 5-8 years behind the current cutting edge. For those people, open/free source is great. Hell, I used to play text Muds as late as 1997, where such models worked great. If you want to play Hearts, you can usually do that for free too on the Internet. But if you want to play something like Giants, Black & White, or Tribes 2 in 2001, expect such a model to fail. (Speaking of which, Tribes 1 suffered more heavily from piracy than most games, I'm surprised Tribes 2 came out at all --- if such piracy rates continue with Tribes 2, I wouldn't be surprised if there never is a Tribes 3).
In response to the person who defined hacking as "quick-and-dirty work". Actually, hacking refers to an even older meaning. A "hack" used to be a playwright who literally hacked piece out of another person's play with a knife and glued those passages into their own play. Quite common in Shakespeare's time. Mark Twain and others gave the word "hack" it's more derogatory connotation.
The term hacker used in computer circles more properly refers to the original meaning: ie, someone who takes someone else's code and modifies it to make something better.
The media has adapted the more proper meaning of hacker to refer to someone who breaks into other people's systems. This is not correct, and is an insult to people who actually are hackers, many of whom do not break into other people's systems, even to "help" them out.
It's true that some hackers do break into other people's systems; it used to be that to break into other people's systems you pretty much had to be a hacker to understand what was going on, or had to have the motivation to search through dry and hard-to-find technical manuals: now you just log onto IRC or hang out on the wrong websites and exploit well-known holes on weak websites. Wahoo!
Breaking into a system does not make you a hacker.
I often wonder why people consider Intel evil. After all, Intel donates substantial numbers of PCs to academic departments around the country (and world, I think). Before AMD starting cutting into Intel's profit share, we used to get a new shipment of top-line PCs about every six months. Plus, Intel funds a lot of research, funds lots of interns, etc. To my knowledge, AMD doesn't behave in a reciprocal fashion. I know there's been no offers from them to MIT, at least.
Sure Intel's engaged in some apparently predatory business practices, but do you really think AMD won't? And how well do you think AMD will do without clearing the path in front of them?
How many people? A few thousand? Are you kidding? This is a console system people can take apart and hack. Trust me, there's a substantial hobbyist market that will love to play with the Indrema. After all, it's only $300.
Is that market enough to sustain Indrema --- maybe, maybe not.
I do know this, though. People always naysay good ideas in droves and gloat when they're right; of course, when they're wrong, they quietly revise history.
Gee, Verant is evil, they don't want people using programs that allow them to cheat. Of course, for Lum the Mad's main constituency, one can see why
this might be a problem.
Verant might be incompetent (ie, the store opening blunder and similar incidents, etc), but they're hardly evil.
In any event, Everquest was a great game. It held my attention for several months, which is unusual.
It also revolutionized a particular facet of the game market. I know it's influenced my own design choices, and has similarly influenced others (witness the huge spawn of RPGs since Everquest, something that I doubt it due solely to Ultima Online or Gemstone). Everquest also taught us what _not_ to do.
Slashdot Mirror
And that's to make it a bloody desktop operating system. The instant you do that, it's just popular software being exploited by big corporations. So what if it's GPL'd? Microsoft will just release Winux and then it won't be your OS. So everyone will have to start over with a new community OS.
I really don't understand why so many people want Linux to gain mainstream acceptance. Its biggest appeal was always that it was a hard core hacker OS, invented by and for the community, with none of the garbage required by the populace at large. Let it stay that way, and let the popular OS's do the pandering to the masses for as long as they can stay in business.
No one is threatening anyone. Xbox attracts developers by demonstrating an incredible lack of spine and kissing their own hiney. I'm not knocking MS for this, mind you; it's drawn a lot of developers into the fold. The idea of a developer-friendly platform is attractive both to old console developers and PC developers tired of the inhomogeniety of their computing base.
Um, maybe you should look into the Wen Ho Lee case a bit more. It pretty apparent he was scapegoated.
For the technologically savvy, $17 is chump change. I sometimes spend more than that going to dinner. Pirates are, in general, not freedom-leaving heroes against repressive corporations, but a bunch of cheap wanna-bes.
It really kind of sucks. Music trading used to be a pretty cool thing (and still is outside the mainstream, although the name itself has gotten tainted). Tapers would bring taped concerts from the Dead or Phish or whatever into circulation within a certain community --- the Internet revolutionized that. But now it has to go underground again.
Too bad really; even though it was sometimes illegal to tape bands, it was a legality few people in music really cared about, since most tapers bought the band's regular albums anyway, as well as paying to go to their concerts.
Committee0based standards aren't the way to do graphics. The industry changes too fast. I don't know if you've ever dealt with a standards committee, but it's a nightware. Having a single core research/industry force define an API has its advantages. SGI used to be that force; now Microsoft is. Maybe NVidia or someone else will be the force in the future.
This way, DirectX or OpenGL or whatever is the standard, and card manufacturers implement it if they want consumers to buy their card. In five years, when card technologies change or maybe when we use O2-like architectures in PCs, or whatever, we will have a different standard to define what is good. Then, whether or not the pixel shading language used in DirectX is good or not is irrelevant. Of course, this does have a downside in that certain games might no longer function on future computers, but this is a common risk.
This does mean game/graphics people have to work harder, but that's why we get the big money, right? Er, wait a minute . . . . Doh!
This is a reply in two parts, first, to some specific comments, and then to stuff I should have said in the first place.
"And of course the DX3, DX5, DX7, DX8 interfaces never changed eh? :)
DirectX is inclusive. If it was supported by DX3, it'll still work. Code written to support OpenGL extensions, however, often breaks. That's the point of an extension; otherwise, it'd be in the standard. Some extensions are likely to remain in place, others, however, are vendor-specific and might just end up vapor.
Eh? NVidia loves OpenGL Alias|Wavefront loves OpeNGL
NVidia loves DirectX more --- well, most of my friends there were pretty pissed at Microsoft for delaying DirectX repeatedly, but that's another issue. NVidia likes whatever the public buys. Right now, games are using more DirectX, meaning that those APIs are more important, at least outside of the Quadro market.
As for Alias, of course they love OpenGL, Alias is an SGI product (SGI owns OpenGL's trademark, remember?). But then again, Alias isn't a game modelling program; OpenGL is pretty good for architectural and CAD applications, probably better than Direct3D.
The ultimate problem is that SGI is no longer the stable foundation that they once were. They were integral IMO to the advancement on OpenGL. No one has a similar commitment. NVidia certainly won't advocate OpenGL for the sake of OpenGL. They'll support it and support it well as long as there is demand, but I don't see them pushing the standard. DirectX, on the other hand, advances because of the heavy-handed advocacy of Microsft and because there is a research division full of top graphics people there (Glassner, Hoppe, Cohen, etc).
Without one hardcore research/industry advocate pushing the technology, OpenGL will continue decrepifying. Different companies will create different non-compatible extensions for new effects, while the standard itself becomes bogged down in standards committee arguments, leaving it lagging many years behind the state of the art.
Note, I'm not saying that there's anything critically wrong with the OpenGL API itself; until DX 7 rolled around, I wouldn't approach DX without a gas mask and a really long pointy stick. But Microsoft has responded well to the community, vastly improving the interface, something that we can expect to continue in future versions.
As a side note, though, this entire article is sort of a red herring, although few people have mentioned it. 3D is not necessary for good games, at least not certain types of games, many of which are popular with Linux users. 3D can ,in fact, be bad for RTS games, where it can be difficult to get a grasp on the tactics of a battle when the viewpoint can spin around, zoom, and where mountains can obscur some of your forces (consider Warcraft III's locking of the viewpoint, for instance, even though meshes are used for mobiles). Plus no card can render mobiles with sufficient tentacles, hair, gnats, floating eyes, or whatever --- and we might really want these things, consider the difficulties of rendering lots of beholders from Baldurs Gate II or many of the Zerg in Starcraft in 3D. In isometric games, sprites can do this trivially, since it's just a picture. After all, graphics is just a tool used to better a game;it is only a tool, however, not the game itself, and it can get in the way of the quality of the game.
More importantly, these "old-style" games can be coded in the true Linux tradition for free. Leave commercial games for commercial OS's. The only reason people don't code games like this for free for Linux is the worry that the code will break --- but that's changing as driver support becomes more standardized.
Of course, there are several companies make DirectX compliant implementations for Linux, so perhaps it's irrelevant anyway.
Fahrenheit is dead.
Using OpenGL extensions is a risky business --- they're not all necessarily the same and they're often not supported on different platforms. Further, because they're extensions, future support is questionable, or if supported, the interfaces might change. Some extensions, like mult-texturing to GL 1.2, will likely remain the same, but that's because it's an ARB-reviewed extension. Further, OpenGL is typically supported as an abstraction on top of the DX driver code anyway nowadays (since it's pretty much a subset of DX functionality). Anyway, who knows what the future of OpenGL is? It doesn't have anywhere near the support it used to. Once, it was nearly standard, now it's becoming more niche. Frankly, DirectX, especially v8, it better than OpenGL, and not just because of vertex and pixel shaders. There's a lot of other free code that goes with DX8 in the extension libraries --- progressive meshes, skinnable meshes, automatic code for computing bounding boxes, etc. One might still choose to write their own optimized versions of these routines, but at the very least, using D3DX8 gives you a faster prototype. As for "crippling" your software to fit hardware specs, well, you always have to do that, whether you use OpenGL or DirectX or any other API. After all, you have to query for extension support in OpenGL if you want to use extensions. Otherwise, your code is broken. The same is true of DX. You would especially have to query for this is you were using NVIDIA's OpenGL SDK. So, you can sacrifice simple Linux support and use DX, or you can prolong your development using OpenGL and get your Windows project out later (note, I'm assuming that one's writing a 3d-intensive appilcation, OpenGL is probably better for 2d and 2 and half d due to cross-platform portability issues). Note, I'm not saying people are stupid for using OpenGL. It can be done. Any reasonable card will provide the same functionality, although it typically isn't standardized across cards and might be less supported. If you're writing the engine close to the hardware, OpenGL might be a good choice (of course, Dx has always been more flexible in making assumptions involving synchronicity and state, so this is only true if one really wants Linux).
Nah, they haven't figured out that Vietnam was really caused by John Wayne movies and that WWII was really caused by TS Eliot's The Wasteland; why would people suddenly be able to string together logical thoughts any faster now, and link Waco and Columbine to the extraordinarily obvious causal source video games?
After all, it's obvious that being a heroic Britsh marine sneaking into a Nazi base to defeat the forces of evil promotes Nazism. Sometimes, I just don't get the stupidity of the American people.
Next thing you know, people won't be able to figure out the Wayne Newton -> George Bush -> Barbara Streisand -> mind control with microwave beams connection. Just remember to wear your tin foil hat!
Oh yeah, for readers other the poster, you'll gain a great deal of understanding of trolls if you read books on the symptoms of paranoid schizophrenia. Maybe these connections will make sense to you too then.
While they do develop in OpenGL, making the point moot, Direct3D and OpenGl are quite similar. In any event, most of the real graphics code is specific to neither DirectX nor OpenGL, but involves higher-level algorithms such as portal, terrain management, and animation-related code. This is not to say that porting is trivial --- it is just that porting from DirectX to OpenGl is a lot like porting from Win32 to Linux, except easier. It's mostly low-level boilerplate code in both cases. Further, it doesn't require someone who groks graphics. D3D and OpenGL can easily be learned sufficiently to do it in no time at all by a decent programmer --- one doesn't really need to understand what the code is doing at an algorithmic level.
It's hardly lack of creativity. Most players (and readers of fantasy) like a particular kind of world --- they like soemthing that's orcish, elvish, dwarvish, etc. Comfortable ideas make the game/story simple to become immsersed in, especially if you like that kind of thing.
Look at Everquest vs Asheron's Call. AC took the route of using new monsters, EQ is very classic; for many people who played EQ, the difference was undesirable (although for others, it was a plus).
The point is, classic references are not examples of lack of creativity, it's what you _do_ with those examples that matter.
Honestly, though, Ernst Adam's various editorials on the topic of game design are less than edifying in general. Perhaps he should look for more creative and deep solutions to problems himself, rather than write articles that only wanna-be game designers will interesting.
As was stated by the previous comment to this one, this already is happening to some extent. However, be aware that licensing an engine doesn't necessarily cut costs. You have to pay the licensing fee, and then you have to pay programmer costs to get used to the engine. Getting used to an engine can take quite a while. From a business perspective, licensing can still be worth it as you just focus on content, rather than needing to worry about programming your own engine.
As a broader question, though, as regards the poster: most people are really talking about hardcore games here. Hardcore games need to be cutting edge. Most game programmers read Siggraph, Eurographics, and the other computer graphics conferences/journals religiously. As soon as a new idea comes out, it is adopted if feasible. Look how fast portal culling became the rage; while the original idea dates back a bit (Teller and Funkhouser are my first recollection in 92, but there was a predecessor, just don't want to look it up), portal culling as it currently is didn't become feasible until Portals and Mirrors in Siggraph 95, I think it was. Pretty much as soon as that paper was published, it was immediately adopted into new game designs. But doing so required a substantial overhaul of the engine. While theoretically portal cullers are similar to BSP tree renderers, they are not identical.
However, it is not unfeasible that a new vein of games might evolve. Text Muds, for example, were popular long after text games were passe, although part of that was due to the internet aspect of Muds, when few games were on the internet. But people like to hack games in their spare time, and some of those games are quite fun.
However, someone still needs to write an engine that is cutting edge; many people demand it. Whether the game creators license it or use it directly is another issue. In any event, it is unlikely that such engines will be free, although slightly more out-of-date ones might well be.
Ultimately, most game developers/programmers just want to get paid reasonable salaries for doing what they enjoy. Any business model that safely supports that is good to go.
Actually, the best selling games are quite easy. Who Wants to Be a Millionaire and Barbie Pet Rescue were two examples of top 10 games in 2000 based on units sold: as a bonus, these games also had some of the lowest development costs.
Many other similar games sold quite well compared to their cost of development.
Hardcore games, on the other hand, have a history of being cutting edge and costing a lot to develop.
In the future, we're likely to see more titles like Barbie Pet Rescue in the top 10 sellers. This isn't bad, they're just completely different markets. Although it bugs game reviewers ;) .
Several comments need to be made here. First, the game industry is not the same as the OS industry. Regular consumers buy games, not corporations. For a corporation, paying thousands for customer support is irrelevant. For a regular consumer, paying _any money at all_ for support is often unacceptable. Second, the profit margin on games is extremely small. You say Blizzard and Sony make money; well, goodie for them. They're two companies. Blizzard also happens to have produced that vast majority of best selling games in recent history. Reality is that only 1 in 10 games makes money. Most companies are created and disappear so fast, you never even know they existed; even those that make a good title typically make barely any profit and then turn into one hit wonders. Further, game programmers aren't the most highly paid programmers as is. Most do it because they like games; otherwise, they'd take a job with fewer hours per week and a substantial salary increase. If you cut the salaries even further, many people will just say it's too much stress for too little remuneration, get a forty hour a week job, and write tetris or asteroids clones in their spare time. Admittedly, this is under the current publisher model, where money is lost to middlemen. One could imagine downloading software for fees, for instance, which has been done for small games like tetris clones and worked well there, but seems unlikely to work well in games with substantial development costs. In any event, developing cutting edge games costs lots of money. Before you go advocating open/free source games, actually go out there and write a cutting edge game: create a portal culler, a terrain engine, custom art and sound, solid game design and make it all efficient. It's a lot of work, which implies that it costs a lot to get it built. How are you going to offset these costs and still make a profit? That's really the thing that annoys me. I hear so many people advocate open/free source software for games, but so few give reasonable suggestions for how companies are to remain in the black. I'm not saying that games can't be eventually open sourced --- once a game reaches "price-saver mode" about a year after release and goes down to $20-30/copy, this is reasonable as the bulk of sales will have already been made and open sourcing it might jump start mod development; even if you don't make money off incidental sales from mod developers and players, at least you get karma with the hardcore community ( a la Id). Of course, there are free game projects, and for those, open source/free source works fine. Some of those games are actually quite fun, although I've yet to see one that didn't lag technologically by a substantial amount. I usually work on free projects myself in my spare time. But then again, I don't need to stay in the black since I'm doing it for free. Of course, maybe some games could be subscription-based like Everquest. Well, that's great for games like EQ, but perhaps not so great for single-player games. Even for games like EQ and Asheron's Call, it isn't necessarily so good. If everything is open source or free source, then people can make their own servers and not pay the subscription fees. I suppose you could still sell CDs with new content and hope that more people than not bought it rather than ripped it, but it seems risky. Also, at the risk of making this post even more disorganized ;), I want to address the piracy issue. Yes, games are pirated and some game companies still make profits. So what? The Finns have been estimated to have a 90% piracy rate on software, but this doesn't put MS out of business? The point is, game piracy is fairly low in most places, largely because most people aren't cheapskates and prefer not to do illegal things. If you make payment an honor system issue, however, piracy rates will go up by an extreme amount --- just basic psychology here, people avoid a stick, even if it's very unlikely they'll get hit, but they don't often do things just for the sake of doing the right thing, particularly if it isn't something _obviously_ ethical like tithing. This is just basic psychology.
People need to start thinking about these issues instead of what's good for them in the immediate future. Pirates need to stop being so bloody cheap and lazy. Come on, it's $50 and a walk to a store or a trip to amazon.com, is that too much to ask?
Of course, some people might not mind playing games that are 5-8 years behind the current cutting edge. For those people, open/free source is great. Hell, I used to play text Muds as late as 1997, where such models worked great. If you want to play Hearts, you can usually do that for free too on the Internet. But if you want to play something like Giants, Black & White, or Tribes 2 in 2001, expect such a model to fail. (Speaking of which, Tribes 1 suffered more heavily from piracy than most games, I'm surprised Tribes 2 came out at all --- if such piracy rates continue with Tribes 2, I wouldn't be surprised if there never is a Tribes 3).
In response to the person who defined hacking as "quick-and-dirty work". Actually, hacking refers to an even older meaning. A "hack" used to be a playwright who literally hacked piece out of another person's play with a knife and glued those passages into their own play. Quite common in Shakespeare's time. Mark Twain and others gave the word "hack" it's more derogatory connotation. The term hacker used in computer circles more properly refers to the original meaning: ie, someone who takes someone else's code and modifies it to make something better. The media has adapted the more proper meaning of hacker to refer to someone who breaks into other people's systems. This is not correct, and is an insult to people who actually are hackers, many of whom do not break into other people's systems, even to "help" them out. It's true that some hackers do break into other people's systems; it used to be that to break into other people's systems you pretty much had to be a hacker to understand what was going on, or had to have the motivation to search through dry and hard-to-find technical manuals: now you just log onto IRC or hang out on the wrong websites and exploit well-known holes on weak websites. Wahoo! Breaking into a system does not make you a hacker.
I often wonder why people consider Intel evil. After all, Intel donates substantial numbers of PCs to academic departments around the country (and world, I think). Before AMD starting cutting into Intel's profit share, we used to get a new shipment of top-line PCs about every six months. Plus, Intel funds a lot of research, funds lots of interns, etc. To my knowledge, AMD doesn't behave in a reciprocal fashion. I know there's been no offers from them to MIT, at least.
Sure Intel's engaged in some apparently predatory business practices, but do you really think AMD won't? And how well do you think AMD will do without clearing the path in front of them?
How many people? A few thousand? Are you kidding? This is a console system people can take apart and hack. Trust me, there's a substantial hobbyist market that will love to play with the Indrema. After all, it's only $300. Is that market enough to sustain Indrema --- maybe, maybe not. I do know this, though. People always naysay good ideas in droves and gloat when they're right; of course, when they're wrong, they quietly revise history.
Gee, Verant is evil, they don't want people using programs that allow them to cheat. Of course, for Lum the Mad's main constituency, one can see why this might be a problem. Verant might be incompetent (ie, the store opening blunder and similar incidents, etc), but they're hardly evil. In any event, Everquest was a great game. It held my attention for several months, which is unusual. It also revolutionized a particular facet of the game market. I know it's influenced my own design choices, and has similarly influenced others (witness the huge spawn of RPGs since Everquest, something that I doubt it due solely to Ultima Online or Gemstone). Everquest also taught us what _not_ to do.