http://news.quelsolaar.com/#post88

Discuss this article. The overdramatic comments after it are no big suprise. Neither will the potential comments in this thread be. Don't take sides and attack the other. Don't.


My Opinion: Cinematics are not a game. Cinematics are good storytelling to enhance a game as a whole. Interaction is paramount.
Don't disagree with an attitude of vengeful anger. Don't.

It's an interesting article. It misses a few important points, like the fact that mobile platforms are now id's primary outlet for smaller/faster-turnaround games, and they have actually released 6 such (at least) in the years since Doom 3. With that in mind I think one big sprawling ambitious title plus 6 smaller titles doesn't look too shabby at all. It may also be argued that Rage had to wait until hardware that was capable of running it was reasonably commonplace (and it might be argued by some that it would have been better to have waited a little longer, but let's not go there), and that maybe id overshot the mark on aiming for target hardware (I'm talking PC here) a little this time around.

I'm also finding myself disagreeing with the point regarding artists. There's plenty of evidence (e.g via Carmack's twitter) that coding on Rage continued right to the end too.

Overall though the core point seems reasonably valid; an extended turnaround time does seem to have hurt somewhat and measures should be taken to address that. I'm just not agreeing with the authors anolysis of the reasons for the time, nor wholly with the measures suggested.

Artists taking over the world sounds like a conspiracy theory you'd have difficulty selling even in Hollywood. Carmack makes one statement about seeing himself as providing a platform for artists to do whatever they want and everybody assumes he's been brain washed or something.

Id has stated repeatedly that they deliberately kept their company small in order to ensure they had the freedom to do what they loved. That was great when the games were much smaller and cheaper to develop, but eventually they had to expand the company which is what they've done with the id tech 5. The engine was specifically designed to make producing games faster and easier then ever before and they've added a lot of new talent to the payroll. Its kept Carmack busy doing other things then just programming and he's said he'll be glad to get back to producing a new game every 2-3 years.

That's quite a valid point. Every engine coder is actually doing much the same job - their role is to let content creators do their thing. A game is primarily content after all; the engine - however important it may be - starts and stops at being the medium through which you get to experience the content.

What's different about Carmack is that he's been very much the public face of id since Romero's departure, whereas most other lead engine coders keep a much lower profile. As a result he's kinda put himself in the firing line a little, and everything he says tends to get dissected to some degree or other, often excessively so.

I get the point that is made, but i consider this article a nostalgic view towards yesterday.
Games have become "planned" big business. The good old times when computer whizkids revolutionized the world by inventing new technologies in their living room are over.

If ID would still go the old school way of just inventing new, never before seen technologies without the manpower to fill these technologies with detail, they could sell their games as downloads on Xbox-live or PSN.
Or even worse: Would have already changed into a company just making demos or benchmarks like 3DMark and the likes.

ID (and Epic) were THE "3D-engine providers" over almost the last two decades, licensing their technology to other companies making great games with it.
With their Doom3 engine, they still went into the classic direction of presenting new, unseen stuff like their great shadowing system at the cost of texture/geometry detail and therefore lack of flexibility where Epic did exactly the opposite by building an engine that is able to render huge amounts of detail in different styles of visuals.

You can count the games using the Doom 3 engine by the fingers of one hand. And all of them looked pretty much the same.
Everybody else went for the Unreal engine. Guess why...

So building an engine that is capable of rendering insane amount of world detail with ease and making it easy to use for the designers is the next logical step to recapture some lost ground.
Gamers of today are not just interested in new effects, they want detail. The game sales clearly show it.
Gamers want huge worlds to explore. The times when adding shadows or lenseflares were generating a big smile on millions of gamers' faces are long gone.

As for raytracing as the holy grail of computer graphics:
I'm sure that very few gamers would see the difference between the standard rasterized and the performance heavy raytraced graphics.
Like most people wouldn't be able to hear the difference between a highly expensive hifi equipment and an Ipod.

Adding detail is definitely the way to go. And raytracing is not going to provide that.
We're already spending way more clockcycles and bandwidth on enhancing the visual appearance of polygons than actually calculating the polygons themselves and raytracing is a lot worse in that respect.
In fact, raytracing will give us the opposite of detail as it uses way more per-object calculations than the good old rasterizing, resulting in less objects with less detail that can be rendered per frame.
Look at the Pixar movies. Is that what we want for games? Nice lighting and effects and NO world detail.

It might be a good idea to reconsider the methods of rendering graphics in general instead of trying to enhance a dated technique.

Btw.: I always loved ID for their radical gamedesigns (Doom definitely kicks Super Marios butt!) and not their technology even if it was groundbraking.

Its not quite that bad. Metro 2033 was supposedly produced by a few guys in a basemant so its still quite possible for small innovative groups to make an impact.

As for rethinking graphics rendering, its a huge business and countless people are investing megabucks into finding the next big thing. Its not a question of ray tracing verses rasterization either, but of how to best use everything possible. To this day 3D games still use sprites if it makes more sense and its not like anyone's proposing throwing the baby out with the bathwater.

P.S.- If Doom still kicks Super Mario's butt its because it has the technology to do so! Can you imagine playing Doom as a 2D scroller? But seriously, many of the floor plans for Ultimate Doom were actually created by the modding community. Id made it easy for people to mod the game and then used the best mods people created in the final version. Many were actually based on real castle floor plans developed over eons to repel invading armies.

A funny this is that this is exactly what Rage is a first step towards. The megatexture technology is certainly revolutionary, a completely uniquely textured world is something that we've never seen before, and the tradeoffs chosen to enable it more or less chime exactly with your description of Unreal.

This makes some of the reaction towards Rage all the more puzzling, and I can only speculate that it threw people out of their comfort zones a little too much. Many reactions focussed on how textures looked up close, quality of bump maps and shadows, and all but completely ignored the non-repeating texturing and insane amounts of polygonal detail in the game. (Aside: Rage's textures have been compared to - for example - an N64. Which is odd as the N64 could push what seems like more polygonal detail than you see in screenshots of many modern shooters that focus on high resolution texturing and pixel shaders. It's all a matter of which tradeoff you choose, but you can't have both - yet.)

In the end I guess that people have become used to seeing games that look, feel, smell and taste a certain way over the past years, and when confronted with something different they don't really know how to handle it. For better or worse, people want to see textures up close, they want to see bumpmapping and they want to see shadowing. Or at least are so accustomed to them that they're not really aware of alternatives, and not certain what to think or do when confronted with one.

This is the inherent risk in trying something new and presenting it to the public for the first time. Daring to be different is almost like putting your balls on the chopping block. It's easy to forget that Quake also met with a bad reception and comparisons being made to Doom and Duke 3D that made it look poor, Q3A went head-to-head with the first UT and lost, and Doom 3 was met with fairly universal derision on release. And all 3 were downrated for entirely the wrong reasons, with the true revolution they presented not really becoming apparent until sometime later.

It will be interesting to find out if hindsight grants Rage a similar legacy.

Such well written replies I feel spoiled for the first time on this forum. I agree with most of the points except that for the part critically and publicly Rage is commended on it's amazing graphics. Sure it has its shortcomings that have been bashed way more than it deserved (low res textures) but let's not forget that most people think the game looks amazing. I remember the comaparisons of past engines as you pointed out Jimmy but the thing that those other games had going for it was also a revolutionary game. Quakes multiplayer was so darn mind blowing at the time it really was the breakthrough title for multiplayer gaming. Rage is a solid game but the game itself doesn't revolutionize the genre, there are parts that are revolutionary but as a whole doesn't bring anything earth shattering to the table. I'm not saying it needs to but if it just did a better job of immersing the player in the world with a more compelling story it could be forgiven that it didn't hit a home run in the revolutionary area...they have always hit homeruns in technology and gameplay and the Rage technology does have a chance to do this with megatexture but the biggest hurdle IMO will be dealing with the enormous size of tech and the sure army it takes to make the environments so stunning. Imagine Skyrim with it's scale with the graphic fidelity of RAGE. I'm dying for an Id on foot deathwatch because til this day no FPS games have done better in the solid feel and exactness of an I'd shooter.

Yeah, RAGE has ruined me... I haven't got so used to having such amazing graphics that when I play Skyrim, I just can't stop imagining what it would look like on id Tech 5. Of course it would take up 100gB + , but I think I'd make room. :celebration:

I think that later on through the engines life, we will see an ability to use High Res textures without sacrificing framerate, as Carmack was saying, Doom 4 is going to have something like 3 times greater graphics than RAGE (not bein stupid, but I can't imagine it). So give it a couple of years and see what comes to pass then.

Last year video games made more money then all the Hollywood movies combined and with the next generation consoles about to start coming on the market we'll begin to see truly cinematic effects in video games become commonplace. These are merely the first serious steps in this direction with games like Rage moving in the direction of ray casting and L.A. Noire incorporating motion capture. I sometimes compare it to the early days of special effects movies like the original 2001: A Space Odyssey and Planet of the Apes. Audiences kept demanding more and the industry kept slowly developing the technology, but it wasn't until Star Wars came out that most audiences really began to appreciate the full potential of the technology and how much progress had been made. In the meantime we can expect some people to continue to criticize all the flaws in the technology mercilessly. That's the price of fame and fortune, but let us never forget its just a stupid game/movie/video game.

COD brought in a cool billion dollars last year while costing roughly 50 million to develop. That's a Hollywood blockbuster by any measure and the size of the task is obviously not a major issue for Rage which cost a mere 27 million to produce. Carmack is a technical wizard by any account, but what we need now is the next George Lucas to step up to the plate and put all the pieces together. The right story, the right actors, and the right directing as well as the right technologies necessary to push the genre to the next level. Don't worry about size, if anything, start thinking big.

I wrote this with the story / content in mind. Should have compared Doom to Mario64 to make it clearer.
Wolf3D and Doom revolutionized the games industry not by their technology alone but also by their radical mature content.

And there's exactly my point when talking about raytracing: You bring up a new, revolutionary technology and the casual gamer doesn't even notice it but complains about crappy textures.
Rage is the only game i know where the landscape has this homogeneous "carpet" that looks like in real life without the usual unnatural edges that clearly shows everything was made out of single objects to make a whole. It completely lacks the usual polygon clipping sections that distort the visuals with unnecessary edges. And it has very few repetition in texture. But most people just won't notice.



I guess life for game programmers was easier in the 90s when games were way simpler in detail and any form of enhancement was immediately noticable even for the most stupid nerd. Especially when considering the fact that all possible target platforms had a distinctive rendering look. You could tell on which hardware a game was running by just looking at it. PC / N64 / Playstation and Sega Saturn all had their special abilities. Today, all available systems (except for the mobile section) use chips by one of the two biggest gpu companies and the results are extremely close in visual appearance.

It's not just casual gamers though, is it? You see even people who really should know what they're talking about completely missing the point about what they're seeing here. I could theorize all day about possible reasons for that (in fact I just deleted 2 paragraphs of such theorizing), but in the end I don't think it would achieve much.

One thing I will say is that I believe that current rendering paradigms are right at the limit of what can be done, and have been for quite some time. You can ramp up the detail, add more bump/specular/gloss/dirt/glow/etc maps, but ultimately all you're really doing is blending a sequence of textures together using standard math operations. There's no innovation in that, it's building bigger, not better, and all you're doing is yet another variation on the same old "blended layers of tiled textures repeated around some geometry"; something that the original Quake did back in 1996. That's something I didn't realise until I saw Rage, and Rage, for all it's other flaws, is certainly demonstrating that there is another way forward towards better realism and immersion, and no matter what your opinion of the gameplay, stability or texture detail, it must be given huge credit for that.

Very old engines could have run RAGE. It's just large areas and vehicles mostly. RPGs have had excessive text in speech and menus since before full 3d (See: Daggerfall)
What are you talking about here other than newer graphics? "The game" is its overall structure, not just graphics. If you want to be technical, Daggerfall is a bigger game than RAGE. Does that mean that Daggerfall's engine is better?
I believe you're talking about the better graphics not game structure.

Its not just another way, but the only way anyone's come up with yet to go forward. The advantage of sparce voxel octrees is that they significantly reduce the demands of ray casting to just what gets displayed on the screen. There isn't even a theoretical alternative that I know of. The situation is similar to when the original Crysis came out. At the time many reviewers commented that it looked great, but almost nobody's computer could play the game. Within 3 years gaming PCs that could max it out were affordable and now there is even news a version will be coming out on consoles. Rage is following that same path taking a dive into the deep end of the pool with a radically advanced engine not knowing exactly how everything will work out in the next few years, but knowing this is definitely the future. Gamers might be frustrated in the short run, but you can bet developers are following it with keen interest.

Yes, I am specifically referring to graphics hardware here. The gameplay, for all that I found it to be hugely enjoyable, is not and - as it's an id Software game - should never have been expected to be revolutionary. id games are always simple, basic, fast-paced, arcadey, kill-anything-that-moves/stay-alive-and-get-out-of-wherever-you-are joyrides; you don't go to an id game expecting deep immersive gameplay, you go to one expecting a fairly non-taxing blast of old-fashioned fun that's not going to try your brain cells but will test your reflexes and/or nerve. Starting out from any other point invalidates any conclusion that may be drawn.

Jimmy - you've written the most intelligent cogent anolysis of RAGE I've read -& possible reasons for so much of the negativity toward it.
Thanks.

As a long time fps PC gamer, I'm utterly baffled by the scale of the negative reaction toward RAGE.

I genuinely feel this title has demonstrated a bold, visionary new way to create video games. And a stunning one at that.
I'm blown away by it & feel dissatisfied with everything I've played since.

id must feel like Galileo, scratching his head, still surrounded by flat earthers!

Great thread.

I see the importance of what Megatexure is trying to achieve. It's a big step in the right direction. A world without repeated textures is a natural progression. The trouble is, many gamers will expect shadows, interactive, physics-based environments and sharp textures up-close because that is what they have got in the last few years. Of course, for id Tech 5 to run at 60FPS on consoles doing what it does, sacrifices had to be made. Even the texture detail is constrained by hardware and is more of a distribution and storage issue than anything. Perhaps if id release an uncompressed megatexture demo of some portion of the game, the naysayers will see.

I wouldn't say that most gamers don't notice id Tech 5's beauty. Perhaps some don't realise it runs at 60FPS and has unique texturing, but they may subconciously know or feel it. I saw Skyrim running the other day at a decent graphical fidelity and my immediate thought was 'the scenery looks a bit bland'. I started to notice - although pretty - it looked like a game, and every other game. To my eyes Rage looks different and not like any other game.

I also think id Tech 5 is a step forwards development-wise. Any technology that frees up artists from technical constraints, allowing them to create content more freely and quickly is also a welcome (and needed) change.

Still, I'm a bit of a techie with a programming background so I'm looking at it from a technology / development point-of-view.

Perhaps Doom 4 will be the title to convince people?

I'm not a guy that gives anything about critics so i might be badly informed there. What i'd call "casual" gamers are the people that buy and play games without knowing what's going on before there eyes. I don't use this expression to differ between gamers that buy many or few games. In the early 90s when games where highly innovative in terms of content and techology it was the "tech expert" that was buying them. Dos was common and most people where able to edit config.sys and autoexec.bat themselves and they were clearly able to judge what they were seeing on screen.
This changed completely over the years. Today everybody is playing games including the newspaper seller, the highschool kid or the pop star and all of them have an opinion even if they're not able to judge over the technology of a game like someone can who knows what's under the hood...
But i'm very pleased to see that this forum has some technology oriented members.
I posted lots of programming related stuff in the Gearbox forums after Duke Nukem Forever's launch to show up the difficulties related to programming games but i had a pretty hard time doing so.



Perfect anolysis. This is the typical result of a product going completely into mass market. People want what they are used to and always more and never less at the same price. Anything that is different will be neglected since it doesn't deliver what they expect. That is exactly the problem Rage is facing. It delivers perfectly in it's new technology except for one point: the (depending on the system) slow texture loading and the somewhat low texture resolution in general. Gamers are used to seeing higher texture resolutions in other games and the negative factors are alyways noticed first. Humans are very bad at keeping their judgement of positives and negatives in balance.

Yes!
The more i think about todays stagnant situation in terms of graphics technology, the more i consider voxels to be the best alternative. But this is just the technological and not the business view. I expect it to be almost a "mission impossible" to introduce this technology into the mass market as it is quite different to polygons and changing the complete industry from one technology to the next is going to be a tough task.
The only way to do that would be to make the transfer as slow as possible and it would need the help of the gpu manufacterers by implementing both polygon AND voxel support into their chips.
Releasing a hardware that was voxels only would definitely fail due to lack of supporting software and software only solutions lack the acceleration that gpus provide for polygons so it would look worse than modern high poly engines at first which is a no go for the mass market.

There is some research being done on creating circuitry specifically designed for processing voxels, but its very basic research at this point and I don't expect any real results anytime soon. Not that it really matters anyway. Its not about just having a more powerful gpu, but being able to feed it data fast enough. That's what the id tech 5 attempts to do by heavily compressing textures and their future plans are to do the same for geometry, but all that compressed data also has to be decompressed before it can put to use. Its that process of getting the huge amounts of raw data all the way from the hard drive to the gpu that needs to be developed before we can even think of creating a new standard for graphics. Ideally you'd want something like faster memory to feed the cpu and gpu and a processor with 8 or more cpu cores and 80-300 simplified gpu compute cores. That's just a gross estimate, but essentially its the same trend we see already with manufacturers attempting to produce faster memory, squeeze more processors onto a chip, and figure out how best to arrange all of them.

You just summarized one of the paragraphs I'd deleted!

My take on it is that people have come to expect a PC to be not much more than a higher powered console (or a standard domestic appliance, like your DVD player or your refrigerator). Trouble is that all the ugliness under the hood that has been hidden by better UI, a more stable OS, saner ways of configuring things --- that ugliness is in fact still there, and when it pops up - as it inevitably does - people no longer really know how to deal with it. You see this in other software too: Java VMs, Oracle client software, Lotus Notes, most "Enterprise" systems, etc.

I'd recommend people read up on http://en.wikipedia.org/wiki/Leaky_abstraction, because the PC itself is one giant Leaky Abstraction. That won't fix anything, but it will give a very good idea of just what is going on.

Instant gratification all the way baby! One possible solution I've heard of is to have computers automatically share information online about what works and update themselves accordingly. If one person with the same computer as you finds a solution it can be shared with thousands of other people automatically and would make a lot of troubleshooting bulletin boards obsolete. There's even some interest recently in creating search engines that check the reliability of information. All designed to enhance instant gratification while hiding the complexity of what is going in the background. To quote Louis Carroll, "Curiouser and curiouser".

What is "ugliness under the hood"? Certainly windose is an ugly thing but it's the halfwit GUIs that break most stuff. OS X is just as compromised and one could make a case for running games from the command line just to avoid the GUI.

Still computers are not at all ugly, well like them a lot, it's the software that is the problem. That's why I run Linux for everything except games. Because all that I require windows to do is run games my system in not loaded with cruft and I have few problems. Rage ran quite well right away and so does Skyrim.

The problem with Rage is that it's just not a very good game. We were expecting a masterpiece. There are also problems relating to it's console base but they are not a huge problem although a bit insulting to my rather sweet machine.

Skyrim is a very good game and although there are problems a very large number of people are having fun.

Wow ... from the Leaky link:

"The TCP/IP protocol is the combination of the TCP protocol, which attempts to provide reliable delivery of information, running on top of the IP protocol, which does not ensure reliability. Therefore, on occasion, the IP protocol will fail, and a developer using TCP/IP cannot trust the combination of TCP and IP to address the problem."

No clue at all. IP is Internet Protocol which commonly includes TCP and UDP transactions and just covers the transfer of packets. If you don't care all that much about how reliable that is, and there are many cases where some missing stuff does not matter, streaming media for one, you use UDP for a bit of a speed increase with a little less reliability. If you care about the data and it needs to be in order you use TCP or Transmission Control Protocol to ensure everything transfers properly. It wants verification that you received the packets and that they are in order.

As we trust billions of transactions every day, many are financial transactions done over SSL, we can see the statement is pure bull.