 |
|
|
| |
Id Software's Willits: Rage Mods May Not Be Possible
by Christian Nutt
|
|
| |
|
September 17, 2008
|
| |
Longtime id staffer and Rage creative director Tim Willits admits that it might turn out to be impossible to mod the title -- even on the PC platform -- but says the issue's still under consideration at id Software.
As Willits discussed Rage's story during his presentation at Austin GDC, an audience member asked him if the game would be as moddable as the company's prior titles.
Willits replied that id is as open to the idea of having the game modded as the company's always been, noting that his own career was launched due to users' ability to mod id titles -- but confessed that the technical complexities of the game may make it less possible for modders to work with.
Willits also noted that id's John Carmack is also positive on the modding community, but sees this technical issue as an obstacle.
"Unfortunately, Rage is going to be more difficult to mod," said Willits, primarily because of the complexity of the game's vaunted megatexture system, which stores the texture data for levels as one huge texture map that streams in, rather than many smaller textures.
Megatextures require huge amounts of processing power to be baked into their final form for distribution on the game disc; Willits alluded to a large number of computers working for a long time to process them, analogous to a CG render farm.
Willits envisioned modders developing modular chunks of gameplay that can be slotted into the extant Rage world, rather than full mods, as a potential solution. While the game's large central wasteland is a streaming hub world, its levels are instance-based.
The megatexture issue aside, Willits said other aspects of the game will actually be easier to work with than the company's prior titles. "The description languages and the tools are a lot easier to use than the things we've done in the past," he explained.
|
| |
|
|
Where's the problem?
You only do this when you're sure about your design, it comes in the end, like for CGI sequences, and this never stopped making films on their own machine.
So again, I fail to see the problem there.
The real problem, imho, is in the new publishing politic. They may not want to have fans make "disputable" mods, when you see how devs get slashed for what "amateurs" do with engines.
The community coped, and desktop hardware got faster.
It's not the tool usage, the tools look pretty good, it's the sheer computing power needed to crunch down the editable data - where you pretty much have no limitations on the amount of unique textures, and can stamp and blend them however you want - to the final unified, streamable megatexture data that the run time engine uses.
And don't get me started on UTengine multiplatform content baking hassles.
The id "tools" are not built following the spirit of a "suite", a middleware. Making the megatexture doesn't only require the tools to generate the texture (which are good, and the texpainter for Rage looks ace), it requires an extremely detailed and good looking terrain first, which couldn't be produced with any id tool, and I doubt they changed that. Depending on your resources, obtaining such a terrain can require one to three other independent tools. Not to say that there's a lot of fiddling you need to apply to files manually as well. -_-'
As much as I prefer the id engine, the whole affair really feels old gen.
To Jarad: It's most absurd to think that a mere processing time where you simply just have to let your computer calculate data for you would be a barrier to enabling mods.
Back then during Quake, it wasn't so much of a problem, up to the fact that Aftershock was sold in retail. They were quite fine with it, actually advocating for modding and LD.
So unless I read this wrong, the explanation they give is pure bollocks.
The explanation they give isn't as bollocks as you think it is. I like to think of it in terms of transcoding H264. On a normal computer at the highest settings, transcoding a dvd to a mp4 takes about 8 hours. On my quad-core with the latest x264 build and slightly less extreme settings, that gets down to 2 hours. That's one dvd (5gig).
If I try and transcode a blue-ray (50gig) at hd settings, that 2 hours is more like 12-20 hours of 100% processor crunching on 4 cores at 3gig.
From the way Id seem to be describing their mega-texture baking, and the texture detail they are describing, I would be expecting a huge number of very detailed textures, And if it takes Id Software a server processing farm to do it, I think we are talking something like a week or 3 of 100% cpu quad-core processing just for a smaller sized level.
That's not untenable, but it puts a serious crimp in your standard "build-test-revise" cycle when the build time is measured in weeks instead of hours. Being realistic, you'd probably break it up and it'd be more like days than weeks, but it certainly gets in the way.
I agree on your sentiments about id's tools though. They build their tools to suit their needs first, then later expand them to be more useful to the community at large. Epic however have been continuing to use and expand the same toolset over 3 generations, so have a much more fully featured editor suite that's in a ready-to-go state.
However, I think it is worth pointing out that, compared to Quake2/Quake3/Doom, the challanges are much greater. Also, while lighting was a fairly large compute task for a moderate creative input (level lighting), Megatexture is a HUGE compute task for a VERY LARGE amount of content creation. Basically, the barrier for entry has just been raised, and that will naturally have a 'chilling' effect on the amount of mods that are produced, and an increase in iteration time. I think his comments imply that this will still be something that they are interested in supporting, but the fact of the matter is that at this point, the engine (more like the technology in general) is just going to be harder to make content for.
The point that he is making is that each generation is pruning more and more of the low hanging fruit of wanna-be developers. You have to pretty much be a skilled professional to get anything into a modern engine these days; not like in the old days.
What is so special about the next megatextures safe that they are 128,000 by 128,000 pixels large, four time greater than those of ETQW?
Willits even says that the newer tools are easier to use. So basically the most bothersome and involving phase in all this has been made simpler.
If the process could be paused, it would even better.
Are id afraid of the lack of mobilization from people to make mods for their games, therefore really wondering if they should put a halt to the modding thing?
Who's fault, really? The UE is used for many games and even outside of the the game sphere.
Actually, apart from the very first comment, I think all of us understand that.