The Sunday Papers

Sundays are for the silent remembrance of the beauty of Giant Anteaters, sipping tea and compiling lists of interesting gaming reading from across the week without revealing the odd 1996 musical flashback that’s been happening in the last few days. So that’s what I did.

Really failed.

57 Comments

  1. KBKarma says:

    I like the TechRadar article. I’m always surprised by how people scream about how Blizzard or whoever is violating their rights as players. Except, while in-game, the only rights a player has are the right to enjoy themselves, and whatever has been set down in the EULA. Lots of rights for these games are rights like in the Miranda Code, such as “you have the right to play this game”, “you do not have the right to hack the game”, that sort of thing.

    Also, curse you Britisher people for having access to Spotify.

  2. jalf says:

    Aw, nothing about Valve banning Kim Swift from the game designy challenge at GDC cos the theme was related to sex? :D

  3. jamscones says:

    “existentialist soul-pop”

    ROMO

  4. Markoff Chaney says:

    Skinny Puppy did put on quite the show in 92 when I got to see them my only time. I never did make the Pigface show the year before either. :(

    Steenberg’s thoughts mirror some I’ve had as well. Something has to give somewhere. Can my eyes really tell the difference between 102349852×2340589 and 10234985×234058 resolutions? I’m sure to some degree, to some point, to some pixel or artifact I can point to the answer is yes. To another consideration, mainly the law of diminishing returns the answer becomes less clear. I think that’s why I’ve changed my hardware adoption schedules regarding PC gaming. My 8800 GTS still pushes out everything I need at the resolution I play. That’s good enough, to me, for anything I’ve really wanted to play, pretty much ever. I’m sure something new will come out (and probably be poorly coded as well, requiring more overhead) but hopefully this box will last the 4-5 my last one did. Bah. Too many tangents. I need my BoJangles.

  5. mandrill says:

    Mr Steenberg is a prophet of our age, a prodigy whose infulence on video game/interactive art production will be being felt for decades to come. Looking forwad to Love very much.

  6. Kieron Gillen says:

    Jamscones: Man!

    KG

  7. Evangel says:

    He said “content” not “resolution”. Dynamically generated textures/models are probably the way to go. During development only, since it’d take way too much processing power to do it client side.

  8. Ginger Yellow says:

    “Takeshi Kitano, perhaps best known in mainstream UK for his riotous imported gameshow, Takeshi’s Castle”

    I do hope this isn’t the case. The guy’s a film-making genius.

  9. Premium User Badge

    Cooper says:

    @ Evangel
    On the subject of procedural texturation:

    link to theprodukkt.com

    An fps game at 96KB

    Done, as far as I can tell, by generating the textures client side (albeit during loading rather than on-the-fly)

    See also:
    link to theprodukkt.com

  10. Quirk says:

    For some reason Eskil Steenberg reminds me a bit of Peter Molyneux.

    His initial point, incidentally, is nonsense. Graphics cards, which are the places where all the important stuff happens, aren’t really behaving as per Moore’s Law from the viewpoint of the average consumer. More powerful cards are becoming available, but their price is correspondingly greater, and depreciation of what was relatively high-range a year or so ago is fairly minimal. I discovered this recently, when trying to replace a broken heatsink fan on my graphics card; a new card of the same capability would have set me back a roughly identical amount to the price over a year before when I got it.

    Let’s suppose this were not the case, though. Resolution, and even polygon count, aren’t actually huge issues. I’m a professional graphics coder myself, and our modeller generally makes his first version to a higher texture budget and polygon richness than we can easily run. This is the low-hanging fruit; modellers will find some of their budgetary constraints relieved. Round objects look rounder and smoother, and you get rid of polygon jaggies – and none of this costs the modeller anything in time.

    Where it does get tricky is adding extra content. It’s all very well to model a room with a chair and some tables, but having the kind of clutter that shows it’s lived in takes time to produce. As we get more and more sophisticated in what we can display, we can show more and more clutter, more complex environments, and if we want to do that, then, yes, we need more modellers. However, his proposed solution does nothing to affect this – you cannot expect to procedurally generate human artifacts and reproduce the variety of them that exist. The best you can hope for is rearranging them in some procedurally generated way, which makes a gain only if you have a large number of rooms that have similar contents.

    Procedural generation comes into its own more when it comes to trees, and terrain. Ironically, having begun with 3D engines showing small interior spaces because of the limitations of the code and processors, we find ourselves overcome by the sheer density of information required to model everything about these interiors accurately, and our 3D engines must go outside to continue to show their improvement, where we can create something that looks somewhat naturalistic. This is something that is very much being done already, though, and Eskil is hardly a prophet for mentioning that.

    What he’s trying to do isn’t really suitable for the big studios of the moment to take into the future – if you decide to just ignore texture mapping as he has, your game will at best look very stylised, and at worst it’ll look godawful. It may prove more relevant for small indie studios trying to compete with the big boys without the resources to hire people to do detailed work; we’ll have to wait and see.

  11. Robin says:

    Steenberg’s wrong, for fairly obvious reasons. Tools have become more powerful in tandem with hardware. It doesn’t take several million times as many man hours to sculpt something in zbrush as it did to draw it in dpaint. Artists aren’t required to handpaint every inch of their square-kilometer megatextured terrains.

    And that’s aside from the fact that you *can* economically use bigger teams if you’re outsourcing to China or Russia. Which is why EA, Ubi and Codemasters routinely do this.

  12. Gorgeras says:

    Well we’ll just have to forgive Steenberg for not thinking like a corporate douche.

  13. Robin says:

    That’s one way of looking at it. I think that the idea of talented people limiting the scope of what they can create (and who can access it) by refusing to engage with “The System” is cynical and self-defeating.

    Maybe Steenberg is fortunate enough to have the talent and resources to realise his vision on his own (like Chris Sawyer). Good for him. It’s not the One True Way to make good games though.

  14. Turin Turambar says:

    Tools are becoming more powerful, but not powerful enough. Point it: both game companies and game budgets increased steadly in the last 20 years. From 2-men teams to 20 men teams, to 200 men. From 40.000$ to 40.000.0000$.

  15. thomp says:

    he kind of does seem to spend a lot of time thinking like a corporate douche, though. or at least like scott adams. which is worse.

  16. StalinsGhost says:

    Interesting. The new Manics track isn’t bad – I’d considered ignoring the new album after the last rather damp effort. But I might reconsider now :P

  17. Mil says:

    Tools are becoming more powerful, but not powerful enough. Point it: both game companies and game budgets increased steadly in the last 20 years. From 2-men teams to 20 men teams, to 200 men. From 40.000$ to 40.000.0000$.

    I think that the fact that larger teams are now economically viable just shows that the market is now bigger, not that the tools haven’t kept up.

  18. pepper says:

    Turin, how arent they becoming more powerfull? With the intregration of Zbrush like tools into modellers like 3dstudiomax/maya, the CS pipeline(Photoshop, flash etc) and the fact that these programs all recognize eachtoher. how is that not keeping up? Especially when looking at the future, the UT3 package system, and the content management from the level editor itself which also has a advanced particle and scripting system. Look at the new unity for windows.

    20 years ago, we could only have dreamed of those tools.

  19. Ian says:

    On of my webchums made a PC in a stormtrooper helmet.

  20. PleasingFungus says:

    Eskil Steenberg had some interesting ideas, and an awful lot of angry shouting. (“Get rid of the designer!” he cries. Hm.) More’s the pity. I’m still very excited about Love, though, which I hope will have at least another trailer sometime this year. It’d be nice.

    (Did you spot his preceding blog post? It seemed… remarkably off-base.)

  21. Turin Turambar says:

    “Turin, how arent they becoming more powerfull?”

    I said tools were becoming more powerful! But not powerful enough. Why? Because what people expect from a retail game has grown faster than the tools.

  22. James O says:

    pepper: Tools are becoming more powerful, but not at a proportional rate to the expansion of rendering power. Photoshop CS 3 is obviously much more capable than Photoshop 1, but hand-painting a quality 2048×2048 texture map is going to take substantially longer than a 256×256 map all the same. Add in new market-driven requirements, like having to also paint a quality specular map and normal map, and the amount of content that needs generation is just exploding at an unsustainable pace.

    Better tools aren’t really a solution here, because better tools just raise the standards of the end-user with regard to asset quality. ZBrush is a fantastic tool for making detailed organic models, and it’s hard to imagine how our pipelines got by without it in the past. However, ZBrush has made generating quality normal maps so much easier than it once was that it’s now become part of a baseline expectation that all AAA games must have this level of detail. Thus, all meshes need to go through another pass now, a ZBrush pass, before they can be considered shippable. That means more time, money, and warm bodies.

  23. Psychopomp says:

    I want Takeshis Challenge

  24. pepper says:

    james, Turin. You guys have a point there. I suppose its a infinite loop that you will wind up eventually. Just as we nowadays seems to need quite a machine for just browsing the internet. Even word processing seems to suffer from this.

    Suppose we cannot easily speed up the content pipeline since that will increase the demands anyway, what should we look at then? Procedurally generated content is a solution, but it isnt yet on the level where it can make all the difference. I read a article about this a while ago on gamasutra. Il see if i can i dig it up.

    The coding side of things doesnt get a lot easier neither, just take a look at the programming required for the PS3 processor, and you instantly realize why those machines are being used in render farms. Same go’s for the X-box but that is more streamlined with the PC.

    Education? I doubt it. Only the develop concepts stay the same, the tools change.

    All I can imagine now is increasing the team size, which brings along bigger risks and investments, and thus could slow down progress, causing team downsizing… You can see the loop were in? if we want to break this then i believe we need to change our ways in which we develop the games. Although i do not see such a thing in the foreseeable future.

    Just my 2 cents from a student gamedesign

  25. Kadayi says:

    I’m looking forward to LOVE, but reading a few of Mr Steenbergs blogs he seems to come across as a bit too much of a dreamer/idealist for his own good.

    I think with regard to increasing game complexity, I don’t think it’s unlikely that we might not see an eventual consolidation of game engine technology somewhere down the line. Right now there are a few players, all of whom have their different strengths, but ultimately it would be quite advantageous from an industry perspective for there to be one game engine everyone uses as their ‘format’.

  26. Kevin E says:

    That Chewing Pixels site has some fine articles. How did you find the author names though, I searched everywhere and didn’t see anything. Without the names, articles tend to give off a faint impression of anonymous spouting.

  27. Kieron Gillen says:

    Kevin E: I know Simon P, so know it’s all by him. I’m surprised it’s nowhere on the site, as he actually links to it from the places he writes with a byline (His GSW columns, for example).

    KG

  28. pepper says:

    Kadayi, that engine would need to have such a huge scalability that the current technology cant coop with that.

  29. James O says:

    Procedural content is one solution and it’s pretty powerful in the right context – from what I can recall, all the mercenaries in Far Cry 2 were randomly assembled at run-time from a library of parts and kit items (bags, pouches, etc,) and 3D packages like Vue and so forth make it clear that procedural methods can generate convincing natural environments.

    However, I think procedural content is just another tool – sure, it can be used to add a great deal of variation, but you still need to make the content to provide those variations (see the Total War games, with their increasingly heterogeneous armies with multiple faces, uniforms, etc per soldier.) Players will just demand an increased level of variety and modular pieces, so you’re kind of back to square one.

    Further, though it’s a great force-multiplier, but it’s only narrowly applicable – it only works on content that can be has easily definable axes of variation. In other words, we have to be able to pre-define certain ranges of variation (i.e. this character should use one of {backpack} kit parts on his back, one of {faces} for his face texture; this grass should be {low…high} inches tall.) Some content cannot be parameterized in this way or is far more difficult to define parametrically – how would you write a procedural function to populate an office building with furniture and gadgets in a consistent and coherent way while still being interesting to look at (i.e. no obvious tiling) ? Procedural solutions are great for simulating natural processes, but not so much for intentionally-designed man-made spaces/objects.

    One big issue in content creation today is the immense amount of re-inventing the wheel. I’m an artist at Turn10 (Forza Motorsports,) and since we specialize in real-world race venues and cars, the duplication of effort of other studios is quite clear. Is it really necessary for us to re-build some of this content from scratch when other companies have already done so? In film, you can just shoot on location or go to the store and buy some of the basic every-day props, like chairs and such. Sure, the film crew may add unique props or make modifications, but basically most of the underlying content is the same. Imagine how much more expensive film making might be if every time a film was set in New York City, the entire city had be to re-built for the film through the use and construction of props and facades and such? That’s basically what games do (how many different Ferrari Enzos do we need for racing games? how many more AK-47s and MP5s are needed for FPS games?)

    I can’t help but imagine that at some point in the future, a significant deal of content will be just be bought off-the-shelves of some content provider company, with the game studio’s in-house artists making appropriate changes to suit the game. To me this is the logical endpoint of outsourcing asset creation – eventually we’ll reach a point where assets are near a photo-realistic quality and thus there’s no need to remake them, so we’ll just buy them as-is instead of paying someone to re-make them from scratch. Obviously, like with procedural content, this wouldn’t be an end-all solution, but it could affect a substantial chunk of games. We already do it for programming (through middleware and licensed game engines,) so I think it’s basically inevitable for art as well.

  30. Okami says:

    Mr. Steenberg should confine himself to creating games, this talk has seriously undermined the respect I have for him.

  31. Gap Gen says:

    Is there a case for running an online world as a cooperative or democracy, rather than a dictatorship-by-developer? Would the game go horribly wrong if the players had votes on in-game decisions, or would they be enlightened and knowledgeable enough to vote for the global good of the game?

  32. Al Ewing says:

    Hooray! Finally get to listen to ‘Something To Write Home About’ again.

  33. Kadayi says:

    @pepper

    I think James O is on the money. I’m not envisaging a universal engine that covers every sort of game, but it wouldn’t be much of a stretch to see one emerge that happily covers the 1st 3rd person single/multi-player sphere of FPS/RPG/Adventure.

    Once you’ve got a standard in place then it makes it so much easier for 3rd parties to get involved. For example presently pretty much everyones doing their own motion capture, which is a hideously expensive process and quite time consuming. With a universal engine in play, it would be quite possible for companies that specialise in that to produce whole catalogues of standard animations (walking, running, climbing, falling over etc) for games that they could sell directly to development houses, therefore freeing those developers up to concentrate solely upon the requisite unique animations their game requires rather than having to reinvent the wheel again and again (as James O points out).

  34. Dinger says:

    Steenberg’s argument against designers reminds me of engineers’ arguments against architects. It all sounds good until you go into a building designed by engineers.

    If you take it charitably, then his point is banal. An unprofessional designer is one who doesn’t take into consideration the costs and limits to implementation; the same is true of architects. But implementation skills (programming, art assets) are not architectonic.

    If you take it the worst way, then it’s kinda like the article linked last week on AI in games. The way to smarter-seeming AI is largely through shaping the encounters with AI to put them in the smartest position (or rather, put the player in a position where only dumb actions are available). In a sense, that’s the cart driving the horse. The availability of resources will always affect the experience you want to communicate (would Film Noir exist if B-movie directors had access to Paramount lighting?), but if you come in saying, “here’s what the hardware does well; here’s what our budget will do well with that hardare; here’s the game”, the result will be another instance of the cookie-cutter shovelware that has permeated this industry for 25 years.

  35. Explosive Coot says:

    @ Gap Gen

    You do have a vote – if you dislike what the developer is doing, cancel and encourage all your like-minded friends to do the same. Most MMOs have a feedback form when you cancel, so there’s the perfect soapbox to let them know why you’re displeased with their service.

  36. James O says:

    Gap Gen: CCP, the EVE Online developers, had a talk at the GDC on just that. I didn’t go myself, but Sirlin has a nice write-up on his site.

    Kadayi: You’re basically seeing that process now, merely subdivided – there’s no singular catch-all engine (that would be too hard to engineer,) but there’s lots of specialists in particular areas – Havok for physics, fmod for audio, Games for Windows Live for online play, and so on. I see it a bit like the automotive industry – the platform, chassis, and engine are done in-house by the car company (as they are what make the car distinct and marketable,) but the brakes are by Brembo, the sound system is by Bose, the tires are sourced from Dunlop, etc etc.

  37. Gap Gen says:

    Explosive Coot: Not quite what I meant. As an analogy, it’s the difference between being told “if you don’t like our dictatorship, emigrate” and living in Switzerland.

  38. Torgen says:

    @Gap Gen: Would the game go horribly wrong if the players had votes on in-game decisions, or would they be enlightened and knowledgeable enough to vote for the global good of the game?

    Ask Karl Marx how trusting a population to make the non-selfish, altruistic choice worked out.

  39. me, ehem. says:

    @James O: I’m rather surprised that content creation houses haven’t been established yet. It does seem to be a rather glaring demand.

  40. Muzman says:

    As far as I’m aware there are content creation houses who do nothing but create licensed texture sets, models (including buildings), animation rigs, motion capture data etc.
    Mostly this is for the film industry, however, for the reasons of platform variety mentioned above I would guess.

  41. Saul says:

    Steenberg is very clever and right about a lot of things, but yes– getting rid of designers is not the answer! Better educated designers is. Letting programmers be designers leads to awful things like MS Windows.

  42. James O says:

    me, ehem. : There are plenty of content creation firms established – Forza Motorsports 2 used at least three – Dhruva (India,) Glass Egg (Vietnam), and Valkyrie (Seattle.) Liquid and Massive Black are US-based companies that do outsourced art for lots of games (I think Massive Black worked on Hellgate: London and Fallout 3, to name a few.)

    However, what I’m envisioning is instead of hiring a firm to make art, simply buying art that is already made. While there will always be demand for custom built assets, there’s surely a great many that really don’t need to be re-built by loads of different companies every year. We already have basically this in Turbosquid, but those assets are mainly built for offline (non-realtime) renderers. The increasing realtime performance of modern game systems is erasing that difference each generation, however, so I imagine we’ll see some sort of professional studio oriented Turbosquid selling licenses to entire libraries of assets in the not-too-distant future.

  43. Heliocentric says:

    The some designers must be programmers, some programmers must be designers. At least a little bit or you get wasted energy. Look oblivion, its environs were not only populated by hand but by generation. These 2 things work well together. I hope love can tap into the player base’s desire to build because otherwise the most exciting aspect of the game would never bare fruit. Other games have cast design to the player before, trackmania for example could be criticised for limited cars but ingame i can go to an online portal and click on a model and have it. All talk of quake live has caused complaints that it doesn’t have the best maps, because they were player made. But ideally love will no programming or design ability to create things. Is that terrible or good?

  44. Gap Gen says:

    Torgen: Again, that’s a little unfair – the man “communist” countries weren’t really democracies, let alone total democracies. Most communist countries were basically dictatorial command economies.

  45. pepper says:

    Helio. Yup, i believe everyone should have some knowdledge what a progammer does, a 3d designer does etc. It makes the process of designing a lot easier.

    James. I personally prefer making environments for games, nature like/outdoors etc. I can tell you why the content needs to be remade for every game, because they all have there own style. Although this is becoming less and less so, because nowdays a lot of graphic data is defined in shaders. One used to see a clear graphics style between games in a genre. Nowadays not so much.

    If we would all borrow content from eachother then games would become very much more of the same. I doubt this is a good thing.

    One thing we do see happening is studios springing to life that ONLY make models/textures for other companies. You can outsource your work to those companies and then get a fully textured, animated model back. I know serval big companies already use this to lighten the workload.

    In the end, we can only accept that this situation isnt going to change over night, and especially not anytime soon. What i can imagine is new peripherals that speed up the process of designing, debugging, and the crunch time. Although the implementation of these new techniques might take a while.

  46. Tei says:

    Only the programmers can change the rules of the game. The designers play withing the rules… and are better players.

    So you want programmers, to maximize the size of the rules, so cool things are doable. And you want designers to fill the world with teh itneresting.

    No programmers, and you haev teh linear. No designers if you have boxrooms.

  47. pepper says:

    And in turn the programmers are limited by the hardware. So if you want to maximize all ends then you want to have the best possible code for the hardware available.

  48. Skwizzal says:

    Not quite from the sunday post, but some more interesting gaming related news from this weekend:
    link to news.bbc.co.uk

    So apparently as gamers we can make safer drivers. I hope they never see me playing Burnout :P

  49. Mihai says:

    That LOVE guy is brilliant.

    Also, any chance anybody recorded Barnett’s “performance” at GDC?

  50. pepper says:

    More about the eye thingy:

    link to livescience.com