**************************************************
For the two people who actually read stuff around here:
Developing for a console makes perfect business sense. Developing for the PC market is very scary, by comparison:
1. With a console, you know your hardware, period. You know, as a developer, what it can and cannot do. With PCs, it's a complete guessing game- Crysis's developers got it wrong, just like they got it wrong before. Their previous game was well-nigh unplayable on typical rigs when it was delivered, or did everybody forget that? I blame the developers for that mistake- the guys working on the CryEngine didn't spend nearly enough time optimizing for slower machines, and gee, the market responded. Bioshock's PC version, on the other hand, plays nicely on my aging rig, albeit with settings turned lower than I'd prefer. Was developed for a console first. Go figure.
2. Consoles are not ruled by the whim of a giant monopoly which doesn't seem to care about video game sales.
Microsoft shafted the PC market in multiple ways over the last two years (for the third time- remember DirectX 3, or the switch between 5 and 6?).
First they screwed everybody developing in OpenGL, by dropping good support for it from Vista, then they screwed DirectX devs, by making DirectX 10 almost wholly incompatible with an approach designed to support XP users as well as Vista users, from everything I've read. Both choices vastly altered the playing field, and not in a way that benefited anybody in the PC development world. I really, really wonder who Microsoft thought was benefitting- maybe their own game publishing arm?
Hmm... but they don't seem to be pushing out many games for PCs these days. I swear, Microsoft's executives must have fallen asleep at the wheel, especially after promising everybody that Vista would be awesome for gaming
3. Budgets for games keep skyrocketing upwards, and consoles are a much safer bet, because you know how much a disc will hold... and you know that you're losing few sales to piracy, so you can do some pretty simple math against projected sales for a given genre, and budget accordingly. Assuming your game doesn't stink on ice, of course.
I think that the budgetary issues are making less and less sense, honestly. I honestly think that the Industry needs to learn how to do things cheaper, and that there needs to be a lot more focus on how to get content made (the biggest single area of cost, other than marketing, usually) at a reasonable cost. A game like Freelancer does not look that much worse than X3, and X3's present successors-in-development don't look that much better than X3.
Of course, so far as I can tell from screens, the successors are all being fairly wise, and are just trying to match X3's visuals, instead of looking a lot better... and probably saving a lot of money. Maybe they're learning. More likely, they're all operating on a shoestring budget, and if their games don't sell, they'll go bankrupt.
I think a lot of that is plain human stupidity. As the requirements for new visuals keep getting harder to reach, we're entering yet-another period where developers are paying more attention to reaching the desired visual result, and a lot less attention to usability by the artists. I've read a lot about the current state-of-the-art production process for a typical game's art nowadays, and meh... give me .S3O. Who wants to have to hack up normal maps, spec maps, shader maps, glow maps, etc., oftentimes in multiple documents, with some special command-line tool used at the end, where you just hope that it looks good?
Moreover, we're in a monopoly situation, right now- if you want to join this workflow, then you must be using Zbrush- no ifs, ands, or buts. Lack of competition is a terrible thing. Zbrush is a cool tool, but it's a very, very expensive tool, and it is designed for professionals who just want to use it, and nothing else- it's a lot more proprietary-feeling than Photoshop ever was. And nobody is competing with them, so it's staying that way.
Modeling at the low end, thanks to players like Blender and Wings, has finally become something rational again, although, having seen that happen once before (in the 3D-renderers-for-2D-sprites game arena) I'm not exactly holding my breath, and hoping this situation continues to improve.
I think that the day when we finally hit "photoreal-enough" and then people start making the process and the tools less painful cannot get here too soon, but it's probably at least a couple of decades out.
I do think that it's incredibly sad that, whereas for sprite-based games I used to be able to literally just paint a guy's animations, or at most rig up a character, animate, render, and clean up and it maybe took a week, now for a first-person shooter, we're talking at least a month, when you factor in all the time rigging, animating, tweaking various textures, making the LODs... and probably longer. Whereas, with Spring, I can go from "idea" to "working" in about 8 hours.
HL 2 didn't have a relatively sparse number of character models because VaLvE wanted it that way- it had to have been mainly an issue of cost. I thought that Oblivion's character engine, which has identical rigging and stock human figures with mesh-tweaked faces was a wonderful concept, but it doesn't look like it's being used a lot elsewhere, and so the costs continue to be a major issue. Yeah, yeah, I know, pro artists are leet and super-fast, but there's a huge difference between their speed runs, doing some random-ass gun model, and doing a fully-articulated character model with a bazillion different issues to deal with.
Static content is of course another problem- there, the issue is mainly that the 3D art market is totally screwy, due to copyright issues and yet more human stupidity. If you want a billion-poly photoreal object, great, it's available at a reasonable price... if you want a low-poly mesh with normalmaps, then you're probably in trouble unless you want to contract it. I would love to offer to change that, by opening a market for, let's call it, "mid-poly" work, but I think it'd fail miserably and I'd rather waste time making stuff for Spring, which I actually play...
In short, because console hardware specs are more limited, content costs can be controlled better. I think this is mainly the fault of the publishers behaving irrationally. See "how to get out", below.
4. Consoles usually arrive with very good developer support.
PCs running Windows? Um, see part 2. MS is "helpful", but yet they may or may not be planning to screw you over in a year. Whereas, with a console, even one owned by MS, they cannot do that, or they'd be abandoned by developers and become Sony
Linux? GTFO! It's Balkanized mayhem... I think it's completely bloody amazing that as many people can get Spring to compile and run as they can, and even play with other folks... but if I had to bet my five-million-dollar budget, no way in hell I'd bother. Maybe after everything else was done. Pretty much like Macs, that way. Sorry, Linux and Mac people. It's not your fault. It's the Linux core people, for encouraging rampant Balkanization, and the Mac people, for not figuring out how to compete with MS. Sucks. I really miss my old Mac, sometimes, but I wasn't about to spend $3500 for a computer again, just to stay reasonably current for a year...
5. Consoles have standardized and simple control schemas. Simple controls? Well, gee, that means we can only make certain choices about how UI works. That's terrible... ooh, it means that these things can be delivered to developers as standard stuff, among other things.
And it keeps costs down, just like the standardized hardware does. "No need for 5 LODs, artists- we know that this console will not handle giant draw distances gracefully at the detail levels we want, so... just do two, for 'fairly close' and 'really close', and we'll have the level designers compensate". Or do like the original Resident Evil, and cheat like mad
I think that the future of PC games that people pay for is incredibly grim, frankly. Piracy is a real issue, not minor, but the rest of the above is also a reason for pessimism.
How to get out?
This is my take on the subject. I've read various gurus' opinions on this topic, but I personally think that most of them are hoping for stuff that doesn't make a lot of sense, such as a thriving indie-game market. So far as I can see, the indie-game market is mainly guys who barely make any money after lots of work, like I did with Silent Dark... not a real solution, imo.
1. PC developers should quit trying to develop for "tomorrow's hardware", they should develop for "midnight tonight". Which means, quit developing with a 3-year horizon, develop a year ahead, tops, with a deadline for Gold of less than two years, this is reasonable.
Don't expect too much, the hardware situation is likely to continue to jump and swerve at random for years, if not decades, to come. Freelancer looked "old" when it arrived, and had it not gone waaaaaaaaay over budget, it would have made money for the publisher (I couldn't find sales figures, I just tried, but I've always heard from the FL modding scene vets that it maybe broke even- but they went through three budgets and a lot of reorganization, and basically rebuilt the game, at least twice... you can do the math).
"Midnight tonight" content looks a lot different from "tomorrow". Game critics will whine and howl, because they're stupid, spoiled brats about graphics. "OMG, it looks like it was released last year", they'll whine.
Players, on the other hand, will be OK, I suspect. Why? Because, these days, "Midnight Tonight" looks fabulous! We're not in the early 1990's any more- being one step behind the bleeding edge isn't a horrible idea. Dawn of War still doesn't look like crap, next to Warhammer: Mark of Chaos, even though it's a much older game engine.
So, why not have a game that plays great on middle-of-the-road hardware? Cheaper, easier, and faster to develop, too...
2. Game publishers should figure out that it's not in their best interests to have everybody writing their own game engines from scratch.
They should grab the people developing engines like OGRE, tell them to keep releasing stuff under an OSS license, then use the resulting engines. Pay them a good salary, have meetings to decide what the engine's specs should be, but quit shelling out money to every house they sign, if they want to go on reinventing the wheel.
Game engines should just be a toolbox, that are well understood, have standard components, and aren't rocket-science. To put this another way... all this time, the game publishers have been like movie studios, except they were not only betting on the creative vision of their people, but also on the type of film they'd invent to depict the show. That's stupid. Consoles are the logical response to that, but surely it could be done with PC games, too. Instead, it looks like the publishers are just vulture-like, feasting when a good OSS engine arrives, or consigning it to indie dev if it's not super-hot. Stupid.
It'd be better to have one engine, with a lot of plug-in modularity. Spring could be that engine, if we could get the rendering / model-format stuff sorted out, I think. In every way but the game code's tie-ins with the renderer, Spring's almost there, in terms of modularity, imo. Get rid of that problem, and hey presto... you can gut the game code, and keep the renderer. Or vice-versa. Or modify both a bit (say, we want to make Freelancer II, let's not bother having maps, write new code for flying around, use LUA for the UI...).
3. Game publishers should create a rational market for art content. This follows naturally from having an OSS standardized game engine sitting around, because content specifications (file formats, etc.) could calm down a lot.
Instead of hiring giant teams of Chinese artists (or wherever is cheap today) they should just figure out a legal formula that will allow them to get what they want and keep it theirs (exclusivity), hire concept people to explain what they need, and bid on a big market. They'd win on price and quality, I think. The way they're doing it right now seems awfully inefficient, when if they just wanted to pick up 50 trees, they could say, "hey, we want 50 trees, here are our concepts, you have one week, GO", and watch 500 trees show up... or just buy from the 450 trees that were made last week...
4. Video game developers should form a union. Seriously! A union, just like the Writer's Union, etc. Make it international (yeah, I know, legal mess, but whatever, it can be done, dammit). Call it the "Digital Creator's Union" or something catchy like that.
It'd force the publishers to start thinking about behaving like rational economic partners in a creative endeavor, instead of rigging the entire game about cost at the developer end. Which is how things have been going, as costs spiral. Instead of trying to limit what they want to do, they're just trying to find who's going to do it cheapest today, and while that makes economic sense in the short run, I think it's been a terrible plan over the longer term.
The best way that the developers could cure this, and wake the publishers up to the mess they have been creating... is to organize. Although many of the publishers does a lot of the work in the Second World, they rely on people, many in academia, in the First World, to provide them with the tools they need. So both Second World and First World folks could benefit.
5. Ask Google to finally get done with its long-rumored OS project... with a few details in place that would benefit them as a whole. Most of the glorious days of PC video games occurred during a period when MS wasn't entirely ruling the Earth.
Maybe they'll start behaving like a rational company again if they see a real threat to their competitive edge- Linux caused them to build 2000 and XP, then Linux collectively blew it by disintegrating into little worlds of customization. I used to think OS X might get Apple back into the game, until I saw just how awful the early versions were, and how little Apple seemed to care about people like me, who did not want the damn OS so full of eye-candy, and so hard to work around. Maybe Google will do a better job, putting out a free (with sponsored links and constant spying, heh) OS that doesn't suck.
6. Come up with a format that's not readily pirateable, for the PC. Oh, wait... this has been pretty much solved, given that most of the market for games consists of people with Internet connections now.
I mean... come ON. DVDs were a stupid idea, because they cannot be changed. Hackers breaking the stuff in Vista was, erm, expected, I would have been disappointed if they hadn't done so promptly.
Steam, on the other hand, works pretty well. When I read about Chris T's piracy whining, I thought, "man, sell on Steam foo". Ok, you cannot protect games that are single-player only very well, with Steam. Meh, that's a tradeoff, maybe single-player games can only be console titles, or niche markets where piracy ain't so bad. OK. But multiplayer games are fine- you wanna play, you have to have paid.
Anyhow, those are my thoughts about all of this. I get tired of reading about these issues, and watching people just chase the same old arguments around...
