Elysian Shadows

Are Game Developers 15 Years Behind The Rest of Us?

What do you think?

good article, but i admit i only read the first half of it.

i think you have to take more of a case by case basis approach to this argument however.

if anything, i think that game developers are responsible for pushing boundaries of existing technology. without games (or maybe medical imaging, engineering simulation) we reached our maximum many years ago in terms of needs and tech requirements.

also, i think while indie developers have a stigma of being unwashed bearded tie-die tshirt wearers, this may only be a response to the elders of this community who are still wearing what they did when they made these games 2+ years ago. I think the new breed of programmers who grew up with games from a young age will change this view.

Uh oh. I think I've made a few Angry Nerds on reddit

k1net1k wrote: this may only be a response to the elders of this community who are still wearing what they did when they made these games 20+ years ago.

Fixed.

i meant 20 lol

I don't know where he get's his ideas to tell the truth. I mean, does he think that games have no source control, or solid backup procedure? - how on earth does he think game companies protect their investment, debug... hell I'd bet that most game developers have more intricate test, debug, and backup systems than he's ever used. I think he makes assumptions that he does not have the knowledge to back up. See, source control is a given, there is no game project out there with more than 1 programmer that doesn't have some sort of source control system. Back when I was in college, business programming was a very different beast - programmers would tend to take on areas and deal with them solely, which is a nightmare when a programmer leaves, but with platforms like Pascal and Cobol, well I guess it made sense back then to compartmentalise. It's easy to look around this millenium and say, wow, we are pretty freakin clever, I can rely on this workflow and all my changes are logged, everythings backed up, all our roles are defined. But it's also easy to forget how business software development got to that stage and assume that it got there all on it's own.

Videogame programming is very different to business programming, they are practically at polar opposites in terms of approach. So much of business software development is tied up in rigmarole and red-tape, I wonder how many developers wish they could work more organically, take ownership of an aspect (at least) of the project. No, they have to work to the book, keep their stakeholders, managers, users, and pretty much everybody happy. That affects the creative flow, and I personally could not work in an environment like that. Imagine every little thing you do, you have to justify it, get it approved, impliment it, debug it... Myself, I'd rather just fix the damn problem and get back to my blowjobs for bigger monitors scheme. I'm a sole developer in a semiconductor fab, and have full control over about 20 systems. Messy, uncommented code, a nightmare for any other developer to use, and I wouldn't have it any other way. If there's a problem with my code, it's my problem to fix, and my responsibility to make sure it doesn't impact on other systems. There's no software on earth that would make things any easier to organise - a problem is judged on it's impact on production or our ability to make money, that's how my priorities are managed. If there was more than just me writing code, I'd go old school, and have each developer take responsibility for a handful of systems each. Source control is only applicable to out of control source - it's certainly not a necessity in most IT departments IMO.

His view of game programmers is a little worrying though, in fact I'm getting sick of that same old bullshit, that we don't wash, can't communicate, are socially awkward blah bloody blah. The most socially awkward people I know are all engineers, nothing to do with computers. And I often program late at the weekends, it's just nice to work in complete quiet, or at least work in only your own noise, when the internet is faster, TV is often better, coffee tastes better, waffles taste a hella lot better - then when you do sleep, you sleep the sleep of the justified and righteus. That's upto the developer to decide when he is most productive, if that's at 3am with youtube and music and all that, then so be it - at least your being productive if not socially healthy.

I think he's a bit of a snob, I think he knows far less about game developers than he thinks he does. He needs to get over it - whatever it is, my guess is that he wen't one way during his education, and his friends all went the other (game developer) way. He probably thinks that game development is somehow easy compared to business programming. Frankly, he doesn't write games, he has a tainted view of game developers, he is a snob, he thinks his opinion counts... I'm sick of him already. I program business systems for a job, but I write games for fun - I just can't take anyone seriously who doesn't see the need to balance things out. Going home everynight and achieving nothing is a lot sadder than spending that time creating something - whether it's a game, some artwork, music... computers are for creating, not for monitoring and controlling you.

Pffft! - its a good job nobody will read this :D, I hit more tangents than Lady Gaga's gynaecologist.

Unit testing is for the most part useless. Can't tell you how many times you change a bit of code only to make all your unit testes useless.

avansc wrote:Unit testing is for the most part useless. Can't tell you how many times you change a bit of code only to make all your unit testes useless.

If changing a tiny bit of your code makes all your tests break, you're doing it wrong. It means you have a huge amount of redundancy in your tests. Why are your tests executing the same code as each other?

Most of the time I've seen this problem it indicates a problem in the design of the code. It hasn't been properly factored and moduarised.

christo wrote:

avansc wrote:Unit testing is for the most part useless. Can't tell you how many times you change a bit of code only to make all your unit testes useless.

If changing a tiny bit of your code makes all your tests break, you're doing it wrong. It means you have a huge amount of redundancy in your tests. Why are your tests executing the same code as each other?

Most of the time I've seen this problem it indicates a problem in the design of the code. It hasn't been properly factored and moduarised.

That, or the fact that requirements are hardly a static thing. I agree that if your design is "bad" then its irrelevant, but a lot of the time your design has to change for whatever valid reason. unit testing is great if everything is set is stone, thats hardly ever the case.

I think you get waaaay more bang for your buck with something like black box. you are testing the application, not parts.

I might have over or is it under exaggerated the "tiny bit" part.

Black box testing can be done on any scale.

But when you've got changing requirements, unit testing is even more useful. When you test units and some subset of requirements change, you don't need to change any of the tests that verify behaviour in the unchanged parts. Quite often you can identify the change as a change in configuration or integration of the parts and the units still function the same.

Unit tests enable safe changes to the code while having confidence that the changes you have made have not produced unintended side effects.

Okay. Maybe I'm doing it wrong.

But I suggest you go read some articles, it's not like its just me chiming this idea.

I have to agree with avansc here. It happens all the time in work. We write tests for the given requirements, we start coding, the tests pass, the requirments change, the code changes, the tests fail. Tests are good to test that functionality that shouldn't change doesn't change, but all the tests relating to the changed requirements break. And as requirements change over time you eventually have tests that don't even closely resemble the originals. This is a HUGE resource drain for little return when most bugs are found in black box testing anyway.

Chris wrote:I'm no expert in game development but..

*Closes tab*

I'm sorry, but I couldn't even read all of that guy's bullshit. That was terrible.

So software developers are "15 years behind," just because they don't practice agile/waterfall development methods and/or use unit testing? Really? That's idiocy.

In my engineering experience, game developers are WAY ahead of the curve. The constant advances in game development drive graphics algorithms, GPUs, CPUs (SSE, anybody), data structures, and so on. Commercial game development is cutting edge as FUCK. Have you ever seen a government/private contractor doing simulation work? THAT is 15 years behind the times. It usually looks like an N64 game. No, game development is way ahead of the curve.

And this new, miraculous thing in the engineering world? The GPGPU? Yeah, you owe that entirely to the game development scene. Your super number cruncher processor was created and developed for game developers who needed to push even more polygons through the pipeline and needed even more control via vertex/pixel shaders. Once again, game development has pushed the ENGINEERING world forward.

If you want to neglect all of this and base engineering efficacy on the method of testing, then feel free, but you lose all credibility from me. Even engineering companies around my area hate that shit, think it's a waste of time, and essentially half-ass it at the last minute to push the product through. This guy sounds like a jealous, management-minded zealot with something against "elitist" game developers. If he wasn't so insistent upon the necessity and superiority of such testing techniques (to be so arrogant as to say NOT doing them puts you 15 years in the past), he would have had a valid and almost interesting argument. Instead it was an extremely biased, over-reaching, generalizing argument that really blew his views way out of perspective and made him look extremely ignorant.

If technology is dictated by the latest software testing trends, fuck this business.

Game developers aren't behind in software/hardware technology. I don't think anyone is saying they are.