"Carmack's" point - anyone agree? Discuss at will

http://www.gamesindustry.biz/content_page.php?aid=30487

Doom creator John Carmack believes the DS would make a great proving ground for upstart programmers if only platform holders didn't stifle homebrew development.

I'm not sure I even recognise his point. Feel free to enlighten me.

Can't you just ask him?!

You don't have to hit the metal to code handhelds of course (both the DS and PSP have LUA implementations, the GBA has spiffy Dragon BASIC etc), but at least you can -- and where you can it's probably advantageous to do so, even if only to get a feel for what's actually going on down there.

I'm going to agree with Carmack. I don't bother with homebrew stuff since I have my GP2X, but if I could work on a DS then I wouldn't need to think twice about which I'd prefer.

I know it's not the same but you could work on Dualis :)

I wholeheartedly agree!

The DS is a cool little device, but homebrew needs special hardware. This at least is easier on the PSP where you can use another firmware, but it isnt as fun to play with imo.
It is kinda fun to program for these little buggers though, takes you closer to the hardware (which is a good thing, it teaches you to be conservative ;)

When the words Carmack, id software, Nintendo and Elves reside in the same paragraph, something's VERY WRONG.

"stayne" is closest.

I think you lot have missed his motivation behind his reasoning.

Try again.

What is the motivation of his point?

Money.

This is interesting. The only problem I'd see is that as an emulator, you won't experience bottlenecks that the original hardware will have unless it somehow manages to limit it realistically. Going to mess with that anyway...

You're going to have to spell it out. I don't see anything odd about yet another id IP appearing on Nintendo hardware -- you've intrigued me now :D

Puki:
Carmack is saying that the DS would make a good platform for beinners to learn how to code at a low level. Ie, using assembly, and accessing the hardware ports and other things directly.

Today's kids don't know how to debounce the input from an analog joystick. They don't know how a keyboard buffer works. A lot of programming is done for custom hardware where you don't have a Windows DirectX API to support you.

Also, Carmack's motivation here is for kids to learn to optimize their code better. Carmack's knowledge of how 3D cards work is at a hardware level, and as a result he can optimize stuff like his texture loading whose performance lapses may be a complete mystery to someone who just knows how to call an API to tell the card to stick a certain texture on a certain polygon. Even simple blending of textures requires knowledge of how the blending pipeline is set up if you want to do anything out of the ordinary, like how I draw solid colored sprites with alpha in my sprite system. (Though pixel shaders have removed that particular barrier.)

That's what Carmack's point is.

Not bad.

Try again.

Make you mind up!

He wants there to be more people in the industry with plenty of technical knowledge and no idea how to make a game fun because then ID won't be at such a disadvantage any more.

Exactly. There is no subtext and this is a wind-up :P

I'm going to have to go with that.

I think you're missing the sort of person he is.

He is telling you, if you have the balls to hit the metal and the balls to work it out and the staying power to get something finished, working and great on that little machine then you have what it takes to go all the way.

Thats the whole point, you're gonna be proving you can do it on so many levels that you can't not be employed.

I'm sorry, did he say that before or after mentioning that he had spent the week writing 3D code for Fountainhead? Because I must've missed it. :P

Seriously though, that doesn't make any sense: If you're going to apply the "not built here" mentality to DS development then you wouldn't be arguing for sanctioned homebrew solutions because, by that measure, the 'worthy' would be the ones that needed no help whatsoever.

I think you're right, mind, that being able to cope with all the low level stuff would increase your employability, but id write engines for people to license and I really doubt that he's trying to argue himself out of business.

[EDIT: Honestly, I just took his comments to be a bit of a lament, from a technologist, that not enough people in the field are like-minded and similarly skilled. If you write homebrew games you probably utilise some sort of API and will fit into a games company as a scripter/level designer I guess; the people who get employed to create the underlying technology write compilers for fun, not games. They're two different sets of people.]

My point is simply, those people who put the work in usually do reap the rewards in every level of expertise.

with the advent of the consoles and the death of the 16bits (well just ST/Amiga) bedroom coders have been limited to working soley on the PC or hacked up handhelds and such like, which means the average gamer cant play your homebrew cos it needs similar botching to get it to work and perhaps locking it out of mainstream games.

psone/yaroze was poorly implemented
360XNA is still a closed environment afaik as only xna subscribers can run other xna games ... imagine if no one but blitzers can run blitz games cos there is no compile rocket
gp32 ... cant see game stocking it

they always seem to spout piracy as a "what would hapen" but arnt they pirated still?

im sure that if they made a version that enabled DS or PSP homebrew from flash cards that people would buy it not just for that but they were getting a hand held anyway ... and then the less generic ripoffs could find themselves in sweaty palms worldwide.

What was poor about the Yaroze?

games only worked on yaroze and not normal psx/spone and it was more bloody expencive to boot

never tried it but afaik it wasnt a fully fledged dev kit it was knobbled ... as a taster it seemed ok on paper ... some uni's got some for videogame courses ... i guess they'll be doing XNA at some point too

Well, they only worked on Yaroze because there was no way to burn games to CD. You could get the games on coverdisk, though.

Yeah, it was kinda expensive and the API was a cut-down version. Not really surprising though, as it was nowhere near as expensive as a pro devkit.

It was a great way to get into pro gamedev at the time though, which is what Carmack is banging on about.

I still have my Yaroze. Anyone wanna buy it? :P