Sony may lose grip in next game consoles war: Article

[Laa-Yosh]

I'm going to assume that the coder with those "mad skillz" didn't necessarely put the performance to best use in that case then. I'm also going to assume that the first coder in your example could find other ways to use that double performance into evenmore clever enemies (or sophisticated AI) or simply in more advanced gameplay challenges thanks to more sophisticated enemies.

I'll try it in another way then:
Coder #1 has low-level coding skills, he can write very efficient and fast code. But he hasn't got a clue about how to design AI, complex renderers with unified lighting, shadows, etc., complex simulations for persistent worlds (RPG-like), and so on.
Coder #2 has the above mentioned high-level skills, but he isn't really able to optimize his code to the metal.

See, I consider the high-level stuff to be more next-gen than a reiteration of the previous gen stuff, but with more more more from everything. I don't want an FPS with ten times as many enemies, or a shallow RPG with even more spells, or more phong-shaded polygons in the graphics.

I guess you'll now try to argue that there are many coders out there who are a perfect mix of #1 and #2...

 

[Joe DeFuria]

Sony and Nintendo are not talking hypotheticals. To us they may seem like hypotheticals, but I'm pretty sure Sony and Nintendo will deliver what they are talking about and showing us.

If they are not showing actual in-game footage, they are showing hypotheticals.

From a marketing perspective, I'm not saying one is better than the other...however, I would argue that the closer you are to launch, the less hypothetical, and more "real" you need to be, otherwise you set yourself up for a backlash.

So it only makes perfect sense for MS to be more "down to earth" because they were always closer to launch than the other two.

 

[Bobbler]

I guess you'll now try to argue that there are many coders out there who are a perfect mix of #1 and #2...

I imagine anyone creating a game will have more than just one coder -- usually a mix between people with low level skills and high level skills (low level people make abstraction layers for the high level people, etc, etc). There shouldn't really be a trade off like you mention. Unless you're very much understaffed you should have high level and low level coders who in theory match the two types of coders you mentioned. Maybe I'm just not sure the point you're trying to get across...

I will say that, in my opinion, you can get done with 5-15 programmers (depending on project scope of course) as much as you can a team of 30+ -- throwing more people at the job is a highly destructive attitude that the industry seems to be in favor of, rather disappointing to see the publishers so eager to do that.

 

[mckmas8808]

If they are not showing actual in-game footage, they are showing hypotheticals.

From a marketing perspective, I'm not saying one is better than the other...however, I would argue that the closer you are to launch, the less hypothetical, and more "real" you need to be, otherwise you set yourself up for a backlash.

So it only makes perfect sense for MS to be more "down to earth" because they were always closer to launch than the other two.

In a non-direct way we are agreeing with one another.

 

[Laa-Yosh]

I imagine anyone creating a game will have more than just one coder -- usually a mix between people with low level skills and high level skills (low level people make abstraction layers for the high level people, etc, etc). There shouldn't really be a trade off like you mention. Unless you're very much understaffed you should have high level and low level coders who in theory match the two types of coders you mentioned. Maybe I'm just not sure the point you're trying to get across...

The main problem here is that especially with the PS3, the high level coders will have to learn hardware details because their existing programming approach would kill efficiency. And you obvoiusly can't get one programmer's high-level code constantly rewritten by someone else... so combining your team's individual strenghts will be pretty hard too.

And the original point was that a programmer's brain is always a better approach than a good cache algorithm, which I disagree with. As programming has to move to solve higher and higher levels of problems, it'll become increasingly difficult to manually build the low-end foundations for it by hand. Having to manually manage all the data around the SPEs might end up like forcing a developer to code each and every pixel and vertex shader by hand, in assembly... You may have a few, or even a lot of clever coders to do that, but the end result will still suffer, because the budget is limited, the deadline is closing in, and your team's efficiency will decrease with its size as well. So maybe Sony should've decided for a good cache algorithm instead.

 

[mckmas8808]

Having to manually manage all the data around the SPEs might end up like forcing a developer to code each and every pixel and vertex shader by hand, in assembly... You may have a few, or even a lot of clever coders to do that, but the end result will still suffer, because the budget is limited, the deadline is limited, and your team's efficiency will decrease with its size as well. So maybe Sony should've decided for a good cache algorithm instead.

Great point, but I guess we will have to wait and see if a good cache algorithm would have been better when the games come out.

 

[compres]

This is getting absurd. I have you telling me I have "have no clue about software engineering", and then you follow through by saying "That said I am no expert in cell assembly code."

Yes, thank you for your contribution to this thread. :major rolleyes:

Next time, drop the rude tone, if you want to have a civil discussion. I won't reply to you again.

How does me not being a cell assembly expert make you have a clue about software engineering?

By simple logic I don't seem to contradict myself like you are implying, enlight me.

I was certainly not trying to be rude BTW. Having a civilized discusion seems rather hard if it's with you.

 

[compres]

If I were sony, I would make the PS3 run Linux without the need of a hack. Ohh man a 300 bux console/PC would dominate the market for sure. Add a mouse and a keyboard(could be sold separatly, for what I care) and PC gamers will be happy.

Guess am daydreaming...

 

[mckmas8808]

If I were sony, I would make the PS3 run Linux without the need of a hack. Ohh man a 300 bux console/PC would dominate the market for sure. Add a mouse and a keyboard(could be sold separatly, for what I care) and PC gamers will be happy.

Guess am daydreaming...

Well what about the HDD that comes with Linux that Sony is going to sell? Will that do?

 

[Lord Darkblade]

Depends, if its a crippled linux (console OS+X) then probably no, if its a full release then you will likely have a lot of people jumping for joy.

 

[compres]

Well what about the HDD that comes with Linux that Sony is going to sell? Will that do?

Can I install my favorite dist. on it?

 

[mckmas8808]

Can I install my favorite dist. on it?

Now I'm not sure about that.

 

[Platon]

If I were sony, I would make the PS3 run Linux without the need of a hack. Ohh man a 300 bux console/PC would dominate the market for sure. Add a mouse and a keyboard(could be sold separatly, for what I care) and PC gamers will be happy.

Guess am daydreaming...

Sounds awefully familiar...

 

[Phil]

how many times must I tell you?

Sounds awefully familiar...

http://www.chez.com/ps2game/assets/images/internet.jpg

...with the exception that the above picture was never targeted for the mass market but for hobby programmers and enthusiasts, which is why it was sold exclusively via Sony's online site and never by retailers. Not to mention that the Kit itself isn't even mass market capable since the average windows user outthere would even fail to know how to use and set up for the reasons you're suggesting.

 

[mckmas8808]

Sounds awefully familiar...

Good try though. Play again later.

 

[Barbarian]

What a joke.
So please go right ahead and list the projects you are referring to. Clearly someone with your experience should be able to demonstrate that:
1) PS2 projects have longer development schedules than similar Xbox projects
2) Or PS2 projects require larger development teams than similar Xbox projects
3) List specific examples of development issues that caused 1) or 2)

Sorry for the late reply, I was away on vacation.
You must be pretty lucky to work for a studio where things go so smoothly.
I wasn't so lucky in my almost a decade of experience, I have worked in teams of 2, 10, 50 and 150. My last project had a team size of 100 of which 35 were programmers. The numbers are projected to double for next gen.
Strangely enough the number of "system" programmers always stays around 10%. So we have the scenario where 5 "system" engineers have to babysit 50+ engineers and keep them away from trouble, while concurrently developing "cutting edge engine". Don't get me wrong, I love spending months doing asm optimizations and rewriting the all important "inner loop" 10 times. But, these days that's not really possible anymore. It seems way more crucial that my time is spend helping all these other 50+ engineers doing their job fast and effieciently, so that they can iterate and polish as much as possible. Optimizations seem irrelevant when you have dull gameplay or crash bugs.
And to answer you question: PS2 had the luxory to be "the target" platform and everything was first developed on it, Xbox and GC "ports" were done on the afterburner with very few people and minimal time investment. All the gameplay people used VC/Xbox to debug and test their code.
As for development time - it seems similar at first glance - exclusives take 2 to 3 years. But this would be deceptive - Christmas shopping dictates these dev times. What we should compare is polish of the final game within same amounts of time, and especially when we go for aggressive schedules like 6-12 months. I have no doubt an Xbox/Xbox360 title would have significantly more polish and performance with equivalent time and effort. This is very difficult to verify, since good developers are naturally biased to the market leader. (ie just like comparissons between Pentium4 and PowerPC G5).
So, ease of development might be irrelevant, but the perception of being so is much more important especially when deciding what is "the target platform".