*Game Development Issues*

Status
Not open for further replies.
...I would say the evidence at this point is on the side of the 3rd party guys who are having a harder time with the PS3. That could change a year from now but as of right now anyone with 2 eyes can see the 360 has better tools or is easier to developer for.

That more or less sums things up.

For an ace coder it might even be an easy platform. Just like there are people who say running a marathon under 3 hours is easy. Some people do it in just over two hours.

Most people wouldn't agree with this statement though.
 
That more or less sums things up.

For an ace coder it might even be an easy platform. Just like there are people who say running a marathon under 3 hours is easy. Some people do it in just over two hours.

Most people wouldn't agree with this statement though.

I think that some here are also forgetting the the 360 is out a year longer on the shelves which surely adds alot of experience for 3rd party developers that has yet to be achieved on PS3 (maybe next year).

Plus those mulitplatform titles are really not that bad most of the time you cannot even tell the difference if you are not told what it is.
Btw. we also have a couple of even older multiplatform titles that disprove this theory since they look and run better on PS3.

With current multiplatformers like Assassins Creed or CoD4 it seems some got alot better at the marathon already though ;)
 
That could change a year from now but as of right now anyone with 2 eyes can see the 360 has better tools or is easier to developer for. It is a shame that people are so quick to blame "lazy developers" for the last year instead of sony,ibm and nvidia.

Looks like COD 4 and Assassins Creed is a year early. Even without the juggernaut Microsoft behind the "tools".

Of course it seems like they did both versions more or less at the same time instead of having a "lead" platform or "porting".
 
The Wii has to prove it has longevity.

In a year are all the games going to have a familiar feel to them, is everyone just going to make games using the wii-mote in the same way (eg will all FPSes be like Metroid Prime)

There are some examples where the wii-mote is great (mainly nintendo own games) and there are some where it's absolute tripe. We've already seen that Nintendo cannot keep a console alive single handedly.
 
I am no developer but I would say the evidence at this point is on the side of the 3rd party guys who are having a harder time with the PS3.

Evidence of what..? This is part of the problem here, people seem to have different conclusions they're trying to present evidence for.

It is a shame that people are so quick to blame "lazy developers" for the last year instead of sony,ibm and nvidia.

I think Sony's dev environment does demand more experience, more skill, and perhaps more conducive circumstances than Microsoft's. You could say Microsoft's is more forgiving of an imperfect world. But if I say there's an issue of competence too on the part of developers, I think it should be clear I'm not blaming them for that. It's just the way it is, some developers are better than others.

I think the whole situation is changing though. We have to remember this is the PS3's first year. Various developers more recently have been coming out saying that they're in a better position now with PS3 development, others saying there isn't a problem at all. I think as more and more developers get PS3 projects under their belt past the first year, this'll be the common situation.
 
here is another take by Morten Heiberg (Hitman, Kane & Lynch):

When asked if developing for the PS3 was a headache for IO, Heiberg said that despite it requiring additional resources and special effort, most developers are "keen" on the platform because of its interesting architecture.

I doubt that generalization is accurate though.

Here's another article that briefly touches on the subject of sales this Xmas.
http://www.gamesindustry.biz/content_page.php?aid=30407
I find it interesting that his opinion on the Wii is that it has still to prove itself. That's true but couldn't the same thing be said about the PS3? If third parties are selling more on the Wii than PS3,why does the Wii seem to be singled out as the system that has to prove itself to third parties,but it's assumed that everything is fine on the PS3.

Since PS3 is getting 360 ports I don't think it needs to prove itself as much as wii.
 
Well, the further continuation of that, they go on to say things like that it's "fun" and that it reminds them of earlier times in their university days... that's a sentiment I've heard echoed by a lot of people. I know a guy at EA who worked on The Simpsons who has said the same thing. Problems tend to come from the fact that not everybody has had that experience before to be able to arrive at such a conclusion in the first place.

The arguments frequently made by many including several on this thread is that PS3 doesn't leave you with any choice. And while certain individuals might not mind that, it's not good at a studio level.

I find it interesting that his opinion on the Wii is that it has still to prove itself. That's true but couldn't the same thing be said about the PS3? If third parties are selling more on the Wii than PS3,why does the Wii seem to be singled out as the system that has to prove itself to third parties,but it's assumed that everything is fine on the PS3.
Well, it seems his argument is more that there are concerns throughout about the Wii's ability to hold onto its current momentum. If the flair fizzles out, then the road carved out by PS3 and 360 will be the preferred one for third parties.

That too, I think the way to really make something of the Wii that could stand as a "good Wii title" is not necessarily an easy proposition for 3rd parties who might be more accustomed to "traditional" console titles.
 
The arguments frequently made by many including several on this thread is that PS3 doesn't leave you with any choice. And while certain individuals might not mind that, it's not good at a studio level.
I think that's where the fundamental disagreements lie. There are those who see PS3 as a funky piece of hardware where it's cool to exercise the brain into using it, and then there are those who's interest is more in creating the game and the mechanics of how to implement it isn't an interest to them. For management and financial departments, how much fun your team are having doesn't normally run high on the priorities list! And for developers who are happy creating software with the systems they know, rethinking everything isn't going to win friends.

What's the retrospective on PS2 like? Surely few developers entered into developing for that comfortably. By the end of the generation were they comfortable with it, or still complaining about the weird hoops PS2 made them jump through?
 
However, I would expect that with VSM, doing the actual rendering is faster on Xenos since it has high precision filtering (32-bit integer) whereas RSX doesn't, and I know from personal experience that VSM with FP16 is ridiculously limited. You'd have to do it all in the shader.
I'd like to see some games doing SAT-VSM on PS3 via SPUs :)
 
I think that's where the fundamental disagreements lie. There are those who see PS3 as a funky piece of hardware where it's cool to exercise the brain into using it, and then there are those who's interest is more in creating the game and the mechanics of how to implement it isn't an interest to them.
I guess you will find ppl enjoying both aspects.
 
I guess you will find ppl enjoying both aspects.
Indeed. And ideally you'd have the people needing to mess about with PS3's innards being the people who enjoy it, rather than those who'll complain! Maybe they're aren't enough of those?
 
I'd like to see some games doing SAT-VSM on PS3 via SPUs :)

I'd like to see a game do SAT-VSM period :)

And I don't know why people think X360 has 32 bit integer filtering. I have the list of supported surfaces in front of me and there is no 32 bit integer. The best one can do is 24 bit depth/stencil surface, and write to it via oDepth, or via edram aliasing. If anyone can enlighten me how to get 32 bit integer that would be amazing.
 
What's the retrospective on PS2 like? Surely few developers entered into developing for that comfortably. By the end of the generation were they comfortable with it, or still complaining about the weird hoops PS2 made them jump through?

My 2c on this, from my personal experience at the tail end of last gen development, had last generation gone on much longer say a couple more years PS2 would have been lead SKU in name only on a lot projects.

It was a lot easier to develop and debug on Xbox, on my last lastgen title we had to take XBox devkits off some peoples desks to make the run and test on the PS2. Because at one point the graphics engine was behind on the PS2, and the pressure for E3, out lead PS2 SKU didn't even compile when we showed the demo.

Having said that we have way more people working on the technical side of the PS2 SKU for the duration of the product and we were still aiming at it's feature set for the most part.

A lot of this was also a function of the team size, and some unique issues on the graphics side (not technical). But I've heard similar stories from other devs.
 
I think that's where the fundamental disagreements lie. There are those who see PS3 as a funky piece of hardware where it's cool to exercise the brain into using it, and then there are those who's interest is more in creating the game and the mechanics of how to implement it isn't an interest to them.

There are a few (possibly more that a few) engineers mostly those in core tech who love the PS3 architecture because it's "cool", and I think this fact alone has bought Sony a lot of lee-way with devs on the studio floors.

Frankly if shipping a game wasn't one of my responsibilities I'd probably love it too.

Now I haven't had to deal with PS3 since the very early days, so take anything I say with an enormous pinch of salt.

For the most part games now unlike in the Amiga days aren't all about the technology, the primary engineering problems are the scale of the software and rate of design change. There are probably core guys who'd disagree with me on this, but I've done that job too and it's just so constrained and static as a problem by comparison.
 
We heard about RSX having half or a quarter vertex shading perf of Xenos, well...as I already said I think there are already 2 or 3 games on the shelves that kind of disprove these statements , but what do I know? ;)

Ok, so my 'quarter' comment is probably a stretch :) Here's something that not a stretch:

Total verticies: 10,103,200
Accounting for GPU only, this frame would take about 33.56ms to run at 29.79 frames per second.

That is taken directly from a Pix grab I just did. Now, this is not from a test app. It's not from someone doodling on XNA. It's from our actual game. The only difference is that I disabled some lod's systems to force a higher vertex count that normal.

So, here we have Xenos, totally on it's own, ripping thru 10 million verticies and still mantaining 30fps. Now, you are telling me that you can get RSX to do the same, if not better, with zero spu help. If you can then I take my hat off to you. You are doing something no one else I know (including Naughty dog who use spu's on Uncharted to help geometry) is capable of doing.

FYI in case anyone is intersted, yes the game does look better with all these "wasted" verticies.
 
Total verticies: 10,103,200
Accounting for GPU only, this frame would take about 33.56ms to run at 29.79 frames per second.
Wow, that's impressive. Isn't it higher than actual vertex setup on RSX?
 
I think it worth pointing out that Heavenly Sword, while it looks very nice in stills, suffers from major technical issues, most importantly a very dodgy framerate and excessive screen tearing.

It was also originally scheduled to launch with the PS3, and recieved over an 8 month delay.

So, while it was a good attempt at creating an exclusive technical showpiece, NinjaTheory definately bit off a little more than they could chew.
 
No offense, but the numbers you quoted (while being very impressive..) are not telling you (or me) that much about the amount of work Xenos (or any other GPU) is doing.
You're measuring the number of vertices that are going through the GPU (for example it's easy to multiply your figures by over a factor of 5 on RSX and 360 with an ad-hoc benchmark), while what you want to measure it's the number of post transformed vertices cache misses per frame.
Whch is the first number you should look at to understand how many vertices you're crunching, otherwise you're doing a poor profiling job, as you think you're measuring on aspect of the chip, while you're measuring another one which is not that much interesting (you're measuring the total number of hit+misses from the pre-transformed vertex cache)
Said that I expect Xenos to be faster than RSX at processing so many vertices, cause it's likely that the amount of pixels generated per primitive is so low anyway, that most the time Xenos will run with 16 virtual vertex shader pipes, though there are many other variables that we should account for (for example I know nothing about your tipical vertex shader, number of input and output attributes, etc..)
 
To Barbarian:

I recall your difficulties with fitting ( debug ) code on SPUs and proposal to rectify it for the next iteration of Cell.

Is there nothing that can be done for you now?

For instance one could view code segments is just another set of streamable data. One could just as easily bring their code to the data as their data to their code. It may be possible to reclaim a good portion of the LS memory consumed by code blocks for you.

There is the potential that some branches may become expensive but it seemed as if you were willing to accept this earlier. However, predicting which code segment to load up next should be far more predictable on the average than which data set to bring in.

There has been some success with approach elsewhere although probably not in respect to this specific need. I would like to interpret this as fringe benefits which could be valuable to you.

What do you need out of the C++ compiler before the generated code is optimal enough for you? (or at least acceptable enough for you to use it freely in abstraction)

Insomniac Games has at least proposed SPU "shader" have you or any of your team members contacted them on their progress?

You feel Sony needs to provide more high level libraries ( not to suggest there aren't any ) where do you think their focus should be?

Sorry for hanging all these questions on you but you were the one who primarily brought up issues which -->can<-- be solved or smoothed over a bit in being software related. As has been said before the hardware is what it is so nothing can be gained in the near term by ruminating over what's good or bad about it.

That said anyone is free to offer their thoughts on actually improving anything.
 
Status
Not open for further replies.
Back
Top