*Game Development Issues*

[scently]

I think that the relative ease of development is actually msft plan coming into this generation ie to help reduce the cost of starting and completing a project. I think that they took consideration of the expected rise in dev cost and so develop a system whereby one could use your pc development knowledge to start. Now imagine if microsoft developed a system that is as hard as the ps3. I think the industry should be grateful about the way things are bcos it could have been much worse.

Also i would like to add that i totally agree 100% with ostepop, i just don't get the whole "leading on a ps3 leads to a better xbox360 game.

 

[Tap In]

I think that the relative ease of development is actually msft plan coming into this generation ie to help reduce the cost of starting and completing a project. I think that they took consideration of the expected rise in dev cost and so develop a system whereby one could use your pc development knowledge to start. Now imagine if microsoft developed a system that is as hard as the ps3. I think the industry should be grateful about the way things are bcos it could have been much worse.

Also i would like to add that i totally agree 100% with ostepop, i just don't get the whole "leading on a ps3 leads to a better xbox360 game.

Those of us who have read Xbox Uncloaked for 360 discovered that that was exactly what they had in mind when designing this system. they spent quite a bit of resources on doing what they could to find what would be needed by devs and to make it easier.

 

[mrcorbo]

I'm not convinced the "lead on PS3 = better 360 version" has much validity these days if it ever did.

The 360 is coming up on its 3rd year on the market with each subsequent title release having to at worst equal the fidelity of prior releases in order to meet customer's expectations. After 3 years of developers aiming to equal or better prior efforts I'd be very surprised if they weren't long past the point of having the luxury to cling to old programming paradigms that don't utilize the 360's hardware resources efficiently and still end up with a product that is competitive in the current marketplace.

 

[Fafalada]

What the best possible practices is for the PS3 isn't necessarily the best possible practices for the 360.

Multiplatform development by definition isn't aimed at best possible practices on any of the supported platforms. As mentioned multiple times over in this thread, you're looking to deliver a product competitive with other software on each of the supported platforms, not something that may look fine on one and completely out of place on other(s).

And the latter has been the common difference between 360 and PS3 leading development - not the silly arguments over whether 360 uses all of its eDram well or PS3 uses enough flops or whatever.

I'd be very surprised if they weren't long past the point of having the luxury to cling to old programming paradigms that don't utilize the 360's hardware resources efficiently

You never know - 3 years into PS2s life lots of SW was still doing the same mistakes as some launch titles did. The obvious things get worked out quickly enough, but paradigm shifts can take ages to take hold, if ever.

 

[mrcorbo]

You never know - 3 years into PS2s life lots of SW was still doing the same mistakes as some launch titles did. The obvious things get worked out quickly enough, but paradigm shifts can take ages to take hold, if ever.

Was the PS2 market at that time not better able to make even titles with marginal appeal financially viable compared to the situation today with the 2 HD consoles? Between the much lower cost of development and the much larger userbase a project didn't have to achieve nearly the market penetration that today's titles do to make money.

Anyway, if you truly think that having the PS3 as lead platform will generally lead to a better 360 title than leading on the 360 would - even now- than I'll just have to be "very surprised" and leave it at that.

 

[obonicus]

Anyway, if you truly think that having the PS3 as lead platform will generally lead to a better 360 title than leading on the 360 would - even now- than I'll just have to be "very surprised" and leave it at that.

How many upcoming 3rd party exclusives are there anyway, at this point?

 

[Quarke]

If I was developing a game. I would target the 360 adn then port the game. Perhaps even be able to negotiate some form of timed exclusive period with MS while i continue work on the ps3 version. esp if its a Western targeted game.

Very good strategy. Create an inferior PS3 version, alienate the PS3 user base and create the perception that your company doesn't have the aptitude to properly develop for the system. Then delay the PS3 version by asking Microsoft for compensation to sign an exclusivity deal, blow your marketing dollars on the 360 version and then silently release the PS3 version to audience, already aware of your inept PS3 development skills, and pray for it to sell. And then ponder why you even wasted your time on the PS3 version when the sales numbers come in.

Fortunately, your strategy for success has been made obsolete and irrelevant for today's MP titles because the development tools and frameworks have improved to such an extent that the need to code in straight SPE assembly language is no longer a required attribute. EDGE has leveled the playing field substantially and is the primary reason why we're seeing parity for titles that are primarily developed on the 360.

 

[NeoTechni]

The fact that he has worked on non-graphically demanding titles makes his complaints about the PS3's apparent lack of power and unpleasant development MORE valid, not less.

I don't see how his comments are made more valid than say, Carmack

Let's take a 2D sprite-based game engine programmer. By your logic his opinion is even more valid?

 

[assen]

I don't see how his comments are made more valid than say, Carmack

Let's take a 2D sprite-based game engine programmer. By your logic his opinion is even more valid?

A good engineering goal for any kind of platform (library, engine etc.) is that it should make simple things simple and complex things possible. An ideal development platform should make making a 2D sprite-based game like Braid* a breeze, something like Guitar Hero/Rock band relatively easy, and still would make something like (the expected graphics powerhouse) Killzone2 possible. PS3 seems to stray far from this ideal.

The impression that it's "hard" coming from someone who tried simple things are more damning than those of someone who tried to make a state-of-the-art renderer, for example.

* Why am I bringing up Braid - here's what Jonathan Blow, who created Braid, had to say on a possible PS3 port:

I think it might be pretty hard to port Braid to the PS3 (speaking as someone who has done PS3 programming before). Jeff at RAD Game Tools tells me they have a pretty fast Vorbis decoder that (I think) uses the SPUs, so that is one of the main stumbling blocks. But there are other parts of Braid that are more expensive CPU-wise than you might think, and they might have trouble running at sufficient speed on the PS3.

 

[FirewalkR]

Sony must give him all the help he needs. (well, short of developing the game for him, heh)

 

[eastmen]

Very good strategy. Create an inferior PS3 version, alienate the PS3 user base and create the perception that your company doesn't have the aptitude to properly develop for the system. Then delay the PS3 version by asking Microsoft for compensation to sign an exclusivity deal, blow your marketing dollars on the 360 version and then silently release the PS3 version to audience, already aware of your inept PS3 development skills, and pray for it to sell. And then ponder why you even wasted your time on the PS3 version when the sales numbers come in.

Fortunately, your strategy for success has been made obsolete and irrelevant for today's MP titles because the development tools and frameworks have improved to such an extent that the need to code in straight SPE assembly language is no longer a required attribute. EDGE has leveled the playing field substantially and is the primary reason why we're seeing parity for titles that are primarily developed on the 360.

You assume way to much.

I wouldn't delay a ps3 verison , it will simply ship when its done. The game also doesn't have to be crap. It can be a very good game. Is bioshock a crap game cause it came out a year after on the ps3.

Also I don't know if we re seeing parity with multiplatform games. I still read of many complaints on the ps3 verisons.

 

[assen]

Also I don't know if we re seeing parity with multiplatform games. I still read of many complaints on the ps3 verisons.

What we're seeing is that parity is possible. Read the latest Eurogamer comparison article, half the titles still have obviously inferior PS3 versions.

 

[eastmen]

What we're seeing is that parity is possible. Read the latest Eurogamer comparison article, half the titles still have obviously inferior PS3 versions.

I would hope that the more expensive machine could reach parity with the cheaper machine. The only question is if that parity is coming from gimping the growth of titles on the cheaper machine.

 

[Arwin]

I would hope that the more expensive machine could reach parity with the cheaper machine. The only question is if that parity is coming from gimping the growth of titles on the cheaper machine.

Depends on where that money was spent, how different the hardware is between the two machines, and how much the software has been written specifically for the hardware of the cheaper machine, don't you think?

 

[patsu]

Dean Calver's latest blog entry: http://blog.deanoc.com/?p=111

My last blog entry mentioned I was starting a new job, now several months on I’m nicely settled in and happy to be getting on with things.

My new job is Lead Programmer at Splash Damage, working on a, as yet, unannounced game. The only thing that has so far been released is that the publisher is Bethseda and that it will ship on the main platforms you would expect a kick arse game to ship on.

Its been a good few months getting into the feel of things, Splash is most known for the Enemy Territory titles (Wolf and Quake Wars) with Id but things are a bit different this time, with more work (particularly the console versions) being done in house, so the company has been recruiting a host of experienced staff to help with key roles. It has an excited time, as people who have worked on various games I respect and enjoy myself join the team. The list of titles that members have worked on, include Heavenly Sword (obviously , Mass Effect, Syphon Filter, Fable 2, Rainbow Six and many more. Were still looking for some good people by the way, particular for Environment artist, FX artists, UI artists, senior graphics and AI programmers, so get in contact if you think you’d like it here.

Apart from settling into the lead role, getting to grips with tasks and schedules, I’ve also been doing a fair bit of PS3 coding. A good tidy up of the graphics backend to a more libGCM paradigm has taking a fair bit of my time. Currently sorting out our SPU framework, one interesting decision I’ve made is to support a SPU-like framework on PC and 360, the idea being that code written for SPU local store, vector units and DMA will run quite well on a 360 core, with cache standing in for local store, VMX128 for the SPU vector unit and cache prefetch as ‘pretend’ DMA. Obviously there are lot of cases were writing a specific version will be better, but in theory anything that runs well on SPU should run better than normal C code on a 360 CPU.

 

[Squilliam]

I was wondering, if graphics technologies such as tessellation which are shared by the Xbox 360 and Direct X 11 are to be used, does it make the it problematic to produce an equivelent PS3 version in terms of coding an engine to match or even being able to produce an equivelent effect in the first place?

The issue doesn't seem to be a lack of triangles, but perhaps the memory/bandwidth savings to be gained from tessellation. If the Xbox 360 already has a memory advantage, then saving a further 10% from one SKU could tip things away from parity.

 

[Cheezdoodles]

If the games were better on PS3, as they should be (imo KZ2 beats UE3 games from what i've seen), then i could just laugh at the 360, and put it in the box. Now however, i always need to consider it as an alternative to my PS3.

So your comparing a $50+ million budget title developed exclusively for the PS3 for god knows how many years now, vs games made on a multiplatform engine?

I doubt thats a good comparison.

Fallout 3 the same: it (reported) runs like crap on ps3. How can that be? The last game from bethesda, Oblivion, ran better on PS3; higher res, smoother framerate, better texture filtering, less popup, faster loading times.

Oblivion ran better because of allmost a year extra development time, not because of PS3 hardware, Bethesda had a year to tweak the game to run well on the PS3.

The X360 version came out only 6 months after developers had gotten the final hardware.

Extra time does wonders for performance, just look at first generation titles vs 2nd generation titles. Same hardware, more time, better results. Its not like Oblivion pushed out everything we can do on either console.

How come they fuck it up this time?

Have you played many Bethesda Softworks games? They are not exactly good at making things run well. Their games are notorious for requiring tons of hardware power compared to what the game actually does.

Do they get money from MS?

Grow up.

 

[AbbA]

So your comparing a $50+ million budget title developed exclusively for the PS3 for god knows how many years now, vs games made on a multiplatform engine?

I doubt thats a good comparison.

Ten years or more of development around Unreal engine....

 

[AlNom]

Ten years or more of development around Unreal engine....

Not on 360.

 

[warb]

Ps3 version has better lighting and texture and drop frame randomly like 360 version.

Dead Space seems very close in the comparisons, with maybe slightly better texture filtering and a few extra effects on 360.