Well gamespot lists Mars as coming out for the PC, 360, and PS3, so I guess they are multiplatforming.

Stands.... *Slow clap*.....*faster*.....*Enthusiastic applause*.

The difference between these guys and Valve is that these guys don't whine -- They actually try when it comes to the PS3.

Kudos to them.

As others have pointed out, this team appears to have talent. Unlike some other whiney studios who I will not mention by name.

@Aaronf,
Are you really saying that Valve doesn't have any talent?

On topic, our projects tend to work like this as well. It is much easier to port stuff to the 360 than the other way around. It slows down development a little but it's worth taking that hit during the entire length of a production than to get stranded near the end with tons of ps3-specific bugs that need to get squashed.

An easy example: Since the ps3 has 256 megs allocated to video memory and 256 megs of all-purpose ram, it's easier to focus on that console since the 360 has 512 of unified ram that you can split in any fashion you wish. It's bad news when you start on the 360 and realize that you are busting the ps3 graphical memory budget by 30 megs, even if things were running just peachy on the 360.

The PS3 isn't that complicated. If you understand multi-threading (a pre-requisite for the 360), or shader language (a pre-requisite for any complex GPU work) then Cell programming is fairly straightforward. An SPU task is basically something you spawn, passing it data and then you wait for it to signal you with a result or terminate with a result. SPU code is written in C using certain intrinsics to speed up processing data. There are certain performance issues to be concerned about such as conditional branches but nothing too difficult to grasp.

The bigger challenge is not the SPU but factoring your code such that you make maximum use of it while sharing as much code between 360 and PS3 platform. For example the 360 and PS3 are both suitable for master / slave architectures so perhaps a main thread should divide up work and dispatch it to slave threads / tasks who do the work. If you have a routine that does collision detection you probably want the master dispatcher to be cross-platform and for the interface and data structures to be cross-platform but the slave implementation to differ. The 360 might spin a thread to service the work, the PS3 might start a SPU task. It's this dispatching and abstraction which is the hard part and it requires proper upfront design.

Obviously the more you can share, the less bugs you will have. So is the PS3 complicated? No not really. The complexity is in producing a proper framework that supports all platforms equally.

As for Valve, the reason for their bitching probably has more to do with producing an engine which does none of this. It's written over DirectX and makes assumptions which mean (shock horror) it doesn't work on PS. Probably why they tossed it at EA to port. Maybe next time they make an engine which is cross-platform out of the gate.

i have never heard of this game before, but hey -- at least i have the option to play it on my PS3!
seriously, devs -- don't punish us for not owning a 360.

@ChaosTeaCup haha now that is a prize winning comment, totally agree!

Now, now, rotten...Let's not let this topic 'snowball' into digression.

Programming is never easy. If done correctly you can make millions so get over it.

@ All:

God forbid Valve choose to not develop for the ps3. Jeez people.

@KrazyKraut, @DrXym I think the reason that most devs are scared of the PS3 doesn't have so much to do with the technical aspects of the development, that much can be overcome (though they like to use it as an excuse)- it all comes down to the almighty dollar. The same crew can develop and debug PC and 360 games and often you don't need to do much to work on pre existing code to dev a DX engine for it so cost wise it is waay better for the company than to dev for PS3. If there were a larger install base it would make more sense to have PS3 only houses or PS3 primary houses but if you consider that the higher costs give you a lower return the choice is made for most smaller houses in favor of the xbox. Personally I am a PS3 and not a 360 owner not because I am a fanboy per say but because hardware wise the 360 is a POS and everyone I know who has one has had multiple RROD's, buttons falling off, busted optical drives, etc and I don't want to be in that boat- maybe I will get one later to play exclusives that I haven't been able to play but most of them end up on PC for cheaper

@glandseck:

When they would rather whine than make the effort to learn, than yes I am saying that have no talent.

I can understand why a company might shy away from PS3 development. It's difficult, it costs additional money, and the returns will probably not be great. Yet. However, in the long run, it may cost them.

Barring some great catastrophe, such as Japan sinking into the ocean, it's not like the PS3 is going anywhere at this point. More people keep buying them, the costs are going to continue to lower - these are inevitable facts. Anyone that is still hoping for the ps3 to "fail" really needs to find a hobby. Even if they remain consistently in third place this generation, it's still on a scale of millions-of-units sold. I believe there's plenty of room in the market for all three consoles.

When ps3 development hits its stride, will these companies that opted for the easy way out be in a position to catch up to the companies that were early adopters? Maybe they won't have to; maybe they'll never have to develop for the ps3. However, it does smack of lost potential.

@whormongr
Yeah thats an another reason. But we had no discussions 4 years ago, when we had cross platform titles for psp, ps2, xbox and gamecube( for example: PoP: Sands of time). like you said: dx runs on 360 and WinPc. openGl is a little bit different. its cheaper to make a game for 360 and PC cause its yeah..similiar workflow.