How should devs handle ports between consoles? *spawn

[zupallinere]

With middleware, as I said I believe they'll have to take into account for esram, no choice.
For in house cross platform titles, I also think there's also no choice. How much time is spent on making the best of it during one title, or improving its use in next title based on what is learned should be interesting.

Indeed. Middleware can be used by devs creating exclusive titles as well so while middleware developers will have to deal with the ESRAM but they are also under some competitive pressure to make good use of optimizations on both platforms.

I think that all things considered, driver maturity, architecture, that the XB1 line up is looking very good and actually bodes pretty well. Development hasn't been so bad that they are laughable.

A high bar you have there for success for the XB1 :smile:

 

[Tap In]

For those of you referencing the latest article from edge magazine about driver quality and performance. Something about the info in this article makes me think that all the anonymous qoutes
from developers are from pre gamescom and post e3. We are 2 months from launch with the time it takes to bug test and manufacture the software do you really think this is current info?

At the end of the article they qoute a dev saying sonys drivers are trash also. With everything we've heard from marc whitten and albert p about new driver performance and the same from sony guys do you think we are having this much trouble from the driver side this close to launch?


I think this info is at least a month or to old.

]

yea that Edge article sounds like "boring two months before launch" fodder

 

[shredenvain]

Since the systems are practically the same, when handling ports the devs are presumably going to use the same code base for each version. In a straight run like this you have a significant bandwidth, CU, and ROP advantage with the PS4. The Edge article give samples where the ESRam in the xbone will provide an advantage. If the majority of games are built around the same few engines as they were this generation, then will we ever see this actually being used outside of 1st parties?

Its been implied by some people that the use of the esram will be built into the system api.
Ive heard that if a developer wants to use the esram in different ways they will be allowed.
Even though we dont know the exact use of the esram I think its being a bit short sighted thinking that the esram is going to be some kind of huge hurdle or burden to use

I really dont think microsoft just threw in the esram becasue of bandwidth deficiency and decided to leave third party devs twisting in the wind on how to use it.

 

[Bagel seed]

There's nothing on PS4 that looks as good as Ryse so I'm not sure where the lowest common denominator claim is coming from. You're just ASSuming PS4 is more capable...yet the real world evidence says otherwise...ie Ryse and Forza.

*AHEM* Less talk on comparison of systems and more talk on the topic -- "How should devs handle ports". Thanks!

Since PC and PS4 are quite similar in regards to bandwidth amount and ease of use, those will probably be developed in tandem, or one after the other. Then you'd take the PC version and pare it down to fit Xbone.

On the other end of the spectrum you'd target Xbone first with a modestly spec'd PC target of 12 CU, get it completely optimized, then simply port over to PS4 and PC and adding extra features if there's enough time and resources.

 

[baten]

It will depend on sales of the consoles. If there is parity in sales and the nr. of XBones are similar to that of PS4s, then I dont see a reason for devs to bother to optimise too much for one or the other.
But if PS4 will take the lead by 2/1 or more - there is an incentive to work a little harder for the PS4 version.

 

[eastmen]

It will depend on sales of the consoles. If there is parity in sales and the nr. of XBones are similar to that of PS4s, then I dont see a reason for devs to bother to optimise too much for one or the other.
But if PS4 will take the lead by 2/1 or more - there is an incentive to work a little harder for the PS4 version.

and if the xbox one takes a lead by 2/1 would there be incentive to work harder to bring the ps4 version up to speed ?


Last generation dispite MS having a lead and the xbox 360 versions running better developers still invested int making the ps3 versions as good as possible.

I think this generation we will see an even smaller gap at the end of the day.

 

[baten]

and if the xbox one takes a lead by 2/1 would there be incentive to work harder to bring the ps4 version up to speed ?


Last generation dispite MS having a lead and the xbox 360 versions running better developers still invested int making the ps3 versions as good as possible.

I think this generation we will see an even smaller gap at the end of the day.

Somehow it is difficult for me to envision what must happen in order for MS to get a 2/1 ratio on Sony.

Last time around they launched one year early and the Xbox was pretty equal in power with the PS3, AND is was easy to develop for.

This time the launch together, but the Xbone is most probably underpowered AND it is not easy to develop for.

Anyway, why would a developer spend extra money to "bring up to speed" the version that will clearly bring less returns in terms of sales?

 

[Shifty Geezer]

Somehow it is difficult for me to envision what must happen in order for MS to get a 2/1 ratio on Sony.

This isn't a thread to speculate on sales! Just consider all possible outcomes and how they affect developers. Leave the business discussion to that thread, or we'll have it everywhere else!

I don't think comparisons with other generations regards investing in ports are directly comparable due to the hardware situation. To port your 360 title to PS3 took a lot of work. If you are going to spend a lot of work, may as well spend a lot making it good. Porting from XB1 to PC to PS4 to XB1 is, theoretically, not going to take much work, in which case you can treat the port as a cheap investment and do the minimum effort which may well afford the highest ROI. Spending an extra $n million to extract the most a platform's unique advantages to gain an on screen difference most folk probably won't notice doesn't sound a great way to get returns on software development.

Then again, if development is easier than you may find developers pushing each platform due to same-platform competition. If Console A is less powerful than Console B and so the Console B version of Death Kill 4 is left at the quality limited by Console A with just the resolution upped a bit, and then Murder Death 3 on Console B looks so much better because devs have invested in it, Death Kill 4 may miss out on sales to the better looking rival. That gives incentive to get the most from every box. Of course, although the incentive is always they to do the best job you can, business is all about compromises (art is about no compromises!) and within the range of softwares made, I expect most to take a rather conservative approach, keeping costs down and development easy.

 

[Billy Idol]

Aren't the Edge numbers a bit weird and not quite realistic (hence a bit fishy)?


1080p VS 900p alone would be a difference of 44% in pixels.

If we have now 30fps on PS4, everything below (roughly) 28.8 fps on X1 would mean that the PS4 is more than 50% stronger for this particular game than the One.

The suggested 20-something fps of the X1 version silently implies even more than 50% advantage.

I mean, there could be some explanations in terms of actual state of optimization and resulting efficiency of system resources. We know that numbers are only theoretical...but at first sight, the difference sounds a bit much.

 

[BoardBonobo]

Aren't the Edge numbers a bit weird and not quite realistic (hence a bit fishy)?

1080p VS 900p alone would be a difference of 44% in pixels.

If we have now 30fps on PS4, everything below (roughly) 28.8 fps on X1 would mean that the PS4 is more than 50% stronger for this particular game than the One.

The suggested 20-something fps of the X1 version silently implies even more than 50% advantage.

I mean, there could be some explanations in terms of actual state of optimization and resulting efficiency of system resources. We know that numbers are only theoretical...but at first sight, the difference sounds a bit much.

Not really. They're talking about a plain vanilla build, so it's DDR5 vs DDR3 bandwidth, 18 vs 12 CU's, and 32 vs 16 ROPS. Plus all the system controlled cache access's and bypasses.

The ESRAM is the more exotic solution this time around.

As side note, wiki have listed the max CPU speed in the PS4 as being 2.75Ghz

 

[Shifty Geezer]

As side note, wiki have listed the max CPU speed in the PS4 as being 2.75Ghz

1) the source is linked to the FCC filing in July, which is for the PS4 devkit and which specifies a maximum clock frequency in the devkit of 2.75 GHz without stating which clock that is.

2) this isn't the thread for determining what either platform's specs are.

 

[brasnial]

EDGE: "Power struggle: the real differences between PS4 and Xbox One performance"



Much more at the link.

http://www.edge-online.com/news/pow...erences-between-ps4-and-xbox-one-performance/

One more particularly interesting quote.

After reading the the full piece I see it as this ....developer says code wrote on PC then ported onto the PS 4 runs faster by as much as 50 percent than the same code ported to Xbox one .
You don't say isn't the PS 4 just a medium PC in a console case .
The real question should be is that PC code using the esram at all and what happens once you configure the code to make use of the esram .
A more interesting question would be how well does a PC or the ps4 handle code written for the Xbox one from the ground up do they coke on that code or run it 50 percent slower than the Xbox one .

The piece is fud designed for hits as its does not go into enough detail or facts in my opinion .

 

[blakjedi]

The edge and cnet.com art8cles are a random mishmash of quotes from anonymous revs. I wouldn't really listen to them. Xb1 is certainly more exotic but lets not pretend like exotic=hard. Every major 3rd party dev house in the world has built games for 360. A system with an in order processor and edram. XB1 with commodity pc hardware could not realistically be considered harder.

Killzone might be the only title that looks like it displays an advantage however, it was designed to showcase particle effects. Ryse on the other hand has the best lighting I've ever seen.

No one can call the lcd yet.

 

[Shifty Geezer]

*Dons Angry Eyes* This is not a which platform is best thread. Let's not be drawn into a "does Console A outperform Console B or not" discussion. Regardless of who performs better, devs face choices. Those devs reporting issues with one platform are reporting what they are experiencing - whether because one platform is harder to use, or slower, or those devs are just lazy or stupid or rushed or whatever - and those experiences affect their development choices.

The discussion here is basically the economics of targeting platform variations and whether devs let one version slide because they can't get it running fast enough cheap enough, or push one version because the platform can do more, or strike a middle ground, which is especially interesting this upcoming generation because there's a hardware common-ground that invites a common approach to software and which may result in platform deviations being overlooked.

So to iterate, which platform is faster/slower is not the subject. How devs handle differences is. If people have trouble talking about the consoles by names without caring whether one platform performs better or not, then take to calling them Console A and B and PC as hypotheticals. But I will be getting arsey with more OT versus stuff. The internet is abuzz with PS4 versus XB1, but we're not doing that on B3D (until we have enough reference material to address the comparison on a technical level, as ultimately the intelligent evaluation of two different hardware approaches to the same product solution is B3D's raison d'être).

 

[Pixel]

"Developers tell me that working on PS4 is much easier than Xbox One. A flipped situation from 360 vs PS3. Sounds like MS dropped the ball."
George Broussard (@georgeb3dr)
https://twitter.com/georgeb3dr/statuses/393012014798229504
October 23, 2013

George worked on Commander Keen, Duke Nukem 1/2, Executive producer on Rise of the Triads, executive producer on Duke Nukem 3D, executive producer on Shadow Warrior, a dozen spin off dukenukem ports/expansions and other dukenukem filler titles, producer on Prey, Executive producer on Duke Nukem Foever until Gearbox took over development.

 

[liolio]

Developers tell me that working on PS4 is much easier than Xbox One. A flipped situation from 360 vs PS3. Sounds like MS dropped the ball.
George Broussard (@georgeb3dr)
https://twitter.com/georgeb3dr/statuses/393012014798229504
October 23, 2013

George worked on Commander Keen, Duke Nukem 1/2, Executive producer on Rise of the Triads, executive producer on Duke Nukem 3D, executive producer on Shadow Warrior, a dozen spin off dukenukem ports/expansions and other dukenukem filler titles, producer on Prey, producer on Duke Nukem Foever until Gearbox took over development.

That is not really disputable, I don't think that there has been a system out there that is as straightforward as the ps4 is. There are APUs in the PC but nowhere that level of graphics performance. On the ps4 devs should be able to do pretty much whatever they want.

 

[FordGTGuy]

"Developers tell me that working on PS4 is much easier than Xbox One. A flipped situation from 360 vs PS3. Sounds like MS dropped the ball."
George Broussard (@georgeb3dr)
https://twitter.com/georgeb3dr/statuses/393012014798229504
October 23, 2013

George worked on Commander Keen, Duke Nukem 1/2, Executive producer on Rise of the Triads, executive producer on Duke Nukem 3D, executive producer on Shadow Warrior, a dozen spin off dukenukem ports/expansions and other dukenukem filler titles, producer on Prey, producer on Duke Nukem Foever until Gearbox took over development.

Should we trust the producer of Duke Nukem Forever?

Honestly another guy claiming anonymous devs?

Where are the actual devs complaining?

 

[DrJay24]

Well Ford, you just posted a comment by a director with a degree in merchandising comparing the two console's power. You can't have it both ways.Take the good with the bad.

 

[FordGTGuy]

Well Ford, you just posted a comment by a director with a degree in merchandising comparing the two console's power. You can't have it both ways.Take the good with the bad.

That director is actually developing a game on XOne and PS4 with no obligations to either console...

http://en.wikipedia.org/wiki/Shinji_Mikami

He might have a degree in merchandising but he has a helluva gaming development career under his belt.

I have no doubt the PS4 is easier to develop on but will this make a actual difference down the road? Not really.

I hate to say it but it can't be too easy to develop on with games like Driveclub being delayed a month from launch.

 

[Pixel]

Should we trust the producer of Duke Nukem Forever?

Honestly another guy claiming anonymous devs?

Where are the actual devs complaining?

Many developers are not going to want to come forward and say things that could hurt their studio's relationship with Microsoft. They are not going to want to be "that guy" responsible for soured relationship. If they were the one responsible for the souring that could hurt their career path in the studio.

Guys like Valve's Gabe Newell can say what they want and not fear retribution from a superior.