*spin-off* Importance of Backward Compatibility Discussion

[function]

If Xbox 3 doesn't let you co-op with and multi-player against all the Xbox 360 owners on your friends list it fails badly as platform who's primary purpose is to support Live. You get your expensive new console home, put it in the old console's place, and find out that not only does your old content not work but the "beacon" that your friends are heading to is simply highlighting your new found exclusion.

Never mind, you can simply dig out your old console (unless you sold it), swap out your new console under the telly for your old one, set up a network bridge or run an ethernet cable through the house, and bam who needs BC to keep up their social gaming and justify that $60 a year Live Gold.

I'd happily pay a premium for a BC console or see the overall power of the system reduced in other areas to allow this. In a theoretical IBM/AMD next gen Xbox I'd be very happy to see a few mm^2 dedicated to supporting BC, and MS could always cut it out once the transition is complete a few years after release.

Put it this way: if they don't, they lose me as a customer and lose my annual Gold subscription (for which they basically do f. all).

 

[tcrews]

To throw in my two cents.....I don't care if the next one has BC or if it does. If (and that's a bit IF) having BC affects the advancement of the next system then definitely leave out BC. I'll still have my 360 for a while after I get the NeXtBox, just like I had my Xbox after the 360 came out. My TV/Surround system has enough inputs to cover both as well as the new one will most likely have built in WiFi, so no need to run an extra cable (though I can do that two.....).

I can't think of a game I want to pull out from my Xbox library that I want to play now....so for me, BC is only important during the first 6 months or so of a Console's life (while the new library of available titles builds). After that, it's on to the new, better, new generation games. My 360 can last long enough to cover that and then I'll sell it like I did my Xbox. DLC doesn't bother me because it's for the game I'm "not" playing anymore anyway.

If BC negatively affects the future systems advancement.....I do not want it. If it's a non-factor than having it is fine.

I've had friends move from PS2 to X360 without care that they lost their old game library, I've had friends move from Xbox to X360 and ditched all their Xbox games without a care.

BC will not (in my opinion) be a factor in people upgrading or not. Hype, wanting the new console/games and getting the latest and greatest is going to be the major factor.

 

[Shifty Geezer]

The more that the next gen consoles are "similar" to each other, the more BC has value. The more that they are different from one another, the less BC has value.

Can we live with that? Then we're left at theorizing how different the next gens will be from one another.

Definitely. That's what I was getting at. BC has value - how much is hard to pin down. As the system approach equivalency, the differentiation of BC becomes a larger factor.

I disagree completely there.

You are assuming that only "hardcore" gamers will buy the new console near the launch window...and not "new" gamers.

I am, although we have to be clear about the foggy definition of hardcore. I'm thinking gamers with lots of games experience will spend the higher entry cost of launch hardware. These'll be tech enthusiasts and CODers who want the latest and greatest. I do consider it unlikely that someone ponying up substantial amounts at launch will be desperate to play their old games, except when there are few new titles and the old platform still has a relevant catalogue. At that point BC is a valuable convenience, saving early adopters from needing to run their old machines alongside the new. this has been mentioned before in this thread - the value of BC decreases during the life of the console. If the hardware cost is high to include in the beginning, it's probably better to bite the bullet and drop it, or suffer considerable losses like Sony. Then again maybe such drastic measures as including the old hardware are needed to get people to switch in those opening 12 months? Although if it were me, I'd spend that money on securing new games rather than implementing BC in hardware! I'd also go with a plug-in board type thing. PS3 could have had an optional PS2 plug-in board routed through its HDMI. Include it in the box, sell it as an upgrade, and remove it in later years for cost reduction.

 

[function]

If someone wants to play games with all their friends, and that's important to them, then making a clean break with Live across generations can't be compensated for by paying for an exclusive game, spending more on existing marketing, or doing an extra FLOPz.

A clean break across Live makes switching to a competitor easy and potentially much, much cheaper for the gamer.

Getting early adopters to continue with your next system drags others in their network across too, when they're ready to switch in a year or two. If a few $$ on BC can get 1 in 10 (for example) additional gamers to continue with your platform and donate hundreds of $$ in money-for-nothing Live fees then the BC will have paid for itself many times over.

I'm not sure how many ultra hardcore iPhone 4 owners would buy iPhone 5 if they lost the ability to speak to and text none iPhone 5 owners. A somewhat extreme analogy perhaps, but connectivity is a core functionality of Xbox Live and you can't easily compensate for that with a little extra eyecandy or a few $$ saved (people wouldn't spend hundreds on Live subscriptions if you could).

And, on another note, I'd love a game of Xbox 1 Rallisport Challenge 2 right at this very damn minute but my Eggboxen is in a crate somewhere and it'd take up a whole damn shelf under the telly if I decided to man-haul it into the living room. And I'm tired tonight, and simply don't want to move. Thank god for Steam - the ultimate in tired-man convenience gaming.

 

[Cheezdoodles]

None.

That's the point.

The architecture of both ps3 and xb360 is scalable.

Cost of BC? Zero.
_______________

In the case of ps3, Sony literally had to have the ps2 chipset in addition to the ps3 chipset. This added cost did nothing for ps3 games as the architecture was so different to ps3 that the hardware (and cost) was there solely for the purpose of playing old ps2 games.

Xb360 had a purely software solution which did incur one cost element on top of the engineering cost which was licensing the nvidia gpu tech for emulation.
_______________

Contrast this with xb720/ps4 potential designs:

Take the Cell, put 4 of them together, viola, ps4 cpu. Then, grab whatever gpu they like and minimal effort will be needed for emulating RSX.

Take the xenon, put 4 of them together for a total of 12 cores and viola, xb720 cpu. The grab the latest ATI GPU which should have minimal problem emulating ATI's prior chip.

Both efforts are low cost as both companies already own the designs of the cpus.

Nextgen the GPU will be leaned on more for gpgpu calculations etc anyway so the cpu "power" is not a major concern.

The above examples are simplified, but essentially, doing just this simple scaling, would result in suitable CPU's for nextgen. Anything more as far as customizations would be icing on the cake.
__________________

As for whether or not my xb360 can play SNES carts, no ... but I can sell, trade, lend, lease, store, or play them whenever I please. One more thing, you don't see SNES carts topping the charts much these days either ... Nor would I expect it in the future...

I fully realize the DLC model, but there is an assumed responsibility with DLC by the platform holder to "take care" of the digital media which was purchased.

And as I said above, as much as they may WANT to, they will have a very difficult time in selling all the same old digital remakes (or HD versions) AGAIN to a userbase that was just screwed out of their prior purchases.

Best case scenario for them is less purchases by the same number of people. Worst case scenario is less purchases by less people.

And what would this sacrifice be for?

Ur forgetting that there exist cheaper and better CPU solutions than putting 4 cells together

 

[jonabbey]

Are there cheaper and better CPU solutions for them than evolving their current processors? That's not at all clear to me.

If you assume that backwards compatibility has no or low value, then perhaps there is. What do you suggest?

 

[function]

Sony appear to be in a sticky situation, but I don't see why MS would need to use Xenon+ to maintain BC. Ignoring the possibility of pure software BC for a moment, could IBM design MS a powerful, OoOE core that could also run (or support fast emulation of) Xenon without having to include an entire complete copy of Xenon? To me it seems reasonable but then again I'm something of an ignoramus.

In the PC space modern CPUs can run old code without dedicating 100+ mm^2 to including complete copies of older processors. Any additional features you might need to add to the CPU to support BC you could remove from later revisions of the chip if the die area was deemed unacceptable in return for back catalogue sales and Live use at that point in time.

 

[Phil]

Great arguments there Joe, I completely agree.

I do consider it unlikely that someone ponying up substantial amounts at launch will be desperate to play their old games, except when there are few new titles and the old platform still has a relevant catalogue. At that point BC is a valuable convenience, saving early adopters from needing to run their old machines alongside the new. this has been mentioned before in this thread - the value of BC decreases during the life of the console.

I think Joe covered this one pretty much at the beginning. The value of BC is dependant on how much you've invested into the consoles eco-system.

Sure, a customer that hasn't bought much will not have much to lose by either non-BC or switching to another platform. I'd argue though that it's been the aim of both Microsoft and Sony to build up a successfull ecosystem through PSN and live in order to keep a userbase attached to its brand even over a transition to new hardware.

I also believe that the factor of building up a social-network within that eco-system will become a factor. On older platforms, the majority of gamers played multiplayer games through splitscreen modes on a single tv. This generation (and partly last, starting on Xbox) however has enabled players to play with your friends through the network and build up a social network. BC in this case not only helps as a marketing bullet point, but also in smoothening the transition between who upgrades at launch and who might follow a bit later while still being able to play among each other.

IMO this is also something that applies to casual gamers as well. Imagine a casual gamer in his local game store being told that his investment is worthless when progressing to the next console or being told the opposite.

I also agree that the importance of BC is growing due to these reasons and I do agree that the graphics have reached a level where the jump to next console will not be as large as previous generations (due not only to economic constraints but also rising costs/risks in game development). This also makes it easier to go back to play "old games" - rather than comparing going back from PS3 games to PS2 or even PS1 games. I still go back to quite a few PSN games here and there that are not graphically impressive (compared to the newer games) but still work - these games would be just as fun even on newer PS4 hardware and better looking games around.

If the hardware cost is high to include in the beginning, it's probably better to bite the bullet and drop it, or suffer considerable losses like Sony. Then again maybe such drastic measures as including the old hardware are needed to get people to switch in those opening 12 months?

Sony didn't suffer considerable losses because of BC. It's also rather difficult to argue how much of a positive having BC had at the time. Who knows - maybe without BC, the losses would have been even bigger after 5 years within the cycle due to loss in marketshare despite the lower initial BOM figure?

Although if it were me, I'd spend that money on securing new games rather than implementing BC in hardware! I'd also go with a plug-in board type thing.

I'd argue that new games are less of marketing bullet point than having BC at the start of new hardware, especially next generation. Unfortunately, there's also no guarantees that good games sell more than bad games - there's potential that they do, but it's not only dependant on that. Securing new good games, might not be worth much if a large part of your current base is unwilling to 'upgrade' due to costs and lacking BC.

The way I see it: The transition to new hardware is always a huge risk. It takes a huge investment on the hardware side and none of the hardware vendors want to start from scratch. This is why they've been investing in providing an eco-system and IMO, they'd be stupid not to everything possible to make the transition as smooth as possible and give every incentive there it to their customers to upgrade and stay loyal to the eco-system.

You argued that people on consoles have different expectancies than people upgrading on PCs. I disagree. This might have been 2 generations ago (although IMO Sony helped to change this by providing in BC in their consoles over the years), but this is changing. One of the biggest reasons Windows is still the most popular OS is because people don't like change and rather stay on something they know and because they can keep using the programs they are used to and have purchased - despite perhaps arguably better alternatives outthere. It would be stupid of them to ever give up that advantage by starting over fresh. The same applies to consoles: why give up this advantage?

 

[Shifty Geezer]

If someone wants to play games with all their friends, and that's important to them, then making a clean break with Live across generations can't be compensated for by paying for an exclusive game, spending more on existing marketing, or doing an extra FLOPz.

That's a fair call for the convenience. If your XB360 owning mates are playing Gears Horde, and getting an XB3 prevents you joining, that's incentive not to upgrade. That's one reason why a decent cross-hardware platform is needed. Of course, when Freddy 360 is playing COD:MW2 on 360 and Joe Xbox is playing the far better COD:MW3 on XB3, the value of the game wins out often enough I think. Again, I don't think it worth making hardware choices over such an issue as I don't think it'll break you sales. If you can't Live! game with your mates on the same games on XB3 because it lacks BC, you're not going to jump to PS4 where you can't chat with them or compete with gamer scores and you lose your media content. People are already sufficiently tied into Live! to be able to live without BC games I reckon.

A clean break across Live makes switching to a competitor easy and potentially much, much cheaper for the gamer.

I'm not sure how many ultra hardcore iPhone 4 owners would buy iPhone 5 if they lost the ability to speak to and text none iPhone 5 owners.

BC is only about the games, right? Everything else about Live! remains intact whether there is BC in XB3 or not.

And, on another note, I'd love a game of Xbox 1 Rallisport Challenge 2 right at this very damn minute but my Eggboxen is in a crate somewhere and it'd take up a whole damn shelf under the telly if I decided to man-haul it into the living room. And I'm tired tonight, and simply don't want to move. Thank god for Steam - the ultimate in tired-man convenience gaming.

Sure, but it's not worth MS adding full 100% BC to every XB360 just for those rare ocassions some gamers like to go retro. I'd love to pay CON or BGDA on PS3 seeing as no-one's created good successors this gen. And SW Battlefronts2. But I appreciate it's a high cost for a low use, and if the software companies actually delivered next-gen sequels to their top-selling games, we wouldn't be stuck looking back! I can look forwards to when technology is sufficiently advanced to allow software emulation without the cost of hardware BC choices.

Are there cheaper and better CPU solutions for them than evolving their current processors? That's not at all clear to me.

If you assume that backwards compatibility has no or low value, then perhaps there is. What do you suggest?

That's tied in to tech discussion about costs of processors and choices. you relly want to go to the next-gen ahrdware thread to discuss what are the alternative CPU options to Cell++/Xenon++

In the PC space modern CPUs can run old code without dedicating 100+ mm^2 to including complete copies of older processors. Any additional features you might need to add to the CPU to support BC you could remove from later revisions of the chip if the die area was deemed unacceptable in return for back catalogue sales and Live use at that point in time.

AFAIK x86 has a lot of bloat for legacy reasons, and it could run smaller and cooler with a clean design. I don't know how much, but my first Googling found this from 2007:

But with each generation of extensions to the x86 ISA, more and more complexity is added to the chips, and support for the older feature remains to guarantee software compatibility.
"There's no reason whatsoever why the Intel architecture remains so complex," said Simon Crosby, chief technology officer at virtualization software start-up XenSource. "There's no reason why they couldn't ditch 60 percent of the transistors on the chip, most of which are for legacy modes."

I don't know how true that is, or how much still remains with the latest round of Intels, but I think it sufficient to show it's a considerable amount of baggage.

The choice of CPU and how BC can be executed is something of a technical thread, not really fitting here which is just trying to add a value to BC, rather than trying to solve it. But I think the x86 example shows the cost of legacy systems. Sony and MS trying to start their future software platform based on tech designed 7+ years ago without the examples of iOS and Android to compare to is asking for legacy troubles. MS may be safe if Live! game development is sufficiently high-level enough, although that won't solve BC of disk games like COD and Gears which affects your original point of wanting to game with your friends on the older system. The only solution for being able to support all games is iffy emulation or expensive, typically cumbersome legacy hardware.

 

[rekator]

Distilling all the arguments down, IMO this topic is answered with this theoretical question. Given 60 million XB360's by the time XB3 launches, how many of those XB360 owners will refuse to buy XB3 if it isn't BC?

If that's a negligable amount, then the value of BC is shown to be not very high.

For me is simple, If there no BC for physical games and digital purchases, I have no reasons to don't go away, due to the fact I have no longer interest in capitalize my purchase on the Xbox system.
In all case I have to rebuy contents ,I have already buy, so why not go to others? I'm probably go to system with no more hardware attachment, so PC or Onlive and Co.

I can always play my ten year ago game on PC, and with HD mod for some, so console need to do the same.

BC seem to be need for the two first years of gen transition, after is not so important, look at Sony with PS3.

 

[Rodéric]

mmh... My plan is to buy every other generation console, so I can get cheap games from previous gen I missed, along with new ones...
I need backward compatibility !
;p

For the record I'd still play some of my Dreamcast games if both my consoles weren't dead
I also played Dark Forces (10 minutes "for fun", makes you appreciate current games a little more ^^) and Jedi Academy (because it's grood!) the past couple weeks.
(Also XWing 95 and TIE Fighter 95 in the last month...)

 

[bkilian]

None.

That's the point.

The architecture of both ps3 and xb360 is scalable.

Cost of BC? Zero.

Incorrect. Even if the architectures were identical, the cost would not be zero. Considering the cost of maintaining _current_ compatibility with all games when the 360 got redesigned into a single chip was significantly above zero (They actually had to add silicon that emulated the old bus between the parts). Maintaining compatibility with a completely new design with different timings and CPUs entirely, even if they are substantially similar, would cost easily in the millions, BOM, design, and testing costs.

 

[function]

That's a fair call for the convenience. If your XB360 owning mates are playing Gears Horde, and getting an XB3 prevents you joining, that's incentive not to upgrade. That's one reason why a decent cross-hardware platform is needed.

There's also those beacon things now, so people can set up a time in advance for people on their friends list to home in on if they can. Not used it myself, my co-op gaming tends to be coordinated over phone or text and at short notice.

Of course, when Freddy 360 is playing COD:MW2 on 360 and Joe Xbox is playing the far better COD:MW3 on XB3, the value of the game wins out often enough I think.

I guess it depends on the game and the people involved. For me the chance to do co-op wins over playing a new game, but that's because it's harder to arrange what with everyones times commitments these days and that.

Again, I don't think it worth making hardware choices over such an issue as I don't think it'll break you sales. If you can't Live! game with your mates on the same games on XB3 because it lacks BC, you're not going to jump to PS4 where you can't chat with them or compete with gamer scores and you lose your media content. People are already sufficiently tied into Live! to be able to live without BC games I reckon.

Live certainly has growing non-game value, but for keeping in contact you have a lot more options than for network gaming once the platforms are set down. If one of my best Live chums jumped ship to PS4 before any of us had gone Xbox 4 that could tip everyone over ... maybe. I dunno. I guess we'll see if it happens. Maybe we'll just trickle over to the PC.

BC is only about the games, right? Everything else about Live! remains intact whether there is BC in XB3 or not.

I think so, the licenses for stuff like videos seem to work across console<->phone<->PC. But for the early adopter hardcore gamers I'd think gaming would still be the most important factor. Bing, movies, media extension might not be any better on the next Xbox (unless the 360 is intentionally not updated) and so it's hard to see anything none gaming really driving sales of the next system like gaming should, at least very early on.

Sure, but it's not worth MS adding full 100% BC to every XB360 just for those rare ocassions some gamers like to go retro. I'd love to pay CON or BGDA on PS3 seeing as no-one's created good successors this gen. And SW Battlefronts2. But I appreciate it's a high cost for a low use, and if the software companies actually delivered next-gen sequels to their top-selling games, we wouldn't be stuck looking back! I can look forwards to when technology is sufficiently advanced to allow software emulation without the cost of hardware BC choices.

Even the 360's limited support would have been enough for me if it had supported Rallisport Challenge 2! I do wonder when this "diminishing returns" thing combined with online distribution will make it a real bonus to offer old games for longer than a single platform's life. The PC has reached a point where older titles can still shift major amounts on Steam (even without HD remakes), and that's just pure profit from games everyone expected to be long forgotten by now.

I'll leave the technical questions out of this thread as you're quite right, they don't belong here. I'll just clarify that for Xbox I'm thinking about software like Windows Vista/7 WoW, with the only instructions supported in hardware (that you wouldn't otherwise have) being ones that can't be mapped to something else. It might not be perfect but as long as it was "good enough" maybe it could offer reasonable value to MS...

The console transitions from now on are the ones where I'm expecting BC to prove whether it's actually worth a damn or not. If smooth online service transition, continued sales straight to owners and building consumer trust can't make BC a vital part of platform provider strategy then I don't think anything can.

 

[Squilliam]

Wouldn't it be simpler if they just paid developers of the top 20-100 played Xbox 360/PS3 games a fair quantity of cash to port their titles and offer a patch? Afterall once you move down the curve in terms of sales the importance of an individual title drops quite rapidly. In this case they could simply make it a paid thing, I.E. If you want it give us $20-50 and you'll have it for X number of popular titles with more to be added later. It's probably easier for developers to take the source code and recompile and test and it gives them a profit incentive. Alternatively it could be tied into Xbox Live Gold as yet another perk for paying that $60 and it could be an incentive for people to pay for PSN+ for Sony as well.

How much work would it be for say a few interns and testers to port an existing title to a new more powerful platform? Would it cost say more than $1,000,000?

 

[kagemaru]

Incorrect. Even if the architectures were identical, the cost would not be zero. Considering the cost of maintaining _current_ compatibility with all games when the 360 got redesigned into a single chip was significantly above zero (They actually had to add silicon that emulated the old bus between the parts). Maintaining compatibility with a completely new design with different timings and CPUs entirely, even if they are substantially similar, would cost easily in the millions, BOM, design, and testing costs.

That's really interesting, I haven't considered any type of compatibility issues that may surface when switching to the new design.

 

[Shifty Geezer]

Wouldn't it be simpler if they just paid developers of the top 20-100 played Xbox 360/PS3 games a fair quantity of cash to port their titles and offer a patch?

I'd say so. That's kinda what MS did with Halo, wasn't it? They wanted the Halo players who were still playing Halo to move over to 360, so gave them an entrance vector, but didn't care about getting 100% hardware BC. They designed the system without BC and ran an emulation layer of some form on top, tweaked for important programs (at cost to MS to implement) which happily enabled other games as a side effect. Wasn't it a year or so in when MS abandoned miproving BC? They had enough to carry the console forwards and let it drop.

So what titles would be needed to be supported? MS will have clear stats on the most played games. They could take the games that represent, say, 80% of time spent gaming on Live! and port versions of those. That'll be Gears, Halo, MW and BF2, whatever. Now if games are very varied that becomes hard, but the cost of BC is so high that I reckon it's the only option. Same with Sony, although their porting costs may well be substantially higher, and their online play maybe not as important yet.

I also question if you want to provide an upgrade path for every customer from day 1. If there can only be...10 million consoles in the first year due to manufacturing and price/demand, if those 10 million will be happy to buy for new experiences you can just wait for the rest to tire of their existing games and move on when ready. Only if those first 10 million won't buy because they become isolated from their current online gaming is it imperative to get those key games running.

 

[Dominik D]

Wouldn't it be simpler if they just paid developers of the top 20-100 played Xbox 360/PS3 games a fair quantity of cash to port their titles and offer a patch?

Not really. Every company has limited resources. If you decide to port, you can't work on something else (as in: people who are occupied with something anyway have to be devoted to this project). It's probably not worth it for both platform holder as well as developer.

 

[Shifty Geezer]

Not really. Every company has limited resources. If you decide to port, you can't work on something else (as in: people who are occupied with something anyway have to be devoted to this project). It's probably not worth it for both platform holder as well as developer.

If you're going to spend great amounts of money on BC to support a few essential titles, doesn't it make more sense to fund ports than actually go whole-hog on hardware BC? It'll be much cheaper in the long run.

 

[AlphaWolf]

Why not add some content and/or an upgraded texture pack to cover some of the costs through DLC. So you get the game on the new box, but they add some additional goodies (new platform only) to help monetize the transition. Then you might even get someone like me interested in the BC.

 

[Cheezdoodles]

Who buys first gen consoles?

Hardcore gamers. Why do they buy it? To play new, awesome games...

Why do other gamers wait? Cost, or because there isn't enough awesome games out there for you to buy. Come launch day for ps4 and xb3 will I buy it? I doubt it, unless there is some killer app that strikes me as a must have, or if they are ridiculously cheap. Otherwise I'm happy playing on my ps3 until the games I like come out...

I seriously doubt anyone buys a new console in order to play OLD games as a primary function. Particularly at launch! So if we can all agree on that point, my next point is that if the hardcore gamers will buy these consoles anyway, then at the point where others starts buying, there are plenty of games out. Further, u will have had a decade to play the **** out of ur old games! 99,9% will be ready to move on to amazing graphics and whatever.

There is no doubt that BC costs something. Who here would trade say, better performance for the chance to play old games? Personally, at the end of the day, when I buy my new Console, I'm buying it to have a new experience. I could care less about coop with friends, nothing beats getting together in the living room anyway