"The problem stems from a catastrophic decision made at Microsoft: not giving DirectX 10 to Windows XP users."
- Brad Wardell, CEO of strategy game specialist Stardock, explains why he think strategy games have been held back over the last few years.
The Stardock founder says in the company's 2012 report [PDF] that strategy games in recent times have been a mixed blessing.
"There have been some great titles released," he says, "but the innovation in strategy games has been diminishing. This is not the result of a lack of game design or inventive thinking."
He says it comes down to Microsoft's decision not to supply Windows XP users with DirectX 10 as standard, instead leaving them with DirectX 9 support.
"As a corollary, Microsoft continuing to sell 32-bit versions of Windows well after the hardware stopped being natively 32-bit has held back PC game development immensely," he continues.
"Game developers have been stuck with DirectX 9 and 2GB of memory for the past decade. While this hasn't harmed first person shooters (they only have to manage a handful of objects at once), it has been poisonous to other genres."
Meanwhile, DirectX 11 allows studios to go crazy with the shader anti-aliasing, says Wardell, and lowers the development capability requirements on having a multi-core-based simulation.
"There are whole classes of games waiting to be made that require these kinds of advances," he adds. "Luckily, after a decade-long wait, we are nearing critical mass. The days of games supporting 32-bit OSes is, thankfully, coming to an end. DirectX 10 as a minimum requirement has also arrived."
Civ5 has DX10 doesn't it? I can't think of anything significant that would change with DX9. Considering the vast majority of strategy games are isometric I think the value of DX10/11 is overstated here.
So not having graphical "flash", like shader AA is what is stopping innovation? There are plenty of game releases that have 32bit and 64bit executables using DX9/10/11 - selectable depending on what type of system a user is running.
What he's saying is that although that is true, the developer needs to develop to the lowest common denominator (32 bit DX9).
Upping the minimum requirement for RAM beyond 2 GB (only doable in 64 bit) will diminish the need to use the swap for AI calculations, which strategy games rely on more than FPS, making them faster or more interesting.
Am I missing something? To the best of my knowledge, the 64-bit versions of Windows XP were never really popular, so if you're targeting XP, you'd be facing the memory limitations of a 32-bit OS anyway.
He probably should've qualified it as "graphical innovation" if (hopefully) that's what he meant. It certainly stifles graphical innovation. But not gameplay innovation.
Although it does feel like (through intention or incompetence) Microsoft has made some rather anti-PC decisions as of late.
This is all growing pains... It too shall pass. x64 is almost the standard and there's always OpenGL if you don't want to be locked to a platform API based on your OS version.
I guess there is no good strategy games on consoles, which has a lower specification than certain desktop PC. Placing the blame on DX9 is a cop out, how about taking responsibility and using ingenuity to overcome the limitation. Its always easy to blame, but how does that solve the problem at hand.
So would you tell someone flying a pre-WW1 plane that the reason they can't break the speed of sound is a lack of ingenuity or innovation? I'd be careful before second-guessing Stardock - they've been very good at what they do for longer than most people have been in the games industry. It's also easy for non-strategy devs to not understand the nature of what the games involve - these are very different beasts technically than most other games out there.
Also - Civ Rev and RUSE both functioned well on console, but both clearly showed the limitations of the hardware in terms of map sizes and unit limits.
There is a marked difference between being a pilot and the engineer that design the plane. Yes...Stardock is just a user of DX9 like all of us, but the pilot nor the engineer that design the plane don't have the convenience of rapid iteration that software design could lend itself to.
Seems to be a mix of real hard limitations, and poor technical skill on Stardock's part. I'd like to see Brad Wardell's strategy game concept that absolutely requires more than 2 GB memory and a DX11 class GPU, but isn't just GalCiv or Elemental with more units or better graphics, and can be done with Stardock level budgets and team size.
As far as actual graphics features go "DirectX 11 allows studios to go crazy with the shader anti-aliasing" doesn't really sound like it's holding back innovation, graphical or otherwise. There are cool things that you might be able to do with compute. If they really want to make DX11 only games that use a lot of RAM they should move to next gen consoles.
Correct me if I'm wrong, but... I thought the point of DirectX 11 was that the API was smart enough to downgrade to DirectX9 environment if it needed to without having to make a separate binary for it?
It is, but XP still has a huge install base and doesn't support DX11 at all. The problem isn't that people have DX9 hardware and Windows 7, it's people with DX10 or DX11 class GPUs that are still running Windows XP.
I find Windows XP far too repugnant to leave on any of my machines, besides the fact that full support for XP ended years ago. It's surprising to hear that so many people continue to use a computer with an OS from 2001.
I'm actually running Windows 8 now, and *ducks* I really enjoy it. It plays ball with DX11 cards and 12-16GB of RAM in such a way that it crushes the performance I experienced on Windows 7 under similar or identical hardware configs. Maybe I just got lucky.
In any case, we're rolling into a DX11-class era of consoles, so it's prime time for PC gaming to shift over to that tier en masse.
Although, as others have mentioned, I would never blame a generation of 3D API for one's inability to innovate in a genre. Presentation-wise, I get it, but that shouldn't be completely to blame for stagnation.
Some people, especially with the kind of downturn most industries have had, in trying to scale back pay and benefits while at the same time increasing their prices on many staple goods (no doubt at least some indirectly tied to fuel prices within 6 or fewer degrees of separation), probably just didn't or don't have the kind of funds to reasonably justify upgrading more regularly.
I for one usually end up waiting a few years at a time and then try to get as much as I can for my money by starting fresh.
Under the hood, Windows 8 did improve a lot of things; partly as necessity to fit in a tablet environment, but many such concessions can (in)directly benefit desktop systems and notebooks as well. I personally can't get past the default interface, however, which is why I'm likely to skip Windows 8 altogether.
That said, if DX9 is so stifling, why didn't they consider OpenGL for handling the graphics? Unless their test beds are all on AMD Radeon cards or something... lord knows I could tell stories of how AMD seemed intentionally crippled at handling OpenGL for years and years.
I hear ya. I didn't expect to like Win 8 or see any performance gains. In fact, I expected bloated crap. But it's been zippy and relatively problem-free.
My first test box for Windows 8 was a machine I originally built as a test box for Vista. A Core 2 Duo 6400 system on a Intel board. Over time it got more RAM (max of 6 GB of PC6400) and a low-end video card. It became a Vista 64 system (requiring a BIOS update), then a Windows 7 64-bit system, and finally a Windows 8 system.
The performance that was extracted from that aging system over time was a major turnaround from the history of Windows. It would have been laughable to take a machine built for Win95 and try to grow it into a decent Win2K machine. It made far more sense to treat a major new Windows release as an excuse to build a new machine.
That has changed. It used to be that the improvements in new machines were obvious to nearly everyone. Now it's become a thing that requires specific needs like gaming to find a big difference. Good for OS upgrades but bad for hardware sales. If it weren't for the objections to the UI (it doesn't bother me, personally, but a lot of people can't deal with it) I'd be recommending a Win8 upgrade for everyone still running Vista, especially when they were going for only $40. It simply uses the existing hardware more effectively.
That is a big change. In the past I've almost always advised clients that the best way to get the latest OS was with a new machine. It just wasn't worth the investment in upgrading an old machine compared to what the same money would get in terms of a new system. But we finally hit a plateau where the base performance needs for a good desktop system have become static. It will take some innovation to deliver the reason for non-gamers or other sorts of power users to get state of the art hardware before it becomes cheap.
If anything, I'd think it's the focus on graphics in strategy games that is holding things back. Worlds are smaller, game designs simpler, because so much of the budget and computer power is going to the graphics.
Heck, one of the deepest strategy games I've played was written years ago in Qbasic for the PC. Sword of Aragon. The city development was sort of simplistic, but the tactical combat was amazing.
"If anything, I'd think it's the focus on graphics in strategy games that is holding things back. Worlds are smaller, game designs simpler, because so much of the budget and computer power is going to the graphics."
/\ this applies to every genre of game these days.
This is utter nonsense. I cannot believe Wardell would make such a specious argument. Who are the people using Windows XP at this point? I'm pretty sure it can be shown the bulk of them are corporate desktops and most of the remainder consumers of the sort unlikely to buy a strategy game or still happy to play the games they bought many years ago. In terms of the hardware, I'd expect to find that nearly all of the XP systems on 64-bit capable machines are among the locked down business systems and were never going o be potential customers for a game, nor do they have graphic hardware sufficient to make DX10 or DX11 use worthwhile, considering how recently Intel IGA became capable of properly supporting those APIs with anything remotely like decent performance.
Let us recall one of the major reason Vista had such a bad launch. Intel, with a severe deficit in IGA functionality and performance, demanded that Microsoft approve the IGA as capable of running Aero despite it performing many DX9 functions in software rather than on the GPU. This caused many millions of machines that barely had DX8 capability to be treated as DX9 systems. Is it any surprise performance was awful? I'm kind of surprised nobody put together a class action suit about this. As it stands, a vast number of machines that are supposedly DX9 systems need a video board installed, where possible, to even be potential hosts for a DX10 game.
The sales of Windows 7 licenses passed 600 million last June. (This includes a large number of corporate machines that were reimaged to XP but most of those will be back to Win7 by now or replaced in the normal three year cycle.) Well over half of those systems are 64-bit just by virtue of shipping with 4 GB or greater RAM. Leaving aside the tens of millions of systems running Vista or Win8, reaching just 1% of the machines out there capable of running a 64-bit DX10/11 would mean sales in the millions of units.
Just 1%.
A hit game is one that reaches a fraction of a percent of the capable machines. And despite the slump in PC sales that base of machines is growing daily. 32-bit systems are now quite rare in the consumer retail channel. Even the crappiest Intel IGA for desktops now shipping is a real DX10 GPU. The soon to launch Haswell CPUs will up the ante to DX11.1 for bottom of the barrel GPUs in desktop systems. The installed base of those machines should be into the millions by the end of the year. Meanwhile, everything AMD has sold with a GPU has been DX11 for quite a while.
This is like a music company blaming Honda because people driving twenty year old Civics are unlikely to buy a new Cd or a sound system capable of playing it.
What is this nonsense with graphics again? CEO of Crytec blabbering that "visuals are 60% of the game", now... wait, really? Stardock CEO? Claiming that DX9 held back innovation? I expected better from you, Stardock...
If anything, it's the pathologic focus on graphics that holds innovation back. Calm down, guys, 5 years ago we've achieved graphics more realistic than most people's imagination, you can now start making GAMES again, instead of animated pseudomovies...
Upping the minimum requirement for RAM beyond 2 GB (only doable in 64 bit) will diminish the need to use the swap for AI calculations, which strategy games rely on more than FPS, making them faster or more interesting.
Although it does feel like (through intention or incompetence) Microsoft has made some rather anti-PC decisions as of late.
Also - Civ Rev and RUSE both functioned well on console, but both clearly showed the limitations of the hardware in terms of map sizes and unit limits.
As far as actual graphics features go "DirectX 11 allows studios to go crazy with the shader anti-aliasing" doesn't really sound like it's holding back innovation, graphical or otherwise. There are cool things that you might be able to do with compute. If they really want to make DX11 only games that use a lot of RAM they should move to next gen consoles.
I'm actually running Windows 8 now, and *ducks* I really enjoy it. It plays ball with DX11 cards and 12-16GB of RAM in such a way that it crushes the performance I experienced on Windows 7 under similar or identical hardware configs. Maybe I just got lucky.
In any case, we're rolling into a DX11-class era of consoles, so it's prime time for PC gaming to shift over to that tier en masse.
Although, as others have mentioned, I would never blame a generation of 3D API for one's inability to innovate in a genre. Presentation-wise, I get it, but that shouldn't be completely to blame for stagnation.
I for one usually end up waiting a few years at a time and then try to get as much as I can for my money by starting fresh.
Under the hood, Windows 8 did improve a lot of things; partly as necessity to fit in a tablet environment, but many such concessions can (in)directly benefit desktop systems and notebooks as well. I personally can't get past the default interface, however, which is why I'm likely to skip Windows 8 altogether.
That said, if DX9 is so stifling, why didn't they consider OpenGL for handling the graphics? Unless their test beds are all on AMD Radeon cards or something... lord knows I could tell stories of how AMD seemed intentionally crippled at handling OpenGL for years and years.
My first test box for Windows 8 was a machine I originally built as a test box for Vista. A Core 2 Duo 6400 system on a Intel board. Over time it got more RAM (max of 6 GB of PC6400) and a low-end video card. It became a Vista 64 system (requiring a BIOS update), then a Windows 7 64-bit system, and finally a Windows 8 system.
The performance that was extracted from that aging system over time was a major turnaround from the history of Windows. It would have been laughable to take a machine built for Win95 and try to grow it into a decent Win2K machine. It made far more sense to treat a major new Windows release as an excuse to build a new machine.
That has changed. It used to be that the improvements in new machines were obvious to nearly everyone. Now it's become a thing that requires specific needs like gaming to find a big difference. Good for OS upgrades but bad for hardware sales. If it weren't for the objections to the UI (it doesn't bother me, personally, but a lot of people can't deal with it) I'd be recommending a Win8 upgrade for everyone still running Vista, especially when they were going for only $40. It simply uses the existing hardware more effectively.
That is a big change. In the past I've almost always advised clients that the best way to get the latest OS was with a new machine. It just wasn't worth the investment in upgrading an old machine compared to what the same money would get in terms of a new system. But we finally hit a plateau where the base performance needs for a good desktop system have become static. It will take some innovation to deliver the reason for non-gamers or other sorts of power users to get state of the art hardware before it becomes cheap.
You’ve had access to OpenGL and yet ignored it, furthering the position you’re in. Good Job.
Heck, one of the deepest strategy games I've played was written years ago in Qbasic for the PC. Sword of Aragon. The city development was sort of simplistic, but the tactical combat was amazing.
/\ this applies to every genre of game these days.
DX9 has not hurt gameplay progression at all, because there are plenty of indie games making up for the AAA rehashes we're getting these days.
Let us recall one of the major reason Vista had such a bad launch. Intel, with a severe deficit in IGA functionality and performance, demanded that Microsoft approve the IGA as capable of running Aero despite it performing many DX9 functions in software rather than on the GPU. This caused many millions of machines that barely had DX8 capability to be treated as DX9 systems. Is it any surprise performance was awful? I'm kind of surprised nobody put together a class action suit about this. As it stands, a vast number of machines that are supposedly DX9 systems need a video board installed, where possible, to even be potential hosts for a DX10 game.
The sales of Windows 7 licenses passed 600 million last June. (This includes a large number of corporate machines that were reimaged to XP but most of those will be back to Win7 by now or replaced in the normal three year cycle.) Well over half of those systems are 64-bit just by virtue of shipping with 4 GB or greater RAM. Leaving aside the tens of millions of systems running Vista or Win8, reaching just 1% of the machines out there capable of running a 64-bit DX10/11 would mean sales in the millions of units.
Just 1%.
A hit game is one that reaches a fraction of a percent of the capable machines. And despite the slump in PC sales that base of machines is growing daily. 32-bit systems are now quite rare in the consumer retail channel. Even the crappiest Intel IGA for desktops now shipping is a real DX10 GPU. The soon to launch Haswell CPUs will up the ante to DX11.1 for bottom of the barrel GPUs in desktop systems. The installed base of those machines should be into the millions by the end of the year. Meanwhile, everything AMD has sold with a GPU has been DX11 for quite a while.
This is like a music company blaming Honda because people driving twenty year old Civics are unlikely to buy a new Cd or a sound system capable of playing it.
If anything, it's the pathologic focus on graphics that holds innovation back. Calm down, guys, 5 years ago we've achieved graphics more realistic than most people's imagination, you can now start making GAMES again, instead of animated pseudomovies...