Why does this Areo thing or compositing thing use D3D? Aren’t most windowed 3D programs using GL?
Personally, I think that this problem would only serve to encourage developers to switch to D3D. End users of software may complain to ISV which would also have the effect of developers switching to D3D.
Given enough time, GL’s share will erode, very much like the egyptian pyramids are eroding.
I don’t know what the case of mobiles is. GL ES also has competition from MS.
Whay some accept the idea of having DirectX ported to linux, and at the same time complain about it in its native platform?
And it will only hurts windows platform, that OpenGL sucks under it so many softare andf game developers will have less interest in supporting Windows. This is because big software like Maya, softimage, Lightwave, and game Quake’s, Doom 3, Unreal and so on maintain different paltforms: Win/Lin/Mac so they have to make at least two ports with OpenGL and one port for Windows. Then they cannot stop support of OpenGL otherwise they are going to loose two platforms from the market share. You see. so the only doomed platform where openGL will be poor and hence will gain a bad reputation however is Microsoft Windows Vista.
Today I tried the (LDDM) ATI Vista drivers that just came out. And I’m sad to report that there is no genuine OpenGL support in them (allthough I expected as much). Just the OpenGL-D3D Translator. For anyone interested, I summorized my findings with Vista and OpenGL on http://www.OrganicVectory.nl
We must think logically. If opengl is strong (realy it is strong), so we should not worry about this matter.
Opengl is on Linux and … , so we can develop on it.
Last, we should not forget that DirectX is good and it is for game programming (Direct3D).
Bye.
If we think logically, there are four great things for game industry:
OpenGL (OpenGL ES)
OpenAL
OpenML
OpenVG
This stuff is open and free. It makes possible to make games for ANY platform, not just Windows. Btw, Sony didn’t accidentally choose OpenGL standard for PS3.
What does this mean? It means that game and application developers will see opportunity to use these APIs for development, since it makes sense:
Larger desktop market: Windows, Linux, Mac…
Less work - investments ($)
Ability to enter console PS3, PSP market
Ability to port to mobile devices
If you think logically, this is exactly what Microsoft doesn’t want. So there is actually no problem (technical or otherwise) to have OpenGL on Windows Vista – the only “problem” is because Microsoft doesn’t want it.
Think about it:
What if tomorrow all systems could have hi quality games (free Linux, Mac)?
What if Sony PS3 could run games for PC (written in OpenGL/AL/ML technology?
You will find answer who and why obstruct OpenGL standard. And this is not Microsoft bashing. It is common sense.
I don’t know about MS monopolistic plan, but I’d like to have some more info about the ( eventual ) techical problems that might rise in having full openGL support through aeroglass.
Is it true that only one API at a time can retain control of the graphic subsystem? If so, is it feasable for the HW vendors/MS to build a virtualization layer that has the exclusive control of it and lets D3D and OGL work toghether?
Does OGL actually have all the features needed to run aeroglass? ( i.e. frame-buffer partitioning, video memory virtualization…). Are they ARB features yet?
As OGL dirvers are pretty different from D3D ones, not to mention the new Longhorn driver model, and the extension fetching method isn’t that clean, safe and comfortable, is it possible for the ARB to issue a new OGL base implementation that better fits MS models?
I think that the compatibility option in the program ejecutable properties is a good solution.
“[x] Run in compatibility mode” System : [Window XP-OpenGL]
“[x] Run in OpenGL ICD fullscreen mode”
“[x] Force run OpenGL ICD and DWM in my own risk”
This maintain the Aeroglass composite desktop while no OpenGL fullscreen application is running. When the game in running the system switch off all the D3D stuff.
This can be only a patch, Of course the best solution is that ISV and IHV work with microsoft to write a ICD that substitute MSOGL and be compatible with the Aeroglass composite desktop.
I haven’t read all the replies, but it seems that M$ wants to strangle OpenGL out of existance not only to compete with platforms like Playstation but perhaps also to shoot for making linux less of a desktop OS choice. How many fewer people are going to switch to linux if games are not ported over when they don’t make the money they’d also make on Windows. What I mean is, the main development would be for windows and secondary for linux. I just hope companies start programming directly for OpenGL in linux. I prefer it as a gaming platform anyhow. I get better framerates in Gentoo than I do in WinXP Pro.
well, guess this is all a good thing.
look at it this way, the serious stuff (maya, catia & friends are already moving to Linux, which sports proper OpenGL support. furthermore, Xorg is going in the direction of running over the OpenGL layer, turning all 2D operations into OpenGL things that do the same thing, and things like gtk are getting build on top of cairo, which in turn is build on those OpenGL things.
What will remain on Windows is the only things that should have been there, that is, games. windows will become a playing OS…
Of course, the PS3 using linux may also get all those games easily ported on desktop linux boxes…
Is it true that only one API at a time can retain control of the graphic subsystem? If so, is it feasable for the HW vendors/MS to build a virtualization layer that has the exclusive control of it and lets D3D and OGL work toghether?
Not really. Technically, only one API can be active on a single window, but you can definately spawn, even on XP, a D3D app while running a GL app.
Obviously, you can only have one fullscreen app running at any one time.
Does OGL actually have all the features needed to run aeroglass? ( i.e. frame-buffer partitioning, video memory virtualization…). Are they ARB features yet?
It all depends on how it is done. Apple, for example, is able to make it work OK. So it certainly is possible. But, then again, Apple uses OpenGL the way Microsoft uses D3D: they provide a non-trivial portion of the code themselves and have driver developers develop to an internal API that can be anything. OpenGL on Windows doesn’t work this way.
As OGL dirvers are pretty different from D3D ones, not to mention the new Longhorn driver model, and the extension fetching method isn’t that clean, safe and comfortable, is it possible for the ARB to issue a new OGL base implementation that better fits MS models?
That’s not really the problem. The issue, as Evan Hart (ATi) put it in the link found on the second page, is that theoretically an ICD can interface with the desktop compositing stuff. However, Microsoft has yet to divulge the necessary information to IHVs to allow them to make this happen. Until they do (which they may hold back for a number of legitmate reasons, the most important of which is that Vista isn’t finished yet, so it may change and break stuff), or until Vista is released and the IHVs are able to deprocess what the compositing needs and replicate it in the ICD, the issue will persist.
I find it very interesting that the average gamer will be able to buy a $49.99 OpenGL based game, run that game on a $219.99 consumer level graphics card and continue to have the full Aeroglass desktop experience with full OpenGL acceleration, while the professional who purchaces a $4999.99 engineering package that requires a $2199.99 professional graphics card will be stuck with a Windows XP themed desktop.
FWIW, high-end engineering applications have been moving FROM *nix TO Windows over the past 10 years – not the other way around. These companies have moved to that platform because of market demands. Those same markets demand performance. It is yet to be seen if loss of the Aeroglass desktop compositing engine will be accepted. Considering that D3D does not support key functionality that is an intrisic part of OpenGL, I’m not sure the market will have a choice.
One other point that I have not seen in these posts is that while the GENERIC OpenGL 1.1 software rasterizer currently delivered with Windows (through XP) is slow – there is one extremely useful aspect of that pixelformat – the software rasterizer is very stable and isolates the application from all hardware anomalies. Nearly every high-end application I have used offers an option to select the GENERIC OpenGL pathway and in many cases (e.g., laptops, legacy hardware, etc.) this is the ONLY way to run the application successfully. If “OpenGL on D3D” is the default pixelformat for Vista, will there continue to be a GENERIC (software only) pixelformat? If not, then that could render many applications unusable on Vista (pun intended ).
After reading all the post here, I think the only thing we can be done is up to the ARB and SGI. Yes it’s their responsibility to protect the OpenGL quality and assure it’s not degraded in any way.
The ARB and SGI should only allow or confirm OpneGL implementation for those who are members of the ARB. This way MS will not be able to provide a crappy implentation and if they don’t have OGL then they will lose many titles on their Vista.
I don’t understand why all that action and intense regarding what MS gonna do. It’s their platform and they can do whatever they want. They are not saying we are going to stop OpenGL dev, but it’s for definite that D3D is more appropriate to the platform and it’s the only way that allow low level frame buffer control under Windows.
Stop all that panic. My first game will be both OpenGL for Linux or even Mac, who cares, and D3D for Windows. Just I replace the calls and re-program the sahders, what’s a big deal. It’s enjoyable experience to master both APIs.
Get real and stop hating MS, it’s the best and greatest software company ever happned to the computer industry.
I don’t know how layered support for OpenGL (and limited implementation at that) is going to fly with the design and simulation application vendors.
Autodesk, Solidworks and Bentely rely heavily on OpenGL for their 3D modelling suites and in some cases demand a full ICD. Hardware graphics acceleration is the single-most important performance factor and I suspect they’ve stayed away from DirectX for it’s lack of support on complex surface geometry.
Given that these programs run exclusively on Windows (no OSX or Linux support amoung them), a stunted GUI or degraded performance seems like a bad choice to hand indirectly to their clients.
I’m a Solidworks operator by trade ( http://www.solidworks.com ) and since their first release in 1995, your card and drivers fully implement OpenGL, you wait…and wait…and wait.
What’s more, these companies are developing their apps to support the new shaders in OpenGL 2.0 and are paralleling their development for OpenGL’s future.
I’ve heard the arguments for the business sense of Microsofts decision, but limiting support for an open standard like OGL seems antithetical to the philosophy behind their push for XML and opens the door for OS’s with a more consistent development philosophy to court these big-ticket customers.
Many people in the Autodesk discussion forums wonder casually when the first Mac release for AutoCAD will show up again. With the kinds of pressures Microsoft’s putting on developers and the attitude that all applications want to be Xbox games or Office, that day may be dawning directly.
I will take this news to Solidworks and Autodesk (though I’m sure they’ve gotten wind of it, already) and voice my support for full OpenGL implementation. Autodesk did a billion dollars in sales last year. Solidworks has two hundred fifty thousand installed seats. I’m sure Microsoft doesn’t want to share this exclusive and lucrative market segment.
-and this last part is just a personal plea. Given the gross technical illiteracy in the US legislature and judiciary the neolithicly crude laws proposed thus far to govern file sharing and intellectual property rights, the whole industry would be better served if the courts were left out of our efforts. It’s the only way to make sure Microsoft’s powerful lobbyists don’t decide this for us.
After reading all the post here, I think the only thing we can be done is up to the ARB and SGI.
Then you clearly did not actually read the posts here. The ARB can’t fix this problem, and neither can SGI. Only IHVs with the cooperation (and possibly without) of Microsoft can.
The problem is caused by Microsoft not (yet) having divulged information to IHV’s about how windows communicate for the purposes of desktop compositing. With that information in hand, whether Microsoft gives it to them or they figure out what Vista is doing, they will be able to create ICDs that work with desktop compositing.
This way MS will not be able to provide a crappy implentation and if they don’t have OGL then they will lose many titles on their Vista.
I find it silly that you’re calling Microsoft’s new GL implementation “crappy”, when it is, in fact far superior to their last one, which is frozen at 1.1, and also doesn’t offer up extensions. Oh, and it was software so it was unusably slow. At least they’ll offer up a reasonable level of functionality this time around. It may not have shaders, but it’ll have VBOs, multitexturing, and stuff. And it’ll have decent performance. Not up to D3D, but that’s not really the point.
The generic implentation has always been a fall-back implementation anyway. It was never intended to be used as the primary implementation on shipping products. And the fact that this implementation is clearly superior to their previous one is cause for celebration.
Both you and korval are both being clowns, you both advocate Microsoft’s current OpenGL Vista strategy as positive, when the majority of the people here making articulate posts do not. The current implementation on Vista cripples OpenGL 1.5 and up. Anyone using gl 1.5, gl 2.0, shaders and extensions is crippled on vista. Forcing end-users to install an ISV driver that disables Aero everytime you need to switch to a high end OpenGL app is a huge deterrent to those apps. People who have already invested their time in developing their apps past openGL 1.4 on windows are going to get screwed by this, myself included. Anyone developing windows using Aero will not be able to take advantage of OpenGL past 1.4. Not knowing whether the problem lies in MS’s incomplete wrapper or the direct3D driver will cause further problems. Given what we know, this is not a good thing, so please stop advocating that it is based on unproven speculation.
As for those of you bashing MS, you’ve already admitted defeat on having proper OpenGL on Vista and might as well move to a new platform now. Microsoft as far as I know is not dependent on any IHV or software developer using OpenGL, if anyone complains and leaves, another competitor waiting in the wings will no sooner take its place and conform to the new MS standard. It is in our best interest that we politely ask Microsoft to provide us with a descent solution, one that allows OpenGL and Areo to co-exist peacefully without compromising the other. We can only hope that the current situation is a temporary measure and that OpenGL will continue to survive on windows, otherwise we must either conform to Direct3D or leave the platform. I regrettably would choose the latter.