Graphics and games engineers angered by the delayed OpenGL spec and threatening to adopt Microsoft's DirectX have been asked to hold out a little longer for promised changes. Neil Trevett, president of the cross-industry Khronos Group leading OpenGL, told The Reg on Tuesday that his consortium hoped to start the process of …
I was under the impression...
...that Vista had nixed OpenGL support, so it was left to the driver vendors? Or was I wrong? (I could be, it wouldn't be the first time...).
It's been intimated in some forums that the CAD crowd are not the real culprit but the very GPU guys, who weren't interested in meeting the OpenGL 3.0 driver rewriting effort, disregarding how that would make things far simpler for future iterations.
The official report from the consortium
the official letter from the group is worth reading.
The gist is that the changes wanted by the gaming community would kill most of the other participants (who can't use DirectX anyway) if they did it in a way that was possible in the time allowed. Doing it in a way would not kill the work of lots of important and interested players would take too long.
Games want FAST. Renderfarms want ACCURATE. DirectX discards accuracy for the needs of gaming and Maya et al use software renderers to get accuracy on Windows.
The requirements demanded of them by the gamer entities also raised lots of problems that the gamers didn't care to answer in time or with the needs of others in mind (because MS just hack something that will work and if it works poorly, they'll blame the user demands).
"This is fucking horse [censored]."
What was deemed so awful placed after "horse" that, while "fucking" is apparently acceptable language, this had to be removed?
I'm at a loss for words here. As far as I know, "horse cuntflaps" is not a perjorative in general use and it's about the only thing I can think of that might qualify.
Anyone else got any ideas?
They're going to take stuff out of <=3.0 in 3.1?
Surely only a MAJOR version change is where you take stuff out? Otherwise it all gets bitty and unmanageable.
"This is fucking horse [censored]. ..."
Interesting that [sh*t] is considered objectionable and f*cking isn't.
Paris, 'cause probably even she'd be mystified by that.
Hmm, something fishy afoot???
The ringleaders of the revolt seem to all have IP addresses from the Redmond Washington area apprarently...
Smells like Microsoft is upto it's usual subversive tactics...
I'd be interested
..in the Reg's opinion of 3.0.
@ "Thanks CAD people for screwing us over because you're so damn lazy to update your piece of junk software"
That's absolutely true.
Aut**** a famous piece of CAD software is a bloated monster still intended to be backward compatible with commands and techniques from 20 YEARS ago! This is like a modern game being expected to load Commander Keen save files, simply because 2 people out there haven't bought a piece of software since then. CAD managers are notorious skinflints; mainly because design industry managers are even bigger skinflints. Having said that, Au****** demand an anual extortion, er, LICENCE fee of nearly $7K per year per seat for said cruddy software, update the entire file format each YEAR to be UN-backward compatible (yet the command set IS? WHY?)
We look at the graphics and toolsets of games like Flight Sim X and Crysis and weep as our CAD drawings are stuck with crap stick figures and trees that look like toilet brushes.
We keep paying, though, because we have no CHOICE - thanks Linux! Where's the CAD apps?
So, after 2 years they fail to deliver a signifcant update, instead bring 2.2 which drops the very object model we were after and then ask us to wait again?
I think these people are on another planet where failure is rewarded with people still beliving them; this is just the latest in a long line of ARB failures, with OpenGL2.0 and the time it took to get Framebuffer Object (render to texture) and Vertex buffer object (vertex data in video ram) to name but 3.
Combine this with the 10month communication blackout (despite last year patting themselves on the back for comminicating with the community) and 8 months without letting the community know what was going on is the reaction a surprise?
No, for many this is the straw which broke the camels back and despite using OpenGL for around 8 years I am now firmly commited to D3D and the up and coming D3D11.
Why? Because MS have a history of delivering what they say they will when they say they will with D3D, the ARB has a history of failure. Who would you rather work with?
- Rob Jones
(aka Phantom, aka bobvodka)
No, it wasn't only the CAD people...
Just to say, I have it on good authority that it WASNT the CAD companies who caused this.
Or to put it another way; OpenGL3.0 without the stuff which is going to be removed looks alot like OpenGL|ES. OpenGL and OpenGL|ES were going to converge anyway in the future.
Now, look at the ARB, which major company which has an intrest in the mobile field and HASNT signed off on the one extension which brings us anything like the original OpenGL Longs Peak plans.
Hint; begins with A and end in MD.
"Redmond IPs"; lolol... dude, can I have some of what you are smoking please?
Re: They're going to take stuff out of <=3.0 in 3.1?
Read the source, Luke.
What 3.0 has is a new way of deprecation and lots more deprecated functions.
The changes demanded (long pull) to remove old cruft are in 3.0 deprecated. You can in 3.0 run in "forward compatible" mode where all deprecated functions are disabled.
Good for testing.
3.1 will have the deprecated functions removed.
That is what "deprecated" MEANS.
If you want to play games go and get a games console. You can't seriously expect a developer to drop CAD users who have been using the technology for 20 years and will continue to do so, in favour of gamers who will be screaming for OPENGL version 4 before the next version of Tomb Raider so that they can get a more moist looking breast?
lock yourself in
lose a crystal
Direct X - ruining the gaming industry since 1997.
thanks to sony and nintendo that their is some sort of competition - isnt like Nokia and Apple are going to do it.
why don't they simply branch the code tree?
OpenGLG3.0 for gamers
OpenGLC3.0 to keep the CAD camp happy
seems like an easy solution to me, if you can't keep both camps happy with the same codebase.
All very well saying you'll stick to D3D, but until MS make D3D multi-platform (and you and I both know that's not going to happen) it's simply not an acceptable (or viable) solution for many.
Doesn't it hurt enough?
Continue being stuck depending on MS and bitching about it or take gaming to another OS and write an OpenGL replacement that servces gaming's real needs. Then again maybe people love their suffering to much to really change things...
Why can't they fork it? If you've got two totally different industries wanting different things, why not have two different libraries? Why?
Can none of these chumps see the damage this does to Microsoft-independent game development?
Through all the moaning
I cannot even see what has changed.
Has performance improved? I am sure a few extra techniques may have been added as well.
But, really all you can do is hone in the 3D world, most of it is already understood, and been written about.
As to CAD programs, why don't a few design companies get together and produce a CAD program from the open source software most doesn't require it to be also opensource, you can just share the development cost of the CAD application.
Don't let that GL fool you into seeing a P inserted there.
Blender had to be GPL because of the style of purchase, and you know what it does sort of work, well especially for the entertainment industry. Product design well, you may want special things.
I damn near died when I read the comment about the breast. Good one.
On a side note as a gamer I havent bothered using OpenGL with any of the games that use it due to no noticable benefit and the fact that DX works for what i need. *shrugs* just my two cents.
We don't all have a choice...
... pity us, here in Appleland, who must take whatever Kronos dish out.
That said, I quite like OpenGL — stuff like GLSL has definitely been smarter than the Microsoft equivalent. I don't know anything about version 3 though.
I'm perfectly able to believe that 90% of the protests are Microsoft people trying to just for once get a Microsoft technology to be viewed positively by the development community.
I'll stick with OpenGL and any modern computing platform rather than DX11 and Windows Vista, Windows Vista or, ummm Windows Server 2008. Is there any technical reason why DX10 and 11 won't be available for Windows XP, or is it the usual strongarmed attempt to make users upgrade?
OpenGL and Vista
OpenGL nearly got broken on Vista, however after everyone went to bat for the ARB and bugged MS the situation was resolved (something else to leave a sour taste in our mouths).
OpenGL on Vista is the same as OpenGL on XP; you reqiure drivers from the hardware makers to get the best out of it. If anything the situtation has improved a little with Vista making OpenGL1.4 avaible by default and wrapping it over D3D instead of the pure software 1.1 of previous Windows versions.
RE: Odd censorship?
Comments over. "fucking, horse cuntflaps" is my new all-time favorite insult. I'm going to unleash it at work today and enjoy the shock-and-awe of my co-workers at how I could use such a vulgar term!
Maybe it's time to fork...
At a time where Linux and the Mac are gaining traction on the desktop, OpenGL is becoming more and more important for game developers, as this is the only API that allows them to target all 3 main platforms (Win32, Mac & Linux). OpenGL not being up to par on the gaming front can cause a major issue for computer gaming, as it seems that Mac and Linux are the growing platforms, and the lack of an up to date 3D library can drive them to consoles.
If OpenGL can't follow the gaming market (speed) and the professional market (accuracy & backward compatibility) at the same time, it is maybe time to fork OpenGL in 2, one OpenGL for gaming and one OpenGL for professional applications.
We need OpenGL
I still feel in spite of the problems there is still a need for something like OpenGL providing a cross platform graphics API.
I am dissapointed how things have turned out but it is not over yet. I still like to at least hope this mess can be turned around.
john carmack manages
Trying to please everyone rarely works
The problem is there are two camps that use OpenGL - both with fundamentally
The game developers want access to the latest hardware technology and don't mind re-vamping their rendering engine often to get it. The graphics are often what sells a game and so to get that 'wow' factor game developers need access to the latest feaures asap - they want a stripped down 'lean and mean' OpenGL so that GPU drivers can support new features easier and more quickly. The fact that in 5 years time these current features will be obsolete and newer drivers will have dropped support for them - meaning the game will no longer work - does not bother them.
The CAD developers 1st and foremost want a stable platform that just works, and will continue to be working in 5 years time - the life of a CAD project can be much longer than that, so the original software used *must* still work (often unpatched) even on new hardware with new drivers. Speed is important, but not the main objective - CAD developers spend most of their time developing the user interface and making the CAD functionality better / correct. They want a graphics API that is extensible but rarely loses core functionality.
I'm sure Microsoft and others opposed to the cross platform and 'open' nature of OpenGL are laughing their socks off at the current situation of trying to please everyone and ending up pleasing no-one.
As I see it, the only way both camps will be happy is if OpenGL is forked into two separate libraries - a fast evolving 'gaming' library and a more sedantary 'CAD' library, both of which could be kept small enough for both GPU driver writers and hardware to support, fully accellerated. Most GPU manufacturers already have separate product lines (and often drivers) for gaming / CAD apps so there really shouldn't be any problems having separate OpenGL's.
because depracation worked soooo well for java right? I still see deprecated methods from java 1 (1.1) in the SDK 1.6.
Warning Autocad to update it software sometime in the next 6 months is not likely to help since they had more than 6 months since abandoning Longs Peak. if it would jsut have taken them 6 months, they could have already done it and we would all be happy with the new obj model.
besides they could just use the outdated runtime instead of screwing the rest of us.
@ Beard and @Mason
I am told that the reason the CAD/CAM folks won't accept a fork or branch, is that they don't want to miss out on all the improvements in direct access and what not. they want to have their cake, and mine, and yours too.
this of course indicates that they never plan to upgrade their old clunkers and all this talk about deprecation and eventual removal is bollocks.
"On a side note as a gamer I havent bothered using OpenGL with any of the games that use it due to no noticable benefit and the fact that DX works for what i need. *shrugs* just my two cents."
Uh, you generally don't get a choice.
Half Life 2? DX.
Your only choice to avoid using OGL is not to play games that do it.
Which is strange, given that you don't seem to have a reason to do that.
Well why don't the game creators do that?
NVidia had extensions to do processing in OGL but none of the others (Intel/AMD/etc) wanted to use it. So it couldn't go in.
So for some changes, they won't go in because the people asking for them don't agree on how.
That's not necessarily the ARB's fault, is it.
Perhaps you should have chosen some less generic names, it might give yourself some credability. Your clue in who you work for, is the DX11 reference....
Griping on the forums.
From what I've read on the forums links to various articles about this uproar the biggest gripe has been that the games developers have to write lots of different code so as to find the fastest way of implementing what they want to do depending upon the graphics card due to driver differences.
In fact, it seems that the biggest problem that they have is actually a driver implementation problem rather than an inherent problem with the OpenGL specification itself. i.e. the vendors have broken implementations and OpenGL is getting the flack. (The vendors probably don't actually test the OpenGL code as much as thier DX code. This seems especially true for Intel from the reactions.)
Now, there is another problem, as a previous poster alluded to. You can have fast rendering and correct rendering as long as you pick any one exclusively.
Also, the features that the games designers are clammering for today will be out of fashion in a couple of years time and will not match the hardware then, which means that the Kronos group would get it in the neck whatever they did.
(The skull and cross-bones 'cos specifying a standard is life threatening.)
Well actually it's amusing that you mention extensions. they are the core of the problem. Extensions need not be implemented by a driver/card manufacture because they are not part of the core spec. much of the promise made with LP was to move this functionality into the core spec/API so that all vendors would have to support it to maintian OGL compliance.
Re: Odd censorship?
Puts me in mind of my favorite political candidate, a local grass cutter who has made repeated runs for city council. He is known for his earnest, hand-lettered campaign slogans (frequently written on the back of another candidate's sign). My favorite: "Vote Quentin Brown. No More Bu**shit!"
Heres an idea - dont upgrade your OpenGL then. If you dont want to use new OpenGL functionality when writing your CAD application, write your application with an old version in mind, and tell your users thats what they need to have.
Otherwise, if you want to use the new version, accept that you might have to do things a little differently. Why should CAD hold back the entire industry? No one is saying they cant use deprecated features, but I think if they want to, it should be their problem.
back to the good old days
When the first thing you were presented with when starting a game was the choice between direct3d, opengl or glide. I actually remember glide being the best looking of the options too at the time :)
Although i can't ever see games going to the expense of developing 2 different rendering engines again with the complexity of modern graphics.
EXT_ARB are core extensions. You don't neet to have a driver producing them or hardware to do it to be "OGL compliant".
EXT_NVIDIA are nvidia's personal extensions.
EXT_ATI are ATI's personal extensions.
You're talking about the last two. I'm talking about the first.
"because depracation worked soooo well for java right? I still see deprecated methods from java 1 (1.1) in the SDK 1.6."
a) it's deprecation.
b) well that's why they are working so hard to NOT make that mistake in the deprecation of Java
c) DirectX10 includes rendering libs for DX9 which includes libs for DX8, ... But you orgasm over that
d) The alternative is to just drop it. Which causes you to HAVE to rewrite all your stuff NOW whilst your users migrate to someone else. Which is nice.
"As I see it, the only way both camps will be happy is if OpenGL is forked into two separate libraries - a fast evolving 'gaming' library and a more sedantary 'CAD' library, both of which could be kept small enough for both GPU driver writers and hardware to support, fully accellerated."
That (if I'm reading the report right) is precisely what OGL 3.0 has instituted. There will be sections that are separate and you *may* if you want implement these sections, but you don't *have* to. I think they're calling them profiles.
So you'd have a gaming driver that ONLY does the core stuff plus the GAMING profile and NOT the deprecated stuff. Or a workstation renderfarm using the core, deprecated stuff (whilst having time to work on removing the deprecated items in a controlled manner) and the RENDERFARM profile.
NVidia may implement all of them so that they have the same driver for renderfarms as gamers.
Erm, this is my real name and I included the akas at the bottom of my first post because I stand by those words posted in both forums linked in the OP. And hey, at least I don't hide way when making comments :)
As to who I work for; Zoe Mode, UK
My first project was on the PS2 and at home I spent the last 8 years working with OpenGL before swapping.
Nothing todo with MS, but thanks for playing.
As for everyone else making noise, the majority of them are people, like myself, who have been around for some time on the forums (in the case of the OpenGL forum longer than me) so are hardly MS "subversive tactics", but feel free to carry on living in your paranoia :)
what I want to know is
if XP's still more favorable than Vista, when will we see something for XP that enables the DX10 functions of my Nvidia 8800?
Indeed OpenGL 3.0 does attempt to do this - and still no-one is happy.
Games developers don't want the 'dross' of all the CAD functionality and the CAD developers won't be happy when their 'depreciated' functions are dropped (as that means hardware-accelerated driver support for those functions will probably disappear).
By forking into two completely separate API's that are both small enough for driver / hardware vendors to support and develop hardware-accelerated you should make both camps happy.
If alternative platforms are to flourish they will need the support and development of both the games industry and professional CAD packages. I don't see that happening without libraries like OpenGL.
Re: what I want to know is
"if XP's still more favorable than Vista, when will we see something for XP that enables the DX10 functions of my Nvidia 8800?"
From Microsoft? Never. Or near enough.
However, it has been cracked by those *nasty* hackers so that XP will install and run DirectX10.
firstly, i won't deign to dirty myself by responding to your remark on spelling.
secondly, your right Java did mess up deprecation by failing to remove the bad methods after the next version. Based on the ARBs statements about CAD stuff needing to last longer, I am pretty sure they will not be willing to deal with the deprecated methods being removed in version 3.1, 3.2, 5.9, or even fricken 20.5 etc.
thirdly, I give not a wick about DX, but I will say, their backward compatibility is not stopping them from moving forward the way they want, while the status quo for OGL prevents the use of the new Object model.
you're quite correct, the only alternative is to drop the old model. The conservatives are going to keep pushing to maintain the status quo no matter how many revisions those methods are deprecated, so unless they receive proper impetus to update their obsolete code, they will just keep blocking the way forward for the rest of us.
I don't trust Microsoft
So, out of the box, Vista and XP don't support OpenGL 2.
I can go and download drivers. Sometimes they're not "signed" by Microsoft.
And I have to, because even the cheapest graphics software is using OpenGL 2
Microsoft don't appear to be interested in graphics work, even though they've bought Caligari. I wonder what Truespace will be like in a couple of years.
I am, however, hearing good reports on WINE and Linux, for the specific software I'm using.
I haven't laughed that hard since I was a little girl.
Wait, it's only the appendix that states what's deprecated. You go through the whole 3.0 spec, and at the end, it's "Oh, BTW, you probably don't wanna use these 40 things". I'm surprised that they didn't keep the dep list in their basement guarded by cheetahs.
"I'm surprised that they didn't keep the dep list in their basement guarded by cheetahs."
A: I eventually had to go down to the cellar.
C: That's the display department.
A: ..with a torch.
C: Ah. The lights had probably gone.
A: So had the stairs!
C: But the plans *were* on display.
A: On display? I eventually found them at the back of an old filing cabinet inside a locked lavatory with a sign on the door that said "beware of the leopard".