Discussion in 'Article Discussion' started by Gareth Halfacree, 23 May 2016.
Well, that backfired.
Fairly obvious that something like that was going to happen.
VR is going to be a knock-down-drag-out fight between competing 'standards', which will only get worse when other players (Sony, anyone else) enter the field properly.
As users, we can hope that an open standard wins the day, with competing products supporting it... but I can't help but think that it'll be a three-way slugfest between OculusVR, Vive and PlayStationVR, each carving their empires out of the VR userbase.
Despite Valve supporting HTC, they don't appear to really have the liquid capital for a prolonged and expensive fight with Sony or Facebook (particularly Facebook, given the sheer weight of public opinion that they can influence, being the currently dominant social network)...
Valve can easily blackmail games to support VIVE. They are the biggest PC store for games
I miss the days when manufacturers got together and made cooperative standards.
Now we can't even get a standard for keyboard LED control, much less a VR headset. Everyone wants to lock users into using one brand forever, complete with a trademarked logo in the taskbar so they can see it the entire time they're using their computer.
Who wants to be "VESA HMD compliant"* when you can be "IN THE RIFT™" instead?
*I just made this up, but knowing VESA it has probably been a formal standard for several years now and everyone is ignoring it. Seriously, there's a VESA standard for everything.
Not surprising to hear that. Not even out on sale in the UK and already becoming a pain in the arse. I'm going to wait until bit-tech do reviews and I'll make up my mind till then.
Hah! Remember the days before OpenGL and Direct3D? You had *dozens* of different 3D rendering standards, and developers had to support each one by hand. Even before that, there were plenty of competing 2D standards as well: anyone else remember having to load VLB drivers in MS-DOS to get games to run in anything other than 320x200?
The early PCs were entirely incompatible; it wasn't until the launch of the IBM PC (well, until the cloning of the IBM PC, technically) that we got a 'standard.' The early 2D graphics cards were entirely incompatible, as were the 3D cards; it wasn't until the launch of OpenGL and Direct3D that developers could just 'support 3D' without worrying about what graphics cards their users had. Hell, even sound cards: Gravis UltraSound, Adlib Pro, SoundBlaster 16, SoundBlaster Pro - hands up if you remember faffing around getting your DOS games to play sound and music at the same time? A220, IRQ5, DMA1 anyone?
We're in the first generation of (the latest incarnation of) VR; products come first, and compatibility comes later. It's not a question of "the days when" - it's a cycle: launch competing products, one emerges as the strongest, becomes the standard - or an outside force introduces a standard and companies adopt it as a means to more rapidly grow their user base.
Talking VR and multiple platforms, I wonder if that google VR controller they announced works over standard bluetooth on PC's and such, that would be nice since you could use it across devices and platforms.
And if oculus doesn't have a VR controller yet by then people can use that /s
No I do not remember that, nor ever heard that, I think nvidia(?) had a third standard a while, but apart from that you had directx and opengl.
And seeing that before that time you had no 3D at all I think you must be dreaming up this memory or dozens of standards you mention.
Please tell me you're joking.
Thinking of only hardware-accelerated standards, you had: Matrox Impression, Matrox Millennium, Creative Graphics Library (3D Blaster), Paradise Tasmania, Nvidia NVlib, ATI CIF (Rage 3D), S3 S3D (ViRGE), Matrox MSI (Mystique), Rendition RRedline and Speedy3D (Verite), PowerVR PowerSGL, 3Dfx Glide 3D. None were compatible with any others, even with the same company; all had to be added to games specifically to support that particular graphics card.
3D existed long before hardware acceleration, of course. 3D Monster Maze was released in 1981.
Damn kids - get off my lawn!
Jeezuz, I knew I was getting old (half century in 3 weeks!!!!) but that lot takes me waaaaay down memory lane - it is easier to remember it than it is forgetting it....
Oh how the mighty have fallen. There is nothing appealing about oculus right now. Tied to Facebook and inheriting the Zucksters greed and disregard for his users.
Regarding standards. Isn't it basically how it always works. Everyone tries to develop their own brand lock in standard which eventually causes enough cluster****ery that either one wins (Bluray) or an open standard is developed ( W3C).
Tell me about it! Remember the joy of building the perfect MS-DOS bootdisk, with everything you needed loaded (MSCDEXX CD driver, Creative Labs SoundBlaster driver, mouse driver) and still managing to have more than 600KB of base memory left? DOS=HIGH,UMB? Running memmaker? LOADHIGH?
I still own a copy of The Complete Idiot's Guide to PC Gaming (bought in the 90s because it had a stack of shareware on the bundled CD); might drag it down for a nostalgia session tonight. I distinctly remember it has a section talking about how the soon-to-be-released Windows 95 might revolutionise PC gaming!
I get what you're saying with competing standards being the norm and all that, but i think what has most people up in arms is that (i think) the majority of competing standards are technically different, i.e one system can't (typically) talk to another whereas what Oculus are doing is (afaik) an artificial restriction.
IIRC memmaker never done a great job, you could free up a little more by shuffling the order that things loaded.
Oh, Faceculus are being knobbers, there's no doubt about that. My argument is simply that to think back to a halcyon time when companies did nothing but build entirely interoperable and freely usable standards for the benefit of the consumer is to ignore... Well, pretty much the entirety of history, really.
I skimmed The Complete Idiot's Guide to PC Games. I'd forgotten there's a section on VR, bemoaning that computers aren't powerful enough to render realistic environments (a fixed problem) and that the hardware is too expensive at $1,000 (we're working on that one still), but predicting that "in the next few years" VR will really take off. Well, swap "years" for "decades" and ol' Dave was on the money!
Close the internet, ladies and gentlemen, we have a winner.
I actually do, though my entry into the brave world of 3D accelerators was largely post-Direct3D.
And, to my recollection, the VESA drivers were "patches" for video card ROMs that didn't support whatever the current VBE standard was.
And yet, most of these myriad different platforms communicated with peripherals over RS-232 and the centronics printer port, and connected to 5.25" floppy drives with the Shugart disk interface.
And the Altair was widely cloned before the IBM PC was. Widely enough that the S100 bus was adopted by the IEEE and ANSI.
The notable thing about the IBM clones is not that they became a standard, but that they've STAYED a standard for so long. There's been no fragmentation of incompatible forks as they've assimilated new technology, nor has the platform stagnated and allowed others to pass it up.
IBM-wise, the early 2D cards were usually clones of IBM's offerings, though IBM's initial MDA/CGA split created an interesting problem.
Or they were clever modifications that retained back-compatibility with IBM's offerings.
It wasn't really until after VGA that the blatant incompatibilities surfaced... and then VESA laid out the VESA BIOS Extensions and all was right in the world again... for a few years.
I suspect without the intervention of Direct3D and OpenGL we would've seen GLIDE become a de facto standard. Because as I recall, 3dfx was the format you implemented first, the rest were optional. Everyone would've started setting up a GLIDE-compatible mode. But that's just a suspicion.
Or VESA would've stepped up and laid out a common low-level 3D accelerator interface.
Mmmm, SET BLASTER... for the longest time, that was the only use of SET that I knew.
I do note that most sound cards post-Blaster wound up being compatible with the SB Pro or SB16 in addition to whatever new game they brought to the table. Much like the Shugart interface or the Centronics port, Sound Blaster became a de facto standard.
I'm still waiting for us to cycle through on keyboard and mouse lights.
It seems to me like the modern market, lacking significant pressure from limited system specs, has no pressing reason to standardize. It doesn't hurt anyone's performance to have a little app sitting in the system tray handling mouse lights and button macros, an app can easily implement multiple VR paths by just building on the latest Unity and letting someone else deal with it, and so on.
This is pretty important: currently there are two PC VR APIs: Oculus VR, and SteamVR/OpenVR. Valve made a huge marketing coup by calling theirs 'OpenVR' when it is decidedly not: both are closed source and proprietary, both are controlled by a single company (without external input, so if you want to implement eye tracking, for example, then you're as SOL with OpenVR as you are with OculusVR), and both are controlled by a company with a vested interest in a single HMD.
The main difference between the two is that Oculus are dedicated to either having the VR experience from their dedicated store be perfect and seamless or not happen at all, while Valve are quite happy for a user to have a shitty VR experience from Steam and leave it up to them to weed out the wheat from the chaff, same as they have done with Steam in general. That leaves more choice up to the user, but also means anyone wanting to try VR will have to spend quite some time to learn what works and what doesn't at their own expense.
This is also why Oculus want to implement Vive support directly rather than implement an OpenVR frontend: they do not want to have to deal with the wave of low-quality OpenVR compatible VR HMDs (there are quite a few already available in China, including ones that attempt to hook into Oculus APIs e.g. Deepoon E2, which is the main reason for the new 'entitlement check') that will be popping up rather soon.
Off of their stores, both runtimes are pretty much the same: you can download whatever you want, from wherever you want (GoG, Humble Store, direct developer distribution, some random .exe on Mega, etc) and run it without issue.
Depends how far back you go: no RS-232 on the early 8-bits, or Shugart, or floppy drives for that matter! Plus, "being able to talk to each other" does not equal "compatible." Software written for the BBC Micro wouldn't run on a Dragon 32, or a Commodore 64, or a ZX Spectrum, or an Oric Atmos, or a TI-99/4A, or a TRS-80, or an IBM PC... Hell, even the BASIC languages were incompatible: I have an A2 poster from Personal Computer World magazine which acts as a look-up chart for converting from one machine's personal variant of BASIC to another. Huge bloody thing with teeny-tiny writing. Standards? We've heard of 'em.
It was indeed, although Ed himself was always peeved that it became "the S100 bus" when it should rightly have been "the Altair bus." My reason for picking IBM is highlighted in your very next sentence: it's the standard that we still use today (more or less.)
I remember 3dfx being very much against that: the company sued sites providing software wrappers, and Glide itself wasn't open-sourced until late in its life.
Creative couldn't even make its own stuff entirely 'standard': there's a reason you had SoundBlaster 16, SoundBlaster Pro, and SoundBlaster AWE32 as separate entries in setup programs!
Actually, I was wrong here. Unlike Valve's OpenVR, the source for Oculus's VR runtime is available as part of the SDK, though it is not freely modifiable and redistributable (i.e. not FOSS).
Separate names with a comma.