There hasn’t been a major DirectX version from Microsoft for a while now. And now with the decline of Windows as the dominant OS and the proliferation of multi-platforms, open standards are winning, like OpenGL/WebGL. AMD also directly mentioned that they don’t believe there will be a DirectX 12 coming, and they have chosen to release their own custom API (Mantle). For me, the future seems to favor multi-platform solutions rather than being tied to Windows only. The market share of Mac OS X and Linux cannot be ignored anymore. Valve also seems to agree when they brought Steam to Linux.
What is everyone’s thoughts on this?
Please log in or register to post a reply.
I wouldn’t be so quick to write off Microsoft and Direct3D. Things like Mantle, Linux, GL, etc. certainly pose threats to them - but those threats may just be the incentive Microsoft needs to rise to the challenge and build a new and better D3D. MS naturally don’t want Mantle or GL to become game developers’ weapon of choice on Windows, so they know they have to get cracking and deliver a D3D that’s competitive with the other options. They may or may not succeed, but they have a better reason to try hard now than they have for the last several years.
Plus, with MS getting a new CEO soon, we may see them changing direction (and hopefully ending their run of bad decisions with Win8 and XB1).
I have doubts about Mantle ever becoming an industry standard. But even if it doesn’t take off, I think it’s a great thing for AMD to have done just because it’s shaking up the industry, getting people talking/thinking about the future of APIs and questioning their assumptions. And the industry could use some shaking up. Who knows what may happen with APIs over the next few years? It’s an interesting time, for sure.
I’m not really writing off Microsoft, though that’s a separate topic. I agree that they’re still in a strong position for gaming.
However, my main point is whether Microsoft likes it or not, and whether they innovate in D3D or not, the world is gradually changing towards a non-Microsoft centric ecosystem, where developing on other platforms now matters. That direction doesn’t play well for Direct3D, since it’s exclusive for Windows (let alone regular desktop applications). Even though Direct3D provides a nicer API for graphics hardware, developers are more likely to adopt multiplatform solutions, like OpenGL. Developers ultimately follow the users and what’s good for their business, not which API is the most innovative.
Microsoft is losing it’s grip on games. Consoles, tablets, and phones and other television game systems are not going to be using DirectX in any large numbers, and the idiot policies of Microsoft, such as not letting older versions of Windows update to newer versions of DirectX has already taken it’s toll. New management might help, but it’s hard to undo 10 years of stupid. If they manage to keep it as a developer and high end game platform, they will be doing good. Not having a recent update is yet another bad call. No one knows what their plans are or where they are going with it.
@Reedbeta Ancient Chinese Curse “May you live in interesting times”
Microsoft is self destructing, hell my windows 7 64 bit laptop is dead at the moment. Microsoft released an update yesterday that got to update 15 of 17 and hung. I know have a £3000 brick.
If a change of management can turn them around is still to be seen.
I think the xbone will be a success, not a huge success, but a success.
They promise it will be easy to code for and indies can use a commercial device for development. They say it supports Unity, which is of no use to me at all.
Does it support XNA? Not sure
Can you program in c++? Not sure
DirectX/D3D ? Not sure
“There hasn’t been a major DirectX version from Microsoft for a while now”
- Well, does there have to be? We already have powerful APIs at our disposal. What I think is more important is hardware evolution. Pushing more fillrate, more polygons, higher resolution textures, faster GPGPU, etc. IMO DX 8/9 and OpenGL 2.0 were the big changes, and then the GPGPU support later. Since then, all the new revisions were just icing on the cake, nothing worth raving about IMO.
I don’t think there will be much of a shift. Xone doesn’t appear to be in a bad position since both MS and Sony soldout all their preorders and they have a better lineup of games too. That puts them in a good position. It also means we need to know both their APIs and live with it. If a 3rd party enters the race and achieves high market penetration, then we’ll just have to add their APIs onto the list as well. It’s just the nature of things, sadly.
Also remember, OpenGL was multiplatform from day one. That wasn’t good enough for Microsoft and so they came up with DirectX. It took them 6 revisions to finally get a payout from that hassel, and it wasn’t really until 8 and 9 before devs officially made the switch permanent. Although a large part of that was the messy situation OpenGL was in, which is an unfortunate consequence when you have an open standard. Getting people to agree is like pulling teeth. Just look at the WiFi industry. Equally bad, if not worse. Nevertheless, you also have to deal with platform specifics. Graphics APIs are one thing, operating systems, programming langauges, audio, platform-specific APIs, submissions requirements, etc. These all differ across platforms. So really there is no standard out there other than vanilla C++ for the most part. If people want a standard, they’ll pick a platform and exclusively develope for it. Consumers won’t have a choice but to cave in and buy the same hardware we developers chose. Unfrotunately, it doesn’t work that way.
As in other thread - did really Valve bring Steam to Linux? As they officially state:
Currently, Steam for Linux is only supported on Ubuntu 12.04 LTS or 12.10 with the Unity, Gnome, or KDE desktop.
Currently, Steam for Linux is only supported on Ubuntu 12.04 LTS or 12.10 with the Unity, Gnome, or KDE desktop.
So Valve really brought Steam only for the Ubuntu 12.04 or 12.10 … which is by far the least stable Linux distribution I’ve ever worked with. Stable distros like Debian, Arch (or literally TONS of others) are unsupported, customizable distros like Gentoo either. Note that I’m not taking as “supported” (F.e. in case of Arch), where you have to literally “rape the system” to make Steam work.
Supporting just Gnome, KDE or Unity is also a bad thing, there is a ton of people using full environments like Xfce, LXDE, or just window managers like Fluxbox, Openbox.
Valve is in my opinion fairly large company that could hire people to test & debug their applications across multiple distributions and on multiple desktop environments.
Well, that’s really the problem with Linux isn’t it? They were trying to do more universal apps ages ago but it doesn’t happen. People have given up and went with Ubuntu. Linux will never be more than a tinkerer’s OS except maybe for Ubuntu.
I don’t agree on point that people went with Ubuntu (they didn’t, at least not anyone I know off switched to Ubuntu (and I’m in sorrounded by area of long term Linux users).
Bloody mess of dependencies is a problem on Linux, especially for inexperienced developers (from my point Valve has inexperienced Linux programmers) - but let’s hope they will learn from mistakes and fix their software.
Note that I’ve created tons of unportable software with tons of glitches and issues in dependencies. I’m proud of it, because the more mistakes you find in your software, the better you are next time :).
At some point that was “except maybe for redhat”, and at some point “except maybe for debian”. =) Currently, I’d wager it’ll be “except maybe for SteamOS”.
Saying that you bring stuff to “linux” doesn’t mean it has to work on every possible incarnation of linux, in my opinion.
Next you’ll be saying that there’s no such thing as linux, but gnu/linux. =)
I didn’t mean to look like a total zealot with this, but supporting only one distro and thats Ubuntu isn’t really supporting linux (generally). I’m kind of allergic on that one linux - it is also well known that building applications on Ubuntu will most likely make them unportable or hard to port to different distributions.
This Ubuntu distro is also well known as linux “that is not linux” - they’re trying to do everything differently and probably digging its own grave (moving to Mir from X might kill it, or at least it will kill it as long as there will be no up-to-date GPU drivers for Mir (as they are for X)) - that would get them to position where they are the only “official” distro able to run Steam for Linux, while unable to actually run any game (because their GPU driver support would heavily suck).
I won’t argument that there is no linux but only gnu/linux - I’m more like disappointed that Valve didn’t really released “good piece of software” on Linux (which is subjective term). And of course on topic of Linux we could flame one year in a row here… and I guess that neither of us i really interested in spamming typical flaming arguments at each other (that are on each common forum - let’s keep discussion at better level than other forums :) )
I thought X was kind of outdated and people just stick to it because they don’t want to change? It’s just one more example of what’s wrong with Linux. You need to get everyone on the same page to a certain extent and that just can’t happen with Linux. I used it for a while and was a big proponent until I realized it was never going to go anywhere. One branch might go someplace, but they can’t agree enough to actually move the OS decently. The dependency issue is horrible and they came up with this kind of work around rather than fix it, so you get this install that goes around and finds all the dependencies. Drivers work on one distro but not on another. It’s like a bunch of kids squabbling and saying I want to do it my way! People that do special applications find it useful, everyone else just scratches their head. The only thing that really works about linux is the kernal. The rest is too dispersed.
Ad X - There seems to be 2 alternatives so far - Wayland and Mir. Intel stated that they won’t support Mir, so basically Canonical teams will have a ton of work (their fault, one day supporting Wayland and the other day creating Mir - only for them).
AMD/NVidia drivers works with Wayland to some extent, and it seems that Fedora will start supporting Wayland - which means that Red Hat Enterprise will also start supporting it - which means that drivers from AMD and NVidia will be released for Wayland.
And also it is important to note, that Wayland is backwards compatible with X, while XMir is to some extent too, from what I’ve heard, Mir won’t be backward compatibe with X.
And of course there are distros like Debian that will keep ongoing support for X.
This all mess will end probably in:
1.) AMD/NVidia will only support X, Mir will be wrong step, Wayland will be probably better (due to backwards compatibility)
2.) AMD/NVidia will support X and add also Wayland (slowly ceasing to support X later), and we will translate to Wayland. Mir will be wrong step.
3.) AMD/NVidia will cease to support X, add Mir support. Canonical will be happy, Red Hat and thousands of other companies will migrate to Intel. AMD/NVidia will lose a ton of customers.
4.) Everyone will support X, Mir and Wayland and we will all be happy.
Now which of these seems like most probable version, Let’s assume point 4 won’t happen (as Intel said they won’t support Mir), I guess it’s not the point 3 also. They definitely don’t want to give a ton of their customers to Intel.
So we will either stick to X, or slowly move to Wayland. In my opinion Canonical does really dumb thing with Mir and they will eat what they created. Intel started it, I guess more will join.
Maybe also Mir is a reason why Steam develops their own OS instead of sticking to Crapbuntu.
I really don’t know why ui framework needs specific hardware support; why not just build it on top of opengl, for instance?
But yeah, revamping the whole ui mess will still require some kind of legacy support, for example like what osx is doing with x (x apps running under “x application”, basically). I have no idea what the gnome/kde stuff would require.
..anyway, I don’t know the systems well enough; I do know things can be much hairier in reality than they “need” to be..