Support NeoGAF

[Rymuth]

I'm confident in mine, I go out of my way to try and objective *shrug*

??

Stop posting that massive fucking spoiler image of the schedule. Fuck off.

+1. The people posting that spoiler in this thread are selfish (censored)s.

I will shoot any messenger I want thanks. if it gives any future messenger a moment's pause then it was worth It

http://www.neogaf.com/forum/showpost.php?p=206612015&postcount=4412

Pretty sure telling people to fuck off and/or threatening to shoot them isn't the definiton of objective.

 

[dr_rus]

In some way yes, but from what CIG said, they have more work on removing legacy codebase from DX9 and moving CryEngine fully to DX11, than from DX11 to DX12.
Which is mindblowing, because CryEngine was already one of the best optimized DX11 engines.
I imagine that for other big engines, like Anvil for example, its even more work.

I think this may have something to do with not making any changes to the shaders themselves when porting from DX11 to DX12. You can but you're not required to. When porting from 9 to 11 you had to rewrite the shaders because the shader model changed and the obvious advantages of 11 allowed to perform a lot of computations more efficiently so there was no reason to keep the old shaders around. This isn't the case with 12 where the biggest change is the engine's (renderer, actually) high level architecture.

It also remains to be seen how efficient CE's DX12 renderer is.

The reason behind the development of Mantle wasn't an ideological one, and I never claimed that. It was developed because DirectX was lacking at this point in time. But this is the past. Things are different now. DirectX12 is a viable alternative to Vulkan. It is also better suited for AAA game development than Vulkan, thanks to Microsoft's focus on AAA development, their good tools and documentation. Prefering Vulkan to DX12 for AAA development only makes sense if you arguing from an ideological point of view. That's also why even developers like DICE, with what they have done for the Vulkan project, seem to be focusing on DX12 for their AAA games.

Yup, the lack of any validation layer in DX12 (present in VK although is IHV specific IIRC) certainly "helps" AAA games development.

Fact is DX12 is in a rather "beta" stage still at this moment. It was close to being unusable in its initial release, they've patched it to somewhat ok'ish form with 1511 but there are still lots of things which it lacks and they'll need a couple more updates before it will be a viable AAA level API which will be used by the industry instead of being pushed by IHVs for their own agenda (like it's happening right now).

Vulkan is a more thoroughly thought through API at the moment because it was a collaboration of IHVs which built it with everyone contributing what was required for it to run across all the platforms. It's rather sad to see that AMD is pushing DX12 instead of Vulkan for example as there's zero reason for them to prefer the one over the other.

 

[c0de]

Timing, feature set, and purpose for existence.

No one can argue that impressive list of arguments. Thanks.

 

[LordRaptor]

I'm being slightly devil's advocate here because personally I'm in the camp that multi-device APIs are a good thing as it gives you more flexiblity/reduces dev burden (whether that's 'common to MS platforms' - UWP, 'common to many platforms' Vulkan, whatever - although obviously UWP vs. Vullkan is apples to oranges).

Vulkan is specifically a Graphics API. It solely deal with Graphics.
Outside of hard limits such as polycount and texture fillrate, graphics are incredibly scalable; you can turn shaders on and off, or use fallback default shaders where more complicated shaders are unsupported on target hardware. You can lower resolution. You can use prebaked lighting. you can use Forward rendering. You can use software rendering. You can use post-processing effects.
There are so many different variables that are purely graphical effects that can be scaled, removed or optimised for a target that OGL still provides the benefits of being a platform agnostic API even when making games for a toaster - or, crucially in the modern world we live in, when making a game playable in a web browser in HTML5.

For all your 'objectivity', you think DX12 is great because its made by MS, and you also think UWAs are great because they are also made by MS. Even though if what you actually believe in is platform agnostic APIs you would support neither.

e:

No one can argue that impressive list of arguments. Thanks.

If you want long editorials about how DX12 is obviously a response to Mantle, or how MS are worried about the erosion of their toolspace at the low end by HTML5 / IOS / Android popularity, none of which go near DX, then google it because there are hundreds of such articles.

 

[tuxfool]

It also remains to be seen how efficient CE's DX12 renderer is.

Not good. It is also built on top of an engine foundation designed for dx9/dx11.

 

[twobear]

Are there any performance downsides to a future where Vulkan is the only API in town? That's the only convincing reason I can think of to not want to use it.

 

[justsomeguy]

??




http://www.neogaf.com/forum/showpost.php?p=206612015&postcount=4412

Pretty sure telling people to fuck off and/or threatening to shoot them isn't the definiton of objective.

Nice try at history trolling, but objective is not the same dispassionate. When people were pointing massive spoilers in the middle of the live E3 thread, they deserve some abuse.

Also on the last point I was responding to someone who said "don't shoot the messenger". Saying i was threatening to shoot something is libellous and incredibly poor taste - I would ask that you reword it.

 

[justsomeguy]

Are there any performance downsides to a future where Vulkan is the only API in town? That's the only convincing reason I can think of to not want to use it.

Only the usual "competition is good" argument. From what I've read, Vulkan/Mantle's existence pushed the creation of DX12, which was a good thing for people who are in the DX camp.

 

[LordRaptor]

Are there any performance downsides to a future where Vulkan is the only API in town? That's the only convincing reason I can think of to not want to use it.

Well, standard Nvidia practice would be to moneyhat studios and implement hacks so that titles which should be running better on AMD cards turn out not to.

I'm not entirely convinced by DX12 or Vulkan performance claims yet, as I'm just not sure there are enough people in the industry that aren't already employed as Nvidia / AMD tech liasons or as engine renderer experts to make use of the extra control and required extra work to make use of them.

 

[roadrunn3r]

Do you work at Microsoft or something? I mean, with so many egregious statements I would think not, but then again...

Heh these type of threads attract a lot of these people and no, they are not MS employees just very dedicated hardcore fans of a huge corporation, lets call them that way. It's nothing new really but just be aware their point of view differs from that of a typical consumer and you can even have some interesting discussions with them.

 

[dr_rus]

Are there any performance downsides to a future where Vulkan is the only API in town? That's the only convincing reason I can think of to not want to use it.

Vulkan will never be the only API in town as not everything requires the performance benefits vs the complexity increase of Vulkan, lots of s/w (games including) will always be better off with OpenGL or DX11. Same is true with DX12.

People seem to think that DX12/VK is the same type of "transition" we had previously between DX versions while in fact it's not. Both DX12 and VK are very high performance, very high complexity OPTIONS added to DX11 and OpenGL. Both DX11 and OGL will still be used for years to come by those who can't justify the increase in complexity brought by DX12 and VK.

It will actually be quite interesting to watch the industry settle down around these APIs as there's a new choice now to make for any game or s/w using GPUs: does it really need DX12 or Vulkan or would it be ok to use DX11 / OGL? That's a new choice which wasn't there before last year.

 

[Caddyshack]

Heh these type of threads attract a lot of these people and no, they are not MS employees just very dedicated hardcore fans of a huge corporation, lets call them that way. It's nothing new really but just be aware their point of view differs from that of a typical consumer and you can even have some interesting discussions with them.

Pretty sure at least one of the posters in this thread has previously admitted to working for MS, actually... not the one that poster was quoting (afaik at least) but it does seem to be getting increasingly difficult to tell these days.

 

[justsomeguy]

For all your 'objectivity', you think DX12 is great because its made by MS, and you also think UWAs are great because they are also made by MS. Even though if what you actually believe in is platform agnostic APIs you would support neither.

Hmm, there's an awful lot of projection going on there... I could easily also say you believe that everything MS does is bad and therefore if you see "MS" you instantly go into attack mode.

If you're genuinely interested in a discussion on this rather then I find it all quite fascinating: MS are evolving from a company where everything was on Windows, to where they got left behind by the smartphone and tablet revolution - and now they're trying to play catch up. The company has embraced a way-wider, way more open stance in many areas (lots of Open Source releases, apps on iOS and Android, Unix on Azure, for example) - and yet in other ways they're still a platform holder that is focussed on primacy the experience on their own platform. Sony don't get grief for not making their APIs available on PC or on other devices... and that's fine IMO, it wouldn't make any sense for them to do that... so why should DX be any different?

I've got no issue with saying "Vulkan is great and makes more sense for targetting lots of platforms" - that's a no brainer. What gets tiresome is the predictable "MS did it, therefore it's bad" agenda that some posters have, whatever the topic, whatever the competitive situation, and whatever the wider context.

On UWP - from a technical point of view I like it, but I'm an enterprise software dev so it makes perfect sense for me. As a game dev technology I'm neutral-to-positive on it depending on how it works in practice. I simply don't have the data to say whether it's better or worse than Win32 for gamedev, however intuitively from what I've used of Win32 and what I've seen of UWP it's a far more productive API because it's designed with a lot of modern thinking and patterns and can eject a lot of the Win32 legacy bloat.

The UWP/Store combo is a fascinating one for me - all of Microsoft's competitors - the ones who are eating their market share - have closed platforms and a sole store operated by the platform holder. I don't think MS will ever go this route - but they have to provide something in this area. That's just common sense... otherwise in ten years time Windows will be dead and gone - I can hear the cheers, but ask yourself which alternative will take over? (and no, it won't be Linux).

So please don't try and pigeonhole me into some kind of cheerleader for Microsoft. As I've said before on here I won't come out and openly slag them off, but that works both ways - I don't tend to be critical of Sony et al when I think they're doing something stupid. I just try and comment on how I see things, without the hyperbole.

 

[Caayn]

It is a beta. It should also be noted that the developer mentions why that was the case.

Which is a nonsense argument. Vulkan support in Talos is currently a naïve port, and is even discussed at length by Croteam as such. Vulkan support is not native and is simply wrapped as a proof of concept. The intention is to get native support for Vulkan into Talos' engine, but it isn't done yet. That work also goes onto the next Serious engine for their new game.

Ah nevermind then, wasn't aware of that.

 

[justsomeguy]

Pretty sure at least one of the posters in this thread has previously admitted to working for MS, actually... not the one that poster was quoting (afaik at least) but it does seem to be getting increasingly difficult to tell these days.

I work at MS (it's in my profile) and actually you'll find us a pretty balanced bunch - we can get fired for astroturfing. Nobody wants to go there.

 

[MaLDo]

I work at MS (it's in my profile) and actually you'll find us a pretty balanced bunch - we can get fired for astroturfing. Nobody wants to go there.

justsomeguy
Occupation:
Work @ Microsoft - nothing to do with gaming - but opinions may be unintentionally biased

 

[michaelius]

Well, standard Nvidia practice would be to moneyhat studios and implement hacks so that titles which should be running better on AMD cards turn out not to.

I'm not entirely convinced by DX12 or Vulkan performance claims yet, as I'm just not sure there are enough people in the industry that aren't already employed as Nvidia / AMD tech liasons or as engine renderer experts to make use of the extra control and required extra work to make use of them.

That's pretty funny considering in AMD sponsored games performance on Nvidia cards fall when async is implemented with 0 benefits for gamer.

 

[tuxfool]

That's pretty funny considering in AMD sponsored games performance on Nvidia cards fall when async is implemented with 0 benefits for gamer.

Please keep the fanboy bullshit to a minimum in this thread please. We already have enough MS people, no need to add another dimension of nonsense to this thread.

 

[justsomeguy]

justsomeguy
Occupation:
Work @ Microsoft - nothing to do with gaming - but opinions may be unintentionally biased

Exactly my point when I said I try to go out my way to be objective. e: can we get back on topic now please? If anyone else has an issue with my posts PM bish and let him deal with it.

 

[Gemüsepizza]

Heh these type of threads attract a lot of these people and no, they are not MS employees just very dedicated hardcore fans of a huge corporation, lets call them that way. It's nothing new really but just be aware their point of view differs from that of a typical consumer and you can even have some interesting discussions with them.

So because I am arguing that DX12 is better for AAA development, which seems to be reflected by reality, this makes me a "very dedicated hardcore fans of a huge corporation"? Sigh. I am actually the complete opposite of that. I don't care about corporations or ideology when it comes to gaming. The only thing I care about is the best technology to get the best games. And because DX12 will make development easier when compared to Vulkan, I am convinced that this is the best technology.

 

[justsomeguy]

I'm not entirely convinced by DX12 or Vulkan performance claims yet, as I'm just not sure there are enough people in the industry that aren't already employed as Nvidia / AMD tech liasons or as engine renderer experts to make use of the extra control and required extra work to make use of them.

Ding ding ding - this is the key point for DX12/Vulkan from what I can tell: it's a lot harder to get good results from than their respective predecessors. That ID guy from the OP says that both Vk & DX12 have great performance advantages - but how hard will those be to leverage in the real world for the 'average' developer?

I guess that's where things like Unity, UE4, CryEngine shine - the top notch graphics techs there can get the renderers working and then the wider game dev community can leverage it.

 

[dr_rus]

Please keep the fanboy bullshit to a minimum in this thread please. We already have enough MS people, no need to add another dimension of nonsense to this thread.

It's hardly nonsense but it's certainly offtopic.

 

[LordRaptor]

Hmm, there's an awful lot of projection going on there... I could easily also say you believe that everything MS does is bad and therefore if you see "MS" you instantly go into attack mode.

I was going to write a longer rebuttal, as you were the one that decided to bring up the subject of post histories and are implying anything I've said in this topic is because I'm an "Ms hater" - which is hugely reductive as to the points I have made in this topic, as well as a mischaracterisation of my very specific problems with very specific MS actions, which can be best summarised into the tl;dr of
"The Xbox division should not be in control of PC gaming as they fundamentally do not understand it and there is an inherent conflict of interest"

but...

I work at MS (it's in my profile) and actually you'll find us a pretty balanced bunch - we can get fired for astroturfing. Nobody wants to go there.

...it would be pointless. I'll just leave that last sentence there though;
Conflict Of Interest.

That's pretty funny considering in AMD sponsored games performance on Nvidia cards fall when async is implemented with 0 benefits for gamer.

I only buy Nvidia cards, but I'm not blind to their shenanigans though.

 

[KampferZeon]

Vulcan will be the king graphics API in 3 years time,
Mainly because of Android, better performance and lower power consumption.
ms has no presence in the mobile space will be it's downfall.

 

[Gemüsepizza]

How does mobile gaming affect AAA gaming on PC and consoles?

 

[OtisInf]

This seems to mostly if not entirely negate the notion that UWP is advantageous if you're a game developer targeting both Windows PC and the X1.

Aren't you confusing 'requirement' with 'advantage' ? Not only that, but using UWP doesn't make your render pipeline work on both PC and X1 with a single code path. At most it's a framework in which your actual game code runs, which for PC and X1 still requires different codepaths if you want to get decent performance.

 

[backstep]

In terms of the APIs themselves, there isn't a huge difference for devs between Vulkan and D3D12. They're both stateless and usage is pretty similar, the biggest difference seems to be how resource views are handled. Someone in this thread mentioned Vulkan has a validation layer that D3D12 lacks, but actually D3D12 does have a very verbose debug layer that you can enable.

In my own opinion the biggest factor, one in favour of D3D12, is the difference in developer tools and support. If you have a problem with some D3D12 feature, or even a feature request, you can contact MS's graphics team and they'll get back to you whether it's an API issue or a driver problem, or a PEBKAC situation. With Vulkan, just like with OpenGL in the past, if you have an issue you can only go to the graphics hardware vendors. Kronos only provide the spec for them to implement (and/or extend).

So that's one less level of support available already, but what about the differences in IHV support? Both D3D12 and Vulkan have much thinner drivers sure, but they still require drivers. Where the IHVs focus their efforts to some extent dictates the choice devs will make. I mean everything you could do in D3D11 you could do in OpenGL 4.x but the variations in driver quality for typical hardware always made D3D11 the no-brainer choice. It was just more reliable across a wider range of graphics hardware. From what I can tell, and I could be wrong, the IHVs seem more focused on reliable and optimised D3D12 support for their hardware at this point. For example's sake, the most popular graphics vendor's software already has D3D12 debug and profiling support while Vulkan is marked as a planned feature.

TLDR: Both API's usage is quite similar, but one has direct support and tools from API creator, and graphics IHVs appear to be focusing more on one API also. For game devs (especially the kind who even care about the lower level access the new APIs provide) those things matter, the same as they did in the D3D11/OpenGL situation.

 

[justsomeguy]

I was going to write a longer rebuttal, as you were the one that decided to bring up the subject of post histories and are implying anything I've said in this topic is because I'm an "Ms hater" - which is hugely reductive as to the points I have made in this topic, as well as a mischaracterisation of my very specific problems with very specific MS actions, which can be best summarised into the tl;dr of
"The Xbox division should not be in control of PC gaming as they fundamentally do not understand it and there is an inherent conflict of interest"

Fair comment - that IS my interpretation of the last few weeks of your posting. If I'm wrong I apologise and I'd much rather engage on specific points though in general. Will try harder next time!

Structurally, the Xbox "divison" is part of the Windows org now as of about a year ago - i.e. Phil Spencer's boss is the head of Windows... so anything they do is done with the mandate of the overall Windows strategy. They used to be separate.

 

[Skinpop]

creative assembly should take a long hard look at themselves.

anyway, this has been obvious for a while to anyone who knows anything about these things.

 

[LordRaptor]

How does mobile gaming affect AAA gaming on PC and consoles?

If that's the line you wish to take, you will need to define your terms; is Ashes Of The Singularity a AAA title? Are indie titles built in UE4 and planning to use DX12 when UE4 supports it (and presumably also happy to use Vulkan when UE4 supports that) part of the conversation?

 

[Krejlooc]

This here is the slight fallacy for me - in the UWP threads there are plenty of people saying "You won't have a AAA game running on phone AND pc, so UWP is useless"... so surely the same logic applies here?

I'm being slightly devil's advocate here because personally I'm in the camp that multi-device APIs are a good thing as it gives you more flexiblity/reduces dev burden (whether that's 'common to MS platforms' - UWP, 'common to many platforms' Vulkan, whatever - although obviously UWP vs. Vullkan is apples to oranges).

However, if you're building a AAA game you're going to be targetting PC, PS, XBox. Two of those use DX12 (with a 1% difference according to another post I saw in the thread), one uses (as far as I know) a Sony proprietary API.

If you're using a common API across multiple different devices (for example top end PC and Android) then surely no way are you going to have *one single set of rendering code*? That would be insane... so even with "one API" in Vulkan you're still doing loads of platform/device specific stuff. So IMO Vulkan isn't the technical panacea some here would like it to be. (e: neither is it a bad thing)

You've set up a strawman and ignored all the arguments about pc segmentation in the process. People are not championing vulkan to bring games to android, they are (and this is the crux of the argument so pay attention) because a minority of what constitutes "pc" arbitrarily cannot run dx12.

This isnt about getting software running on unfit hardware. Its literally the exact opposite - letting fit hardware run the way it should without arbitrary software limitations.

 

[InsaneTiger]

 

[justsomeguy]

You've set up a strawman and ignored all the arguments about pc segmentation in the process. People are championing vulkan to bring games to android, they are (and this is the crux of the argument so pay attention) because a minority of what constitutes "pc" arbitrarily cannot run dx12.

This isnt about getting software running on unfit hardware. Its literally the exact opposite - letting fit hardware run the way it should without arbitrary software limitations.

Ah - actually I had something in about the PC segmentation, I must have taken it out before hitting post (I tend to over-analyse what I write, re-edit ten times, and still cock it up - hah)... I think that's a totally legit argument personally - why isn't DX12 back compat to at least Windows 8?

edit: bugger yes that '1% difference' I put in... wasn't my intention to disregard the segmentation, just a cock up on my part... have edited my post accordingly.

 

[hooijdonk17]

Vulkan is the future and I for one hail our id software overlords.

 

[Colbert]

Hmm, there's an awful lot of projection going on there... I could easily also say you believe that everything MS does is bad and therefore if you see "MS" you instantly go into attack mode.

If you're genuinely interested in a discussion on this rather then I find it all quite fascinating: MS are evolving from a company where everything was on Windows, to where they got left behind by the smartphone and tablet revolution - and now they're trying to play catch up. The company has embraced a way-wider, way more open stance in many areas (lots of Open Source releases, apps on iOS and Android, Unix on Azure, for example) - and yet in other ways they're still a platform holder that is focussed on primacy the experience on their own platform. Sony don't get grief for not making their APIs available on PC or on other devices... and that's fine IMO, it wouldn't make any sense for them to do that... so why should DX be any different?

I've got no issue with saying "Vulkan is great and makes more sense for targetting lots of platforms" - that's a no brainer. What gets tiresome is the predictable "MS did it, therefore it's bad" agenda that some posters have, whatever the topic, whatever the competitive situation, and whatever the wider context.

On UWP - from a technical point of view I like it, but I'm an enterprise software dev so it makes perfect sense for me. As a game dev technology I'm neutral-to-positive on it depending on how it works in practice. I simply don't have the data to say whether it's better or worse than Win32 for gamedev, however intuitively from what I've used of Win32 and what I've seen of UWP it's a far more productive API because it's designed with a lot of modern thinking and patterns and can eject a lot of the Win32 legacy bloat.

The UWP/Store combo is a fascinating one for me - all of Microsoft's competitors - the ones who are eating their market share - have closed platforms and a sole store operated by the platform holder. I don't think MS will ever go this route - but they have to provide something in this area. That's just common sense... otherwise in ten years time Windows will be dead and gone - I can hear the cheers, but ask yourself which alternative will take over? (and no, it won't be Linux).

So please don't try and pigeonhole me into some kind of cheerleader for Microsoft. As I've said before on here I won't come out and openly slag them off, but that works both ways - I don't tend to be critical of Sony et al when I think they're doing something stupid. I just try and comment on how I see things, without the hyperbole.

Signed.

 

[Synth]

Yeah DX has become much better, was just reminiscing on how strange its "victory" over the alternative APIs was, considering how easy it was to see how bad D3D ran everything.

I'd imagine the Unreal Engine circa Unreal Tournament would have had something to do with this. OpenGL was basically crippling to lower end machines running it at the time, whilst Direct3D ran amazingly smooth by comparison. Could never work out what the issue was as Quake 3 Arena fucking flew on the same machines.

Funnily enough though, today I can't for the life of me get UT99 running without framerate hitches in Direct3D, whilst OpenGL runs flawlessly on newer hardware.

 

[Skinpop]

Anyone know what the ratio of middleware led to pure Dev games is now days? Aren't ue4 et Al isolating most devs from this kind of thing?

most serious devs still make their own engines and I'd argue indie developers should do so as well if they have competent programmers. I have no clue about ratio but it would certainly be a disaster if every game made in the future was made using ue or unity.

 

[hooijdonk17]

I'd imagine the Unreal Engine circa Unreal Tournament would have had something to do with this. OpenGL was basically crippling to lower end machines running it at the time, whilst Direct3D ran amazingly smooth by comparison. Could never work out what the issue was as Quake 3 Arena fucking flew on the same machines.

Funnily enough though, today I can't for the life of me get UT99 running without framerate hitches in Direct3D, whilst OpenGL runs flawlessly on newer hardware.

The reason was drivers. The reason is drivers.

 

[Krejlooc]

most serious devs still make their own engines and I'd argue indie developers should do so as well if they have competent programmers. I have no clue about ratio but it would certainly be a disaster if every game made in the future was made using ue or unity.

You arent a serious dev unless you build your own engine? You want cash starved indie developers to waste their time reinventing the wheel?

Neither of these statements are realistic nor reflective of modern games development.

 

[charsace]

The issue with Khronos Group is documentation and getting help. That's the bottom line. Blender 3D had the same issue and has recovered from it and has gotten more users because of it.

If KG improves support people will hop on the wagon.

 

[digitalrelic]

As a trojan horse, yeah.

Hahhaaha.... A trojan horse? For what exactly?

 

[LordRaptor]

The reason was drivers. The reason is drivers.

I'd say DX's 'secret' killer app was DirectInput, which enabled finding and remapping of any connected device to be basically transparent to an end user.
Bear in mind before we had super useful things like USB ports, people were hooking up controllers via the frigging printer port

 

[Krejlooc]

I'd say DX's 'secret' killer app was DirectInput, which enabled finding and remapping of any connected device to be basically transparent to an end user.
Bear in mind before we had super useful things like USB ports, people were hooking up controllers via the frigging printer port

Er, this isn't right. The gameport standard had been ubiquitous for a good half decade before direct input came around. Using your printer port is some 1986 tech lol.

 

[blu]

Sigh. Most of these companies have been members of the Khronos group for ages. This doesn't mean anything. At all.

This is not a generic list of Khronos members. At all.

 

[LordRaptor]

Er, this isn't right. The gameport standard had been ubiquitous for a good half decade before direct input came around. Using your printer port is some 1986 tech lol.

I had a gamepad using LPT1 in 1999

 

[dragonfart28]

Meanwhile, no vulkan update for Doom yet.

 

[justsomeguy]

Meanwhile, no vulkan update for Doom yet.

Eh? It just came out, there's a whole thread about it.

http://www.neogaf.com/forum/showthread.php?t=1245006

 

[hooijdonk17]

I'd say DX's 'secret' killer app was DirectInput, which enabled finding and remapping of any connected device to be basically transparent to an end user.
Bear in mind before we had super useful things like USB ports, people were hooking up controllers via the frigging printer port

Actually it was the serial port, and the serial port is like.. ancient. I recall there being USB even back when DX was at v5? Maybe even v4? Regardless, nobody on PC actually gamed with gamepads until Xbox 360 became prevalent. If anything, Direct Input mostly benefited people who played niche games requiring things like wheels and flight sticks.

Meanwhile, no vulkan update for Doom yet.

What?

 

[DouglasteR]

Hmm, there's an awful lot of projection going on there... I could easily also say you believe that everything MS does is bad and therefore if you see "MS" you instantly go into attack mode.

If you're genuinely interested in a discussion on this rather then I find it all quite fascinating: MS are evolving from a company where everything was on Windows, to where they got left behind by the smartphone and tablet revolution - and now they're trying to play catch up. The company has embraced a way-wider, way more open stance in many areas (lots of Open Source releases, apps on iOS and Android, Unix on Azure, for example) - and yet in other ways they're still a platform holder that is focussed on primacy the experience on their own platform. Sony don't get grief for not making their APIs available on PC or on other devices... and that's fine IMO, it wouldn't make any sense for them to do that... so why should DX be any different?

I've got no issue with saying "Vulkan is great and makes more sense for targetting lots of platforms" - that's a no brainer. What gets tiresome is the predictable "MS did it, therefore it's bad" agenda that some posters have, whatever the topic, whatever the competitive situation, and whatever the wider context.

On UWP - from a technical point of view I like it, but I'm an enterprise software dev so it makes perfect sense for me. As a game dev technology I'm neutral-to-positive on it depending on how it works in practice. I simply don't have the data to say whether it's better or worse than Win32 for gamedev, however intuitively from what I've used of Win32 and what I've seen of UWP it's a far more productive API because it's designed with a lot of modern thinking and patterns and can eject a lot of the Win32 legacy bloat.

The UWP/Store combo is a fascinating one for me - all of Microsoft's competitors - the ones who are eating their market share - have closed platforms and a sole store operated by the platform holder. I don't think MS will ever go this route - but they have to provide something in this area. That's just common sense... otherwise in ten years time Windows will be dead and gone - I can hear the cheers, but ask yourself which alternative will take over? (and no, it won't be Linux).

So please don't try and pigeonhole me into some kind of cheerleader for Microsoft. As I've said before on here I won't come out and openly slag them off, but that works both ways - I don't tend to be critical of Sony et al when I think they're doing something stupid. I just try and comment on how I see things, without the hyperbole.

Thats some quite nice thoughts, appreciated.

 

[cakely]

There is literally no downside to having another graphics API available as an alternative choice to Direct X.

If some developers feel that it's the better choice, it will get picked.