Awwhhhh… I just noticed this today, http://unity3d.com/unity/roadmap
Anyone know why it was pushed back from 5.3 to 5.4? And whether other engines are having similar delays? Not that it matters much in the long run, but just curious.
Awwhhhh… I just noticed this today, http://unity3d.com/unity/roadmap
Anyone know why it was pushed back from 5.3 to 5.4? And whether other engines are having similar delays? Not that it matters much in the long run, but just curious.
Re-jiggering the entire render pipeline probably took longer than they thought. Unreal support where you dont lose FPS will probably take the same amount of time.
I don’t think Unity has to rework the “entire render pipeline” to ad DX12 support - especially now that they reworked the graphics API integration/picking (notice how DX11 is no longer just a checkbox).
Using the DX12 version of the Infiltrator demo actually increases FPS.
DirectX 12’s primary feature is a multi-threaded pipeline. Unity’s graphics pipeline is having to be reworked for it.
@Art-Leaping This doesn’t cover everything but it is a quick look at what they have/had to do, http://aras-p.info/texts/files/201508-SIGGRAPH-PortingUnityToNewAPIs.pdf.
@Frpmta I have seen it both ways in both, some of it is the drivers fault, some of it is that while the old DXs had tons of overhead, their are now things the engines need to redo that the drivers had very well optimized.
Yup, must admit on this one I don’t envy them…
Everyone went on vacation during the summer so we had to push it back ![]()
There’s been experimental DX12 support since 5.2 right?
Rather they push it back than release some buggy implementation…
Well true, but it’s been a very unfortunate pattern in the 5.x cycle so far. Features get announced, get rolled out in a completely unfinished state, and many deadlines gets pushed back all the time. At this point with the scale of Unity, maybe a few key hires to actually get Unity5 up to speed might be something to consider ( it’s not like this engine is a few people working on it…) I do understand that Unity5 requires a lot of rewritting, but it is getting pretty ridiculous.
Adding more people though won’t necessarily speed anything up. If anything it may actually slow things down as they’re brought up to speed with how the engine works and are integrated into the team(s).
I understand that, I know that brigning more people on board now would most likely mean a decreased output in terms of feature implementation and bug fixes, but what we’re seeing now, with features being pushed back, and back even more ( i’m mostly talking about my own issues with 5.x, which may differ from other developers) to the point that I’m loosing hope that some of the features that are being perpetually delayed are actually going to make it in before Unity starts rolling out Unity6 alphas. So still… in my eyes, adding more hires, while maybe dragging down efficiency in the short term, would improve the 5.x cycle in the end.
No, it will not. Epic’s team is tiny. Most engine teams are small. You can’t add more people, that is demonstrating complete ignorance of programming in general.
Also, where do you find these rock star programmers? Last I checked all the best ones are tightly hired up. Unity has some very smart people working there - some of the smartest developers in the world, responsible for many of the AAA games you’ve played.
They also have 500 staff or so. Adding more will not fix this because the problem is that a given job relies on other parts of the engine to be coded. If you program hello world with 10 people, is it done quicker or slower?
Adding more artists, now that works.
@hippocoder - you’re wrong - JUST WRONG! haahaha jk. Sorry I had to butt in for a minute lol.
EDIT - But naw on a real note.
I do have a DX 12 card, however I’m still wanting to use my Windows 7, just because it’s what I’m used to.
being I have a DX 12 Card - will I benefit from DX12 at all?
Because last I heard it’s only a Windows 10 feature?
You need Windows 10.
(or a small team working on post effects non stop? I can’t understand why Unity’s post effect filters haven’t been really updated since… 3.x? Or why there isn’t a decent mobile DOF filter included)
awesome-graphics-confirmed:
However Epic has their full engine source code available for free, and tons of developers are actually submitting fixes that are integrated into the engine at about all patch releases, offloading dev costs and fixing errors that should not be in the engine. I mean come on, don’t try to tell me that Unity has no errors that have been gliding along for the last few years or more.
And I’m not saying “add more people”, I’m saying that from a project management standpoint, when a product team continually misses deadlines and PM’s allow bugs to filter through betas and patch releases and that continues on for a while, there’s a broken structure within the studio. If my team got back to me telling me " deadline will be december 15th, then on that date, i get a …“hmm, we’re not sure actually when we can get that done, let’s push it out a month or two from now and see where we are?” and this happens for a lot of the highlights of the development roadmaps, maybe rethinking the way it’s actually developed, or planning to hire more staff would be key in my book.
“If you program hello world with 10 people, is it done quicker or slower?” I dont think that you are making a comparison between DX12 implementation and hello world. DX12 implementation should be broken down into major modules and each assigned a team to work on, especially when you have a feeling that deadline is going to be missed. Hello world… well, 1/10th of a programer would get it done ![]()
That’s fine, the whole world is wrong and you are right ![]()