Page 1 of 1

2.0 and UWP

Posted: Mon Mar 14, 2016 10:05 pm
by ProBAR
Norbo, please tell me will Bepu support Universal Windows Platform? Especially .NETCore 5?

Re: 2.0 and UWP

Posted: Mon Mar 14, 2016 10:27 pm
by Norbo
.NET Core and native compilation friendliness are some of the main targets for v2. I haven't done extensive research into UWP-specific requirements, but I don't think there will be any issues there. At worst, some conditional compilation may be required. A physics engine doesn't have much in the way of OS dependencies.

The demos application, on the other hand, may not be UWP. If it's easy enough, I might, since there is value in having a widely running debug renderer. I doubt I'll bother supporting sub-DX11 hardware feature levels, though.

Re: 2.0 and UWP

Posted: Mon Mar 14, 2016 10:33 pm
by ProBAR
Norbo wrote:.NET Core and native compilation friendliness is one of the main targets for v2.
Glad to know that, thanks))
Norbo wrote:A physics engine doesn't have much in the way of OS dependencies.
This is why I choose your Engine, besides it is not hardware dependent.)))

Re: 2.0 and UWP

Posted: Mon Aug 22, 2016 10:46 am
by vicviper
Hi, I'm new to the forum, although I've been following Bepu engine for quite a long time.

I've finally had some time to tinker a bit with it and I found (and correct me if I am wrong) that the latest version is Bepu 1.4 , but there's a 2.0 in development.

I am interested in doing something that is platform agnostic, so I am also in the process of porting some of my libraries to Net Standard portable class libraries, and preferably, using System.Numerics.Vectors as default math types.

I've seen some people have unoficially ported Bepu 1.4 to portable class libraries and so on.... but... what is the current roadmap? bepu 2.0 is going to be released soon? or we better stick with 1.4 for a while?

Thanks in advance!

Re: 2.0 and UWP

Posted: Mon Aug 22, 2016 7:40 pm
by Norbo
I'm not technically working on v2 full time yet. The vast majority of its development will be in the open, visible on github. For the last 10 months or so I've been busy working on other aspects of the game, but once I'm done with this phase, I'll be swinging back to physics full time. And that development will take a while.

So, in other words, while I'll probably be starting physics development soon, I wouldn't say v2 is going to be complete soon. Given my history, it would be irresponsible of me to suggest that it will be mostly feature complete any sooner than 9 months from now. I hope that's enough time, or that I can get it done even faster, but if you need physics now, v1.4.X is probably the right choice.

Re: 2.0 and UWP

Posted: Sun Nov 27, 2016 10:17 pm
by kelthar
This makes me very glad to hear. .Net Core is a bright light in the sky, if I were to be poetic. So we can expect something in the middle of next year I guess. Unless it has been pushed forward.

What I want to try it to run the physics on a Raspberry Pi 3 under Windows 10 IOT just to see how many objects that can be simulated. It would be interesting to see if one could use RPi as light game servers. Maybe for instances or something like that. I just want to see the capabilities of a system like that. Would be isolated and sweet.

Re: 2.0 and UWP

Posted: Mon Nov 28, 2016 6:41 pm
by Norbo
So we can expect something in the middle of next year I guess. Unless it has been pushed forward.
Hopefully! The current non-physics work has, uh, expanded a bit and will probably reach into January unfortunately, but I'm still hoping to get at least the meat of the new version solid by mid-2017. I'm real bad at this.

(On the upside, a very large chunk of the non-physics stuff I've been working on for the last year will probably go open source by necessity at some point. Not sure when.)
What I want to try it to run the physics on a Raspberry Pi 3 under Windows 10 IOT just to see how many objects that can be simulated. It would be interesting to see if one could use RPi as light game servers. Maybe for instances or something like that. I just want to see the capabilities of a system like that. Would be isolated and sweet.
That's something I was wondering about too. I think I ran the numbers at one point and found the cost effectiveness on the current models to be underwhelming for serious compute server-ish purposes, but they win massively on novelty and adorableness (very important metrics).