Page 1 of 1

Should I use OGRE?

Posted: Wed Sep 12, 2018 7:59 am
by JanKleemann
Hello everyone.

I have an existing application written in QT/Qml, which targets iOS, Android and Windows. My next task is to write a 3d viewer and right now I'm trying to find out which engine to use, especially considering Apples announcement to deprecrate OpenGL.

The viewer itself should be quite simple: It will show a few objects in different poses, that can be rotated. Materials will include reflecting materials like gold, normal maps, etc.

OGRE 2.1 suppports Metal, but does not support Android. What are the plans for Android support? Can I expect Android support anytime soon (like this year)?

OGRE 1.11 currently supports all target platforms, but will that be true in the future? With Apple deprecating OpenGL, will OGRE 1.11 be viable on iOS in the future? Or will I have to port to OGRE 2.1?

Assuming I will start with OGRE 1.11, how much work will porting to 2.x be, once Android support is there?

Thank you in advance
Jan

Re: Should I use OGRE?

Posted: Wed Sep 12, 2018 12:05 pm
by paul424
Heh a 3dviewer , are you kidding me :) ?

Maybe you could write something like http://wiki.ogre3d.org/Ogre+Meshy and supporting the ogre3d's .mesh file format is a must :) !

Re: Should I use OGRE?

Posted: Wed Sep 12, 2018 12:53 pm
by JanKleemann
I assure you that I'm not kidding. Why? You think this is too simple? Or do you think I'm underestimating the complexity of it?

We have existing assets in different file formats (dae, c4d, 3ds, blend). If needed, I will code an importer using assimp. Or maybe there is a converter to OGRE's .mesh...
These assets come from basically the same viewer I'm writing, which is written in BabylonJS. I'm to rewrite the viewer, so it can be integrated into my QML app and into our other applcations (most of which are written in Delphi).

Re: Should I use OGRE?

Posted: Wed Sep 12, 2018 6:32 pm
by zvd2
I wasn't actually aware Ogre 2.1 doesn't have Android support. Mr. Goldberg is the main 2.1 developer and he seems to be working on screenspace decals at the moment. I figured he would have implemented Android support long ago considering he is developing a mobile game himself.

Ogre 1.11 seems to be widely used enough that I'm sure it will stay supported and otherwise Ogre 1.11 supports WebGL through Emscripten, which iOS isn't dropping support for.

I don't think Ogre 2.1 would have a big performance impact on a model viewer anyways, but mr. Goldberg can probably answer anything related to that better than I.

Re: Should I use OGRE?

Posted: Wed Sep 12, 2018 7:06 pm
by zvd2
Moreover for conversion you can just copy paste from the ogre-assimp command line converter.
https://github.com/OGRECave/ogre-assimp

Re: Should I use OGRE?

Posted: Thu Sep 13, 2018 7:28 am
by JanKleemann
Thank you for your reply. The dae assets are around 200 MB, but I think the scene won't show more than a few hundred objects at any one time.

WebGL support may be interesting. Maybe we can replace our old BabylonJS viewer. But the viewer also has to work outside of a browser. I think in Qt/QML the only method to show WebGL would be the WebView item, which we don't want to use. I'm not sure though. If anyone has any experience with WebGL and Qt, I'd appreciate some insight.

Who would know about future OGRE 1.x iOS support? Whom should i ask?

Currently I'm implementing the viewer in Qt 3D, but Qt 3D doesn't seem to be very mature at this point. There is a lot of stuff missing.

Re: Should I use OGRE?

Posted: Fri Sep 14, 2018 8:11 am
by zvd2
I can't answer the rest, but Paroj leads 1.11 development and would definitely be the person to ask.

Re: Should I use OGRE?

Posted: Fri Sep 14, 2018 11:06 am
by paroj
we are evaluating backporting the Metal RenderSystem from 2.1 to 1.11. So far it seems feasible - although some features like the dynamic shader generation (RTSS) might be more involved.
However in the end it depends on whether Eugene will join in, as Mac/ iOS is not my primary platform but his. (He also has a shipping product on those)
I guess community interest (as in contribution/ testing) will increase the chances. Probably I could also provide a Metal RenderSystem skeleton based on 2.1, but somebody else must do the platform fixup/ testing.

On the other hand Qt3D neither provides a Metal backend and I dont think Apple can just drop GL support like that due to the amount of legacy Applications - I mean Carbon is still around..

Re: Should I use OGRE?

Posted: Fri Sep 14, 2018 11:12 am
by JanKleemann
That sounds promising. Thank you for you reply.
I will continue to test out OGRE 1.11.