i do have a question regarding the gfx-engine of unity. Recently i was looking at the new release of ogre (Dagon) and whilst watching the demos (shadows, ocean, compositor, …) i thought about what kind of advantages and disadvantages there would be if otee would use ogre instead of their own gfx-engine. I mean you’re already relying on middleware in the fields compiling (.net) physics (ageia) and audio (oal) why not doing the same with the gfx-engine and freeing the ressources for porting the ide to windows, improving the ide/workflow, making things as simple as possible for the users, enhancing game relevant stuff, ai…? There sure are ups and downs with this:
-) Time for integration, not having control, size for the plugin (in terms of filesize irrlicht would be better)
+) Lot’s of effects not only for PRO, shadows, freeing ressources in a long term view for other fields
I really would like to hear what you are thinking about this and what would be the ups and downs from your perspective.
Not to bash Ogre, but there is a huge difference in making some feature in a tech demo versus making some feature generally useable.
Take for example shadows: it’s a piece of cake (say, 2 hours of work) to implement “look ma! we have shadows!”. It’s relatively easy to make a cool looking tech demo (say, a week) like “look pa! we have lots of shadows!”. And it’s f*****g hard to make them actually work in a generic case (I’d say at least two months of fulltime work for a really smart guy).
In other words, never judge anything by watching tech demos. They just don’t work in real life.
Not to bash Ogre, but there is a huge difference in making some feature in a tech demo versus making some feature generally useable.<<<
Hmm i only used ogre (the former version) once for a small project and everything worked just fine but i also wasn’t going into depth so much so i really can’t judge this.
Sure it’s a difference showing something in a demo or making it a general purpose item but that wasn’t exactly my question except you wanted to say that ogre’s features don’t work fine in a general 3d engine scenario?
Until then i still would be interested in the ups and downs from your point of view.
Again you’re already using physx, oal, .net, sdl so i think it’s a reasonable question to ask if due to limited ressources this could be an option for you. Additional as there are features in ogre already which are misssing in unity today.
So…
( ) We will investigate this when we have some more time.
( ) We have no idea what ogre is capable of and so can’t answer this.
( ) Ogre just sucks but we don’t say why.
( ) We simply don’t wanna change our graphics engine.
( ) We don’t like such questions.
( ) ?
Ogre fits very badly with Unity. Remember that Unity is much more than an engine, and the editor part is very much connected to how the engine works.
Losing some features, gaining a few, and then having to spend a lot of time integrating, testing, polishing, and fixing other people’s code: sorry, it’s out of the question.
And remember that the features you see listed on the Ogre website are possibilities in the code, not necessarily nicely exposed or easy to use. See Aras’s comment above.
Remember that Unity is much more than an engine, and the editor part is very much connected to how the engine works.<<<
So much teethed? No abstraction layer between this?
Losing some features, gaining a few, and then having to spend a lot of time integrating, testing, polishing, and fixing other people’s code: sorry, it’s out of the question.<<<
As i already said i can’t judge this. I just thought that after the time for integration/adaption which you’ll have to invest you could spare time regarding a)implementing features they already have and b) further development of the 3d engine.
And remember that the features you see listed on the Ogre website are possibilities in the code, not necessarily nicely exposed or easy to use. See Aras’s comment above.<<<
Abstraction layers are for wimps! Sorry, could not resist… :roll:
But to second David: integration time is usually much larger than the actual implementation time. As with everything in software, you spend 20% of time implementing something, and then spend 80% time actually making it work. So just taking Ogre (or pretty much any other engine for that matter) would not get us very far.
So much teethed? No abstraction layer between this?[/quote]
Abstraction layers are for wimps! Sorry, could not resist…<<<
Without a lot of these you would not have been able to write these words… :O)
But to second David: integration time is usually much larger than the actual implementation time. As with everything in software, you spend 20% of time implementing something, and then spend 80% time actually making it work. So just taking Ogre (or pretty much any other engine for that matter) would not get us very far.<<<
Then i will be happy as we’ll get all the neat stuff like shadows, aa, deformable meshes, and lot’s more in a pretty short time, right?! ;O)