tbz wrote:
Yes, but Ogre is the basis of both performance-heavy apps and non-performance-heavy apps. Just because Ogre may be used for non-performance heavy apps does not mean it won't be used for performance heavy apps.
Regarding Ogre, sure. People would like it to be faster in various areas. This does not, however, mean that one has to use C to speed up everything in a 3D app, and meticulously avoid C++. There are plenty of apps even in 3D land where the minor performance differences between C and C++ are irrelevant. Even the major performance differences between C and Lua are irrelevant, sometimes. It's appropriate to script a modal dialog or minigame in a single player RPG. The sky won't fall. Nothing else important is going on with the computer. (Well, if it's a sysadmin bogging down a production server just so they can be entertained playing a single player RPG, such a person is not to be pitied! They can either find a clever sysadmin way to get more performance, or lose their job for goofing off.)
I did, however, say that 3D engines benefit from leaner coding styles and that C is more conducive to said styles.
What you said is
tbz wrote:
the coding concepts introduced to C++ result in slower code, period.
As the rest of us have established by now, you're wrong. The nature of your error, is symptomatic of obsession with optimization to the exclusion of other factors. If you had actually said "C is more conducive to said styles" in the 1st place, I doubt we'd have as much to argue about.
So, that's why Ogre gets outperformed by many engines, mine included? Right. I see your point! They have definitely proven it! (please, please, note the sarcasm!)
Where is your open source 3d engine, or shipped game, that I can evaluate? Note I do mean
open source engine. Your engine might run rings around Ogre, but if we can't use it with full control over the code, that does the rest of us no good whatsoever. Not an apples to apples comparison in that case, as Ogre would be providing something that you are not providing. Sure, doing everything "your own way" could have advantages to you personally. Doesn't mean squat in an open source context, if nobody else can make use of it.
If you haven't shipped a game with your engine, or at least produced a good chunk of a game with it, that the rest of the world can look at, then I don't care how great it's supposed to be. Until the rest of the world can see and interact with it, it's vaporware.
Nevermind how many other developers have shipped stuff with your great 3d engine. Ogre has a track record.
But none of the "how to speed up Ogre" stuff has a darned thing to do with C vs. C++
That is far too dogmatic of a statement to be true.
Actually there was 1 thing in all of the forum posts that does have to do with your thesis. An argument was made that commercial engines simply license code, rather than trying to support C++ plugin architectures. If developers want something in the code, they just modify the code, and don't worry about conforming to any external C++ class standard. Given the amount work I just did cleaning up the Ogre libs area on the wiki, I see merit in this. Because even with the plugin architecture, most people in open source land are not maintaining their Ogre lib projects. The
inactive projects list is really big. It looks like the ecology was a lot healthier 3..5 years ago, but now much has bitrotted. Since the plugin ecology for the most part doesn't exist in the real world, why bother to have it as foundational to Ogre?
Please stop with the straw men, I am finding it quite tiresome to constantly rebuttal your fallacious arguments.
Methinks thou protesteth too much. How about you go read those various Ogre performance improvement threads, then contribute something that would make Ogre faster in some narrow area? You have the technical expertise to do it, if perhaps not the ideological motive. But really, what's tiring is all these stones you're throwing at Ogre or C++, if you don't have your own better 3d engine project to offer anyone else, or a shipped game / game-in-progress that we can all give props to. Are you offering something other than ideology?