I know you are taking a swing at me here, but let me say this:
I didn't specifically think of your name one time while writing that. In fact, I'm only aware of your work on GBA emulation.
If you happen to see a nice video plugin, would you like to rewrite a entirely new piece of software that mimics the behaviour of what that previous software is?
Again, it depends upon my intention. Do I want the games working now, at playable speeds, on modern hardware? Or do I want to understand how the hardware works at the lowest level, to document it for the future when Glide, OpenGL2, etc are all long since deprecated and not supported by drivers?
So why is a wrapper such a big fucking deal?
It isn't, and I never said that it was.
Then you talk about the problem with replicating graphics effects with hardware. Isn't that the nature of HLE, or do you know some better approaches?
No, like you said, that's the nature of HLE. The only alternative is LLE, which also has serious problems.
Or why don't we use a software renderer that goes < 20 FPS, and the general public yells at us for it due to quality and speed.
Again, the general public will bitch regardless. How about we do what we want, and ignore them?
Bye.
I seriously don't know why you're so pissed off at me. If you read my other posts, you'd understand my point.
We all have
opinions on how things should be done. Opinions don't mean shit, unless you're actually working on something. Am I working on N64 emulation? No, so my opinion doesn't matter.
1) HLE works for now, yes. It gets games playable now. But it has the problem that video card APIs
will become deprecated in time. Thus, it's a temporary fix. Do you agree or disagree with that?
2) LLE gets around this by implementing the raw drawing routines in pure math. Math will never become deprecated, so once we have a software implementation, it will work for all time. The problem is that it's ridiculously slow, and not
at all feasible for playing games today. However, it is the superior solution for when PCs are 5x faster than they are today. Do you agree?
I'm not saying that people using HLE are wrong. I can understand that nobody would want to write an emulator they couldn't even play because it was too slow for any existing hardware. I understand that HLE is the best (and let's be honest -- only) possible choice for
today's hardware. My only point is that it's not a permanent solution. Surely you can agree with that. Eventually, the work will have to be discarded. That, or you'll have to keep porting the Glide wrapper to newer and newer APIs into perpetuity.
Or perhaps it's best to just keep up with the wrappers until PCs are fast enough, and then focus on a pure software implementation at that time. I don't claim to have a magical solution to make everyone happy. PCs are too slow for low level emulation, so compromises are
mandatory. You're going to take slack from people no matter which option you choose, and there's nothing you can do but ignore it.
But I'm not the one being verbally abusive about your choice of compromise, and I respect your decision, so you have no reason to be upset with me. I'm just stating the very real, very obvious to anyone, compromises that are being made. If that upsets you, then it's your problem, and not mine.