Homebrew The Truth behind GBA emulation

  • Thread starter Thread starter Normmatt
  • Start date Start date
  • Views Views 39,864
  • Replies Replies 168
  • Likes Likes 1
Status
Not open for further replies.
The possibility or impossibility is irrelevant, without a demonstration it hasn't happened.

Oh yeah and HAY GUYS I GOTS NINTENDO 64 RUNNING ON MY DS! FOR REALS! I JUST CONNECTED THE PROCESSOR MATRIX TO THE FLUX CAPACITOR AND CUT THE N64 CART CONNECTION DOWN TO THE SIZE OF THE DS THEN PLUGGED IT IN AND IT TOTALLY WORKS YOU SHOULD BELIEVE ME.

The difference is that Normmatt is well known in the DS Scene.
 
I believe Exophase. He did an amazing GBA emulator on PSP that's near full-speed (gPSP, if you don't know it), so he knows alot about the GBA architecture and about the possibility of doing something like that.
 
If the next DS doesn't have a slot 2 then this is probably gonna be important..
Otherwise its a waste of time.
 
I... don't know what this means. Has someone done something groundbreaking, or have they just sat down and put the hours into something that others decided would take far too long and too much effort to be worth doing? I'll admit the technology's way over my head (I read Exophase's post 3 times and still don't get it), so I can't tell.

The developer doesn't want to release the software for piracy conerns, that's perfectly fine and all you people who think it isn't can go to hell. But I have to wonder, what was the point of it all? Did he have some kind of legitimate use in mind while developing (for the concept by deriving some other application from the same principles, even if the scale was only to be used by himself) or was this just an exercise in seeing if he could do something no-one else has done before? I know some developers think like that, I suppose they're the researchers of the future. But is it really worth it if there is no use for their findings? (dependant on the answer to paragraph 1 I guess)

and no I'm not changing the picture, it was true at the time it was made god damn
tongue.gif
 
[...]or was this just an exercise in seeing if he could do something no-one else has done before? I know some developers think like that, I suppose they're the researchers of the future. But is it really worth it if there is no use for their findings? (dependant on the answer to paragraph 1 I guess)
I think that's the right answer. He did it just to prove that it was possible.

Nonetheless... I'm not asking for a release, but a video would be nice
tongue.gif
 
I... don't know what this means. Has someone done something groundbreaking, or have they just sat down and put the hours into something that others decided would take far too long and too much effort to be worth doing? I'll admit the technology's way over my head (I read Exophase's post 3 times and still don't get it), so I can't tell.

Sorry, let me try to be clearer.

What Normatt's saying is impossible.

Whatever anonymous programmer he's referring to isn't running GBA ROMs, commercial or homebrew. If it runs off of a cart then this isn't going to work. Maybe some results were seen with multiboot games. That isn't the same thing. Or, he has some results playing at a very low speed. That isn't worth anything.

You won't ever be able to run real GBA games at an acceptable speed using this method.

They would run much more slowly than if done using dynamic recompilation, which is a traditional emulation technique. Even with this technique it is very unlikely that GBA games could ever run at full speed on Nintendo DS via slot 1.

If Normmatt or the anonymous programmer would like to prove me wrong I'm open to counter arguments.
 
I believe Exophase !

Just think about it again... 4 MB RAM thats not enough for even caching the stuff fast enough, so how do you want to process it at the same time ?
 
I... don't know what this means. Has someone done something groundbreaking, or have they just sat down and put the hours into something that others decided would take far too long and too much effort to be worth doing? I'll admit the technology's way over my head (I read Exophase's post 3 times and still don't get it), so I can't tell.

Sorry, let me try to be clearer.

What Normatt's saying is impossible.

Whatever anonymous programmer he's referring to isn't running GBA ROMs, commercial or homebrew. If it runs off of a cart then this isn't going to work. Maybe some results were seen with multiboot games. That isn't the same thing. Or, he has some results playing at a very low speed. That isn't worth anything.

You won't ever be able to run real GBA games at an acceptable speed using this method.

They would run much more slowly than if done using dynamic recompilation, which is a traditional emulation technique. Even with this technique it is very unlikely that GBA games could ever run at full speed on Nintendo DS via slot 1.

If Normmatt or the anonymous programmer would like to prove me wrong I'm open to counter arguments.

It's possible he does have some kind of proof of concept, but like you say it won't be on a flashcart. I'm thinking it's possible he's got it up and running a DS emulator (DeSmume perhaps) where the emulation isn't 100% accurate and it's taking advantage of loopholes that don't actually exist on a real machine. It would also explain why using page buffering works quick enough for the code to actually page swap, the difference between a hard drive write and a flashcart write is immense.

As far as I know (and my technical knowledge is pretty limited) it's impossible to access both the GBA and the DS bios at the same time, and once you switch to GBA mode it shuts the DS bios off completely - therefore making it impossible to run DS code and passthrough data to the GBA bios, as is hinted at by the "hypervisor" technique. Wouldn't you still have to emulate the GBA bios AND the actual GBA CPU (as there's no access to the GBA stuff) for the hypervisor to have something to pass data to?

I think the problem that people are having here is that they're looking at the DS part of the DS as an "add-on" to the GBA whereas they're both two entirely different machines inside one case. Isn't it like the GB/GBC on the original GBA's where the actual GB/GBC circuitry was built into it?
 
For me it's not about weather it can be done, it's why. Why would you spend so many hours on something people already have overcome using slot 2? Even if you sell it for money, you're main audience is the pirate scene, which means that the chances of being pirated rather than bought is almost 100%. However, if you do this for self satisfaction and for dedication to the scene, then please let me bow to you.
bow.gif
If this was to happen, I am willing to go all the way to where you live and lick your balls for you.
 
I believe Exophase !

Just think about it again... 4 MB RAM thats not enough for even caching the stuff fast enough, so how do you want to process it at the same time ?

The new DS is rumored to have more RAM.
I stick with my statement. If the new DS doesn't have a slot2 then this is going to be very important.
If not, its a huge waste of time.
 
I believe Exophase !

Just think about it again... 4 MB RAM thats not enough for even caching the stuff fast enough, so how do you want to process it at the same time ?


The new DS is rumored to have more RAM.
I stick with my statement. If the new DS doesn't have a slot2 then this is going to be very important.
If not, its a huge waste of time.

If the new DS doesn't have a Slot-2 then this is completely pointless. If it doesn't have a Slot-2 on it then it won't have the GBA internals in it, rendering this technique completely pointless.
 
I believe Exophase !

Just think about it again... 4 MB RAM thats not enough for even caching the stuff fast enough, so how do you want to process it at the same time ?


The new DS is rumored to have more RAM.
I stick with my statement. If the new DS doesn't have a slot2 then this is going to be very important.
If not, its a huge waste of time.

If the new DS doesn't have a Slot-2 then this is completely pointless. If it doesn't have a Slot-2 on it then it won't have the GBA internals in it, rendering this technique completely pointless.

then this is pointless either way...
 
If the new DS doesn't have a Slot-2 then this is completely pointless. If it doesn't have a Slot-2 on it then it won't have the GBA internals in it, rendering this technique completely pointless.

It has to have the GBA internals, all of them are part of the DS design. If it doesn't have slot 2 then they'd be forsaking GBA compatibility just to save a little space and very little cost.

I'd say it was a really bad idea if I didn't know Nintendo would manage to make it sell somehow anyway.

Either way you won't see GBA games playing on it through some other method unless they add new hardware to rig the cart bus to dedicated memory for it, which would be weird.
 
If the new DS doesn't have a Slot-2 then this is completely pointless.  If it doesn't have a Slot-2 on it then it won't have the GBA internals in it, rendering this technique completely pointless.

It has to have the GBA internals, all of them are part of the DS design. If it doesn't have slot 2 then they'd be forsaking GBA compatibility just to save a little space and very little cost.

I'd say it was a really bad idea if I didn't know Nintendo would manage to make it sell somehow anyway.

Either way you won't see GBA games playing on it through some other method unless they add new hardware to rig the cart bus to dedicated memory for it, which would be weird.

I don't doubt that Nintendo will keep the slot-2 cart as backwards compatibility is a major selling point in consoles these days, it's just the original poster mentioned that if the DS2 (which I doubt we'll see for a long time yet) had no slot-2 then the technique would be useful.
 
Status
Not open for further replies.

Site & Scene News

Popular threads in this forum