Homebrew [WIP] 3DeSmume (DS emulator for 3DS)

  • Thread starter Thread starter shutterbug2000
  • Start date Start date
  • Views Views 734,814
  • Replies Replies 2,759
  • Likes Likes 153
I think we have 64MB to put a game there. Also we can create a HB that moves the game between the SD and the NAND. I would do it but I don't know how to do it because I don't know C. It would take some time to move the game though


Enviado desde mi iPhone utilizando Tapatalk
No we don't, and no we can't. Why do you think nobody did it? xD At first you have to read and search, not ask here. Then, program what you think you can do if really think you're capable to do it.
 
did u alredy test B2W2?
There's no point really. Games like Mario Party and Pokemon Pearl still have really low FPS. The most a test will tell you is if the game even runs in the emulators current state which no one will be using to play games normally
 
Now we just need an updated version on /r/3DSHacks the mods keep deleting the posts claiming it's not an important update even though there is nice progress

RIP

And it's kinda true. No offense intended, but making posts in reddit for every single update made to the code will get old fast. Specially considering they aren't final releases nor meant for general use yet. I'd only make that kind of post if there was a significant improvement (for example, dynarec or hardware acceleration so the games are playeable at good fps despite not having sound; and then when sound is finally added)

Make no mistake though, the progress that has been made so far is amazing, but most users in 3dshacks won't care until it's ready to be used.
 
And it's kinda true. No offense intended, but making posts in reddit for every single update made to the code will get old fast. Specially considering they aren't final releases nor meant for general use yet. I'd only make that kind of post if there was a significant improvement (for example, dynarec or hardware acceleration so the games are playeable at good fps despite not having sound; and then when sound is finally added)

Make no mistake though, the progress that has been made so far is amazing, but most users in 3dshacks won't care until it's ready to be used.
I understand that side completely. That is the main reason I haven't posted every single update video there :p

From the looks of the replies however they did seem to care about this a lot still.

Either way whats done is done, the whole goal was to get more attention to this project and more people aware of it and it did it well.
 
@MasterFeizz you wrote a new rasterizer but was that in software or hardware? I've got a day off on work and was thinking of contributing. If graphics is being worked on I might busy myself with simple menus. That does bring me to a question: say if I were to write a loading menu instead of just automatically loading from game.nds, is there a way I can detect if the application is running from a CIA? The menu could be skipped and load from the romfs instead, leading to a "packaged" DS game like the Blarg "VC".

Beyond that I need to look at libctru and figure out how to scan a filesystem. Anyone know a repo of a file explorer I could look at?
 
@MasterFeizz you wrote a new rasterizer but was that in software or hardware? I've got a day off on work and was thinking of contributing. If graphics is being worked on I might busy myself with simple menus. That does bring me to a question: say if I were to write a loading menu instead of just automatically loading from game.nds, is there a way I can detect if the application is running from a CIA? The menu could be skipped and load from the romfs instead, leading to a "packaged" DS game like the Blarg "VC".

Beyond that I need to look at libctru and figure out how to scan a filesystem. Anyone know a repo of a file explorer I could look at?
I didn't work on the rasterizer, but the latest version of desmume does allow for threaded rasterization which I did test it and doesn't benefit much. Honestly the rasterizer isn't the slow part, the interpreter is. You could try porting the jit or threaded interpreter from NDS4droid.
 
I didn't work on the rasterizer, but the latest version of desmume does allow for threaded rasterization which I did test it and doesn't benefit much. Honestly the rasterizer isn't the slow part, the interpreter is. You could try porting the jit or threaded interpreter from NDS4droid.
I'm just curious: What is the program language you are using? I'm thinking of try to learn C or C++
 
I didn't work on the rasterizer, but the latest version of desmume does allow for threaded rasterization which I did test it and doesn't benefit much. Honestly the rasterizer isn't the slow part, the interpreter is. You could try porting the jit or threaded interpreter from NDS4droid.
Those would probably be the best performance gains we could get right now. I'll take a peek at the nds4droid repo and see if I can make heads or tails of it. I'm guessing it's written with the Android NDK so hopefully porting won't be too hard. Do we have access to dynarec with 3ds homebrew? Might be worth trying to implement after porting the threaded interpreter.
 
Those would probably be the best performance gains we could get right now. I'll take a peek at the nds4droid repo and see if I can make heads or tails of it. I'm guessing it's written with the Android NDK so hopefully porting won't be too hard. Do we have access to dynarec with 3ds homebrew? Might be worth trying to implement after porting the threaded interpreter.
Whith elevated privileges its possible, and the threaded interpreter will only benefit the N3DS.
 
Whith elevated privileges its possible, and the threaded interpreter will only benefit the N3DS.
I'm guessing the O3DS just doesn't have the CPU mettle to get much benefit from the threaded interpreter? How much is the JIT going to benefit O3DS, then? I always figured this would be more of a N3DS project, but I'm interested in the reasons.

Looks like nds4droid is using a different version or at least a reasonably altered version of the desmume core. I think I've got a general idea of what's going on. I need to port over the JIT and interpreter code, and then those eventually go into NDSSystem via armcpu_exec_inline from what I can see. I wish GitHub had a search in file function. Well, I can certainly take a stab at it.
 
I'm guessing the O3DS just doesn't have the CPU mettle to get much benefit from the threaded interpreter? How much is the JIT going to benefit O3DS, then? I always figured this would be more of a N3DS project, but I'm interested in the reasons.

Looks like nds4droid is using a different version or at least a reasonably altered version of the desmume core. I think I've got a general idea of what's going on. I need to port over the JIT and interpreter code, and then those eventually go into NDSSystem via armcpu_exec_inline from what I can see. I wish GitHub had a search in file function. Well, I can certainly take a stab at it.

Hey, want to get on the discord server? It's in MasterFeizz's sig. :)
 
I'm guessing the O3DS just doesn't have the CPU mettle to get much benefit from the threaded interpreter? How much is the JIT going to benefit O3DS, then? I always figured this would be more of a N3DS project, but I'm interested in the reasons.

Looks like nds4droid is using a different version or at least a reasonably altered version of the desmume core. I think I've got a general idea of what's going on. I need to port over the JIT and interpreter code, and then those eventually go into NDSSystem via armcpu_exec_inline from what I can see. I wish GitHub had a search in file function. Well, I can certainly take a stab at it.
Any luck on something? :)
 
  • Like
Reactions: laharl22

Site & Scene News

Popular threads in this forum