Freedo Gp2x

[Playdough]

Hay Jaguarandine, no matter what happens in the end (success or not) I just want to say thanks for your efforts liaising between us and the dev, and for the donations you've already made.
If this emulator does indeed get completed someday (and runs at a good speed) this would be a massive victory for the GP2X.

 

[imhotep]

Hay Jaguarandine, no matter what happens in the end (success or not) I just want to say thanks for your efforts liaising between us and the dev, and for the donations you've already made.

Ditto from me too. It's awesome what your doing, Jaguarandine.

 

[slaanesh]

When I first saw this I thought it was an April Fools Joke.. but that's next month!

This does sound very exciting and I'm eagerly looking forward to see how this pans out.

One thing I'd like to know is why is the Magic Eyes documentation such a big secret? Why would the chip manufacturers want to keep this to themselves? Why not share that info so that devs can make full use of the hardware? Sounds frustrating to me.

Anyway, good luck hope it comes together!

 

[Squidge]

Doesn't give me much confidence if you have doubts

So you're saying that he can't use Linux even if he wanted to because the source is closed?

Maybe I didn't make myself clear enough. He can certainly use Linux if he wants and keep his source closed.

However, if he decides to kill Linux, he can't use parts of the Linux source code in his emulator, or any other GNU GPL source code. Using parts of the Linux source code would help greatly in reading from sd card, and communicating with USB devices (as well as other things such as sound in the early stages).

So therefore, he would have to rewrite a lot of code, which will be not needed if he decides to leave Linux alone (which is basically why no one else has bothered killing linux as of yet - its really not that much of a resource hog, and greatly helps in the development cycle).

slaanesh: For the same reason everyone else keeps things under wraps and NDA's - money and competitors. They will hand over most of the information if you give them enough cash, but some things are just not for people outside of MagicEyes, or only for people they trust, simply because they want to keep it away from certain people such as competitors.

 

[DaveC]

Wish you luck, but I'm skeptical. For example, full support of the 2X's and the BoB's hardware without Linux is going to take a LOT of time and knowledge, along with a lot of rewriting of code (as since he's using NDA'd documents he can't use any Linux source or other GPL code, as that would require him to release his source too).

To me, it seems like a lot of theory, but little practical experience, but I'm willing to be proved wrong.

BTW, You can run native ARM code on the 2x across both cores without killing Linux (ie, no need to emulate CPU if the emulation target is also ARM), which means you have full access to USB, SD card, etc without having to rewrite, but it seems he's already gone down the other route...

I'll be very interested in watching this project

Doesn't give me much confidence if you have doubts

Well considering we don't even have *NES* running fullspeed at 200 MHz full speed 3DO seems a bit dodgy. The 3DO is much more complex than an Amiga and that is a choppy mess at 240 MHz. Then there is this "it will take a long time 6 months or more" thing. That just sounds like a way to make sure we forget about the grand claims by then.

Then there is the fact that fullspeed 3DO is *impossible*. Not just a CPU power thing but a SD card DMA problem. The 3DO relied ALOT on spooling from the CD in realtime. Since the MagicEye hardware is incapable of doing that at a good speed, it is impossible. Even IF a miracle was performed and the emulation was fullspeed, there would be all kinds of pauses and lags from the SD card reading. This whole thing reminds me of that guy saying he got Halo running in the GP2X and to give him money to release it

 

[Jaguarandine]

Squidge: Thanks for the info. I'll let him know why he should be using Linux.

Well considering we don't even have *NES* running fullspeed at 200 MHz full speed 3DO seems a bit dodgy. The 3DO is much more complex than an Amiga and that is a choppy mess at 240 MHz. Then there is this "it will take a long time 6 months or more" thing. That just sounds like a way to make sure we forget about the grand claims by then.

Then there is the fact that fullspeed 3DO is *impossible*. Not just a CPU power thing but a SD card DMA problem. The 3DO relied ALOT on spooling from the CD in realtime. Since the MagicEye hardware is incapable of doing that at a good speed, it is impossible. Even IF a miracle was performed and the emulation was fullspeed, there would be all kinds of pauses and lags from the SD card reading. This whole thing reminds me of that guy saying he got Halo running in the GP2X and to give him money to release it

I think NES emulation has less to do with pushing the limits of the GP2X and more to do with developer preference. I'm sure that if we had a decent coder who was interested in bettering NES emulation, then we would have had it already. It's just that no one is. It sucks, but that's the situation.

The 6 months thing has no basis on anything except Admin taking as much time as he wants. I'm sure that if he focused on it, he could have it done in little over a couple months. However, he's a busy individual with a family and a life. I don't want to put any pressure on him with this port. The comment was directed more towards the gp2x community suggesting that it would be good to have patience. If everyone manages to forget, great, it'll come that much sooner.

Sorry, on the speed issue, you're saying that the GP2X's SD card reader is slower than the ancient 2x-speed CD-ROM drive of the 3DO? Or maybe I'm looking at it the wrong way?

 

[DaveC]

Sorry, on the speed issue, you're saying that the GP2X's SD card reader is slower than the ancient 2x-speed CD-ROM drive of the 3DO? Or maybe I'm looking at it the wrong way?

Yes, Look to the Sega CD thread for info there. The lags on the Sega CD are due to SD card DMA hardware problems. BTW the SegaCD was only a single speed and it has trouble. You would have 2X the pauses and stutters on the 3DO.

 

[Yono]

fullspeed 3DO is *impossible*. ...SD card DMA problem.

Referencing Squidge's post from that same Sega CD Topic:

That said, it is possible to create a faster SD card reader - Rob Brown has created one, but it will take a lot of work to actually get it to run as part of the kernel (Rob's was written for HH mode), and its currently unknown how much faster it'll be with the rest of linux leaning on it.

Considering that the developer has extensive documentation and plans to use a HH structure of his own making, it is unknown what is possible from him. I understand your skepticism, but I don't think we really have a clear picture of what this developer is or is not capable of at this point, so really all we can do is sit back and wait.

 

[ste_167]

I'm going to enjoy reading about the development of this emulator, just as much as (hopefully) using it.
I must admit that I am very sceptical too, but this can be seen as a compliment to the programmer if he does manage to achieve his goals.

 

[___]

sounds interesting, though I have never been a great fan of the 3DO. Even if the project does not reach fullspeed at all, a proof of concept would be neat just to show off ;-)

 

[skiffex]

Hello everybody. I'm the one who's writing FreeDO GP2X Port.
Since i have very limited time. I'll be checking in once or twice a week. If you want to ask me any questions -- you can post them here. But -- not more than 2 questions a time (give me a break ... i need time to develop, not to answer the questions).

As for not using Linux.
FreeDO will be unloading OS completely in order to acquire as much speed and flexibility as possible. I will not use any linux code -- since i'm using this exact microcontroller in my other project -- majority of HW routines are written by myself. Also for your curious minds -- it will have it's own RTOS - Integrity by GreenHills -- i have valid royalty-free license (used in my other projects). All this will allow me to be flexible in multi-core environment.

 

[Playdough]

Just wanna be the first to say thanks for your efforts, and I look forward to reading about your progress!

 

[Rivroner]

Thank you for porting FreeDO to GP2X
^_^

 

[jimid2]

Many Thanks to you for making this effort! I'm really looking forward to watching this project evolve...

 

[KingCheese]

Thank you for spending so much time on developing for the gp2x, because I think it's a great handheld and it's worth.I also think that you're one of the BEST free devs round here and it's great you've joined the gp2x community.
Most people here don't want to put any pressure on anybody at all (me too) and THAT is the great thing, that it's free and joinable for everybody.
Good luck

 

[skiffex]

>>>Small status update:
Got preliminary application up and running.
Doing following on 2nd core:
-- video display line processor (VDLP) is working (displaying dumped screen, built from 5 bitmaps)
-- CEL processor working (drawing dumped 12 sprites)
-- DSP working (playing a sample from dumped FIFO state)
Main core is running idle loop.

TODO: ARM code translation.
ETA: 3-4 weeks

 

[naples39]

Sounds like you have your stuff figured out. Good luck to you on the project.

 

[Lazrhog]

Wow, and everyone laughed at me when I bought my 3DO !!! I still have all the games I bought sitting on the shelf next to me lol.

For the uninitiated my list of 3DO games is ...

FIFA International Soccer (Jaw dropping in 1995)
Striker world Cup Special (thanks Rage for the worst game ever!)
World Cup Golf (OMG it was slow)
Cannon Fodder (Brilliant)
Slam n Jam 95 (Amazing for its time - looking forward to this one)
Wing Commander III (yawn)
Return Fire (OMG OMG Brilliant game - best 2 player on 3DO)
The Need for Speed (The original and best - 15 minute long courses !)
Mad Dog McCree (sorry - what was I thinking)
Dragons Lair (now that could be cool on gp2x)


Please stop laughing at me :blink:

 

[Jaguarandine]

3DO was a great system. No reason to be ashamed
It had its good and bad like anything else. I purchased my 3DO secondhand after the fact (goodwill store $20 ftw!), because it was inhibitively expensive when it came out (I got a Jaguar instead). It had some great games for its time.

 

[DaveC]

I don't know why but I really liked POed and Monster Manor on the 3DO. No one seemed to like POed much but I beat it a few times. I loved the missile with a camera on it that you could steer to your target in real time. The jetpack rocked too.

If this emu is real and not a fake and it plays POed I would be thrilled with just that.