Hacking DSTWO and iSmart Linux: A petition.

[sirocyl]

So, earlier there was a release of Linux on the Supercard DSTWO. The author of the thread, and presumably the coder behind it, also posted a similar version for a similar card, the iSmartMM.

Problem is, there was no source code provided for the Linux kernel included, any GNU utilities (which I presume are unmodified from the Dingux distribution), or the bootloader/host which runs the Linux kernel and displays its output/audio to the DS.

Why it's such a problem, aside from the GPL license violation, is that we've had numerous[citation needed] reports of people bricking their SCDSTWO's by bugs either in an application, in the Dingux distribution, in the Kernel, or on the loader.

These are "killer" bugs, which should be fixed as soon as possible by anyone that's able to do so.
Unfortunately, tysong hasn't been seen since December 2012, over 60 days ago today, and beyond a reasonable statute of limitations for releasing the GPL source code (even though, GPL code should be released with the software binary, as soon as the software binary is released.)

We need to get two things in order to be able to fix, and expand upon this awesome development opportunity:

1. Source code for the bootloader/application/plugin launched from the DSTWO plugin menu
2. The Linux Kernel source code, or the necessary patches and a reference to the existing source.

I don't want to come off as some Free-Software lunatic raving about proprietary software being a sin while I sic the SFLC lawyers on whoever's responsible for not immediately releasing the source code.
I just want to develop cool stuff using the SCDSTWO and Linux/Dingux as a base.
And I'm sure the developers of us, want that too.
And I'm sure the users want the same.
So how about it?

 

[sirocyl]

Post reserved for any communications, errata, news and such, in the same style as the NDSGBA thread.

 

[Deleted member 319809]

As I understand it, the bootloader acts as a sort of BIOS and filesystem driver to boot the kernel, and if the bootloader is unstable, that compromises the kernel more than a kernel bug. Therefore, I selected the bootloader.

With the bootloader code fixed, efforts can then go on to fix the kernel, rebasing from the current code in Linus's tree and remaking patches as needed.

(I never understood people who don't post their code along with a GPL release...)

 

[sirocyl]

I agree with you, Nebuleon. It's harder to code a kernel around the underlying system's issues, than to fix the firmware itself.
I should know, I patched kernels to make them load on pre-OpenFirmware PPC Macs. MkLinux was a goddamn beast.

 

[sirocyl]

Okay, so I'm going to bump this.

It's been a few months since we've seen Tysong here, and the GPL source code is way overdue.
I already PM'd him, with no reply.

Right now, if we want any further development on this, one of two things will need to happen:
1. We keep asking about the source code and hope he'll eventually post it,
2. We take matters into our own hands and make our own bootloader and Linux kernel/distro, similar to how NDSGBA's lack of gpSP source brought us TempGBA.

Option 1 would get us the code for the bootloader and kernel as-is, with all of its flaws (of which, there are many).
Option 2 would let us build our own system, likely based on/around OpenDingux, with better support for the hardware.

Thing is, we'd need some development for Option 2 (and probably some reverse-engineering of the existing loader and SCDSTWO firmware).
I've been studying the architecture, the CPU, the SuperCard and its SDK a bit, but haven't had much to contribute.
It'll take quite some effort, and I can't do it myself in a reasonable amount of time, so if any developer's interested, feel free to join/take over.