Cynthia

So uh, a few months ago I made a post on my Mastodon about being psyched about getting a very early Wii-U devkit (a white one "V4"), the first one that had been sold/dumped publicly. Well uh today I can say we successfully secured 4 (Factory/Nintendo Only) Wii-U devkits that are just as old. These units (lovingly referred to as "Ketchup" & "Mustard"), were never documented anywhere in any SDKs, contain internal only JTAG boards (it is not currently known if the JTAG goes to the Starbuck ARM CPU, or the Espresso PowerPC CPU), and have some extra goodies. I've documented a bit about them on the RGD Wiki: wiki.raregamingdump.ca/index.php/CA... (I think I've written all of this page at this point heh). We'll be working on dumping them, and bringing support to them to Sprig (https://codeberg.org/rem-verse/sprig -- our open source tooling for units like these). And before you ask "Do these units have hard-drives/Any Data?" Wii-U Devkits were commonly used in "PCFS" mode meaning they got all their files from a PC (these units are all PCFS), and switching between PCFS & not triggered a secure wipe. So the hard drives are empty (we checked). The internal nand is also likely empty of any games. There is a small chance we may get some factory firmware/system files, but it is being the most charitable we can be ~20% chance. I've been working on a remote dumper to be able to dump the NAND of PCFS mode CAT-DEVs, that we will use on these.

With how many projects and times have step like 15 turn into “okay now compile LLVM from scratch”, I should probably just start compiling LLVM as a step one, right alongside buying a domain name for it.

My partner dragged me to a chilis, and they had a braille menu??? Out of thousands of restaurants I’ve been to, the one who accommodated me was chilis???? Okay damn.

I forgot it was tomodachi day today, but luckily my partner let me know though.

@codecat I do not, and I’ve run into patterns at least in the past where Narrator has conflicted with NVDA behaviors so I generally try to stay clear. Also incase it’s helpful, every so often WebAIM will do a survey of a couple thousand screenreader users to try and identify support and trends. Not everyone, but generally a good “gut check” IMO: https://webaim.org/projects/screenreadersurvey10/ , and Narrator was at 0.7% back in 2024.

Happy to say I have acquired the first CAT-DEV v4 (one of the prototype Wii-U devkits), and even happier to say that the tools I’ve been developing for these devices mostly “just worked”.

This is not as useful as the “v2” cat-dev (aka “Toucan”) would’ve been, but is just as rare (there only ever being one that publicly was known to be sold). It is running firmware and has stickers from early 2012.

I was able to dump the usual suspects from the “MION” board (the board that holds the internal hard drive, and drives the debugger, power, etc. for the Wii-U board also in the case). The hard drive was empty unfortunately.

Unfortunately the device is in PCFS boot mode (meaning it requires a Pc to get its filesystem), and the older firmware does not support bypassing this like the newer firmware does. So we cannot power it on, or dump the internal NAND yet.

I’ve already been working on a dumper to remotely dump CAT-DEVs using PCFS. So the current plan will be, finish the remote dumper and test it on a modern cat-dev, then boot the v4 with PCFS using a modern CafeOS (hoping our PCFS implementation is compatible and the older FW doesn’t have any tricks up its sleeve), and then use the remote dumper to dump the internal NAND. The remote dumper should do this without any writes to the underlying NAND so we shouldn’t lose anything. Although to be clear there will almost assuredly be no game data on the NAND, but we may get some earlier OS binaries.

Incase anyone was super curious about what didn’t work with Sprig, it’s really only the fact the MION does not support “extended” information queries, and does not use the MIONs internal parameter space. What this means is I have to go parse some incorrectly formed HTML (which I already had implemented but was not using), in order to determine what version this thing is, what boot type it’s in, etc. everything else, including the dumping of MION just worked. We’ll see if PCFS does though.

Just thought that it's funny to mention, the largest files on the Wii-U CDN by far are from the game "One PIece Unlimited World Red" which means not only is it the longest running anime, it also eats up the most storage space of Nintendo's CDN (pre switch era), as well as my mirrors.

Oh this like isn’t anywhere as bad as I thought. This all like makes quite a bit of sense. I definitely understand how there’s like a much longer learning curve I’m probably not super suited for, but the basics of shaders and buffers like, damn. I get that example now that overwhelmed me earlier??? Anyway guess I’ll put up a GX2 triangle demo for my toolchain tomorrow to make sure graphics keep working, but like that’s nowhere as bad as I was dreading.

After asking people for shader learning materials like 3 weeks ago. I've finally ported the GLSL compiler for the Wii-U to build on Windows/Mac/Linux, Packaged It, and wrote "Good Enough" blessed CMake, and I can actually finally go through these tutorials.

Wish me luck, as I try to make something appear on screen, and actually understand this world.

Me before finding out about the latest release of k8s: K8s is alright I guess, lots of cool people work on it, but it's not for me, and i feel like a lot of people use it when they don't need it, so i'm generally not a fan.

Me finding out the latest k8s release is called uwubernetes and has an adorable logo: Alright when are we deploying k8s.

@luna heyyyyyyyy 🔥