AmpBenzScientist @AmpBenzScientist@qoto.org

So the Architectures of the last 4 supercomputers might be surprising.

x86-64 with Radeon
Arm and something else.
Power9 (2 Processors paired with 4 Tesla GPUs)
Sunway

PowerPC is not dead and Sunway is believed to be most closely related to DEC but that's a stretch.

PowerPC, OpenSPARC and other open RISC designs aren't going away anytime soon.

It would have been easier to develop on a microkernel if I didn't pick four of them. My relationship problems in the past too.

Updating UEFI on a Thinkpad just feels like it's going to break. It does use TianoCore so that's something cool. It just keeps me updated with compatibility for the latest malware.

The PowerPC Architecture is supposed to be around into the 2030s.

I've tried another FSF Approved Distribution and it is surprisingly good. The test computer was a 4th gen Intel board with iGPU. Linux-libre did what it could to mitigate and it managed to do well. It's a 4rd Gen Intel CPU and microcode patches don't exist for all the vulnerabilities. 5 are mitigated by Linux-libre, 3 require microcode patches and one is just there. Perhaps there are other exploits but I didn't install this to be secure.

Performance is great and I would definitely recommend this to anyone who wants a FSF distro. It's Guix and I'm enjoying it. It's not like every other distro unless I'm using a terminal. I haven't compiled and tested my usual software for it yet. It feels very solid and mature. I used Trisquel for a few years and I might use Guix for a few years too.

Tl;dr I have been testing a FSF Approved Distribution and it is really cool. Skip their other distros and pick this one. It is one of the few Linux distros that really impressed me.

Mixed news on two old projects. One would just need traces cut and appropriate firmware.

The other project has a new volunteer that is older than the previous microcontroller. Older interface but a good performing controller. I think I could get some basic firmware built for it and flashed. I haven't checked out what communication protocol it uses but maybe it will function with open source.

That's progress.

I'm impressed with microkernels on a regular computer. It's a new experience but it's really flexible. I think it's an L4 Microkernel but I could be wrong.

Okay so I thought one project was going to be more difficult than it is. I'm already making progress on it. It's just going to require a few very close cuts to disable a troublesome feature. It should just disable that one thing. Physically removing it would be difficult but it can't be used without those connections.
That's disappointing but I should have expected as much.

It can be disabled and that's good enough for now!

It seems like I need to make a change in my tools if I am going to have any success in my projects. I was hoping that I would never need to go this route but I think it's called for now.

It's not optimal at all and I'd rather have a useable device the same as it was from the store. Enter different controllers, protocols and interfaces if needed. If the controllers aren't IP then the protocol is IP and maybe both of the above with the interface being IP.

At least it won't be challenged. Refuse to play their game. I'm going to revisit some former projects and see how valid they could be. I'm not above using a scalpel on traces or removing components. It's time to check them again.