All the best firms continvoucly morg these days.

It was a literal 100-level course project in my CS programme in 2000 or so.

You didn’t even do it with a programmed CPU, you used 74xx logic gates and counters wired on a breadboard

Floats for currency in a payments platform.

The system will happily take a transaction for $121.765, and every so often there’s a dispute because one report ran it through round() and another through floor().

Some mainboards have very few PCI-e slots.

I ended up with a similar adapter because the onboard SATA on my board was flaky with optical drives and I rip CDs.

I am disappointed it’s not a VLIW platform.

The Internet boom didn’t have the weird you’re-holding-it-wrong vibe too. Legit “It doesn’t help with my use case concerns” seem to all too often get answered with choruses of “but have you tried this week’s model? Have you spent enough time trying to play with it and tweak it to get something more like you want?” Don’t admit limits to the tech, just keep hitting the gacha.

I’ve had people say I’m not approaching AI in “good faith”. I say that you didn’t need “good faith” to see that Lotus 1-2-3 was more flexible and faster than tallying up inventory on paper, or that AltaVista was faster than browsing a card catalog.

I have to think that most people won’t want to do local training.

It’s like Gentoo Linux. Yeah, you can compile everything with the exact optimal set of options for your kit, but at huge inefficiency when most use cases might be mostly served by two or three pre built options.

If you’re just running pre-made models, plenty of them will run on a 6900XT or whatever.

I believe the huge mistake in HTML wasn’t having some sort of element-level addressability.

People went insane over “the page flashes for 15ms because we have to reload the header and footer and it doesn’t look NAAATIVE!” and the response was to SPA/AJAX everything, inviting a huge Turing-complete nightmare of possibilities when 95% of what peopleneed would be delivered with < form action=“blah” replace_with_response=“#foo” >

That and a dearth of native widgets-- a < combobox > and a < menu > that worked like the system menus might have kept JavaScript as the sick oddity it should be.

Grok went into a new conspiracy k-hole?

At this point Void feels like Slackware for the 21st century. It’s comprehensive and less full of “modern linux” hairballs than some others, but they seem pretty good on package updates. I like it being non-systemd as a first class thing rather than trying to backport it on an uncooperative parent distro.

This is big “if we break your old toys, you’ll HAVE to play with the new ones” energy.

Tell me when they port FVWM. Seriously. FvwmButtons-- a pretty trivial dock except it can swallow other windows-- seems like it would be out-of-bounds on Wayland unless it was owned by the compositor itself to access the other windows. I don’t see any of the new taskbar-tools used with Wayland compositors offering similar functionality (I could be wrong) and that seems an amazing loss of feature parity.

Try RiscOS for a glimpse of a world most of us missed.

Not to mention I bent all the pins on my Cyrix MII ttying to get it in that stupid socket.

I understand pre-OS X Macintoshes used colons.

I guess the assumption is more that for me, a fresh install is often about decluttering as much as anything-- the five Wayland compositors, three music players, and six pseudo-IDEs I tried and didn’t like don’t need to follow me to the next build.

In a conventional install, that just means “don’t check the checkbox in the installer next time”. In a Nix-style system, this is a conscious process of actively deciding to remove things from the stored configuration, no?

I suppose the closest I’ve gotten was recently migrating my setup from a desktop to a new laptop. Mostly copying over some config from my home directory, but even then, I wanted enough different stuff-- removing tools I don’t use on the laptop, adding things like battery monitoring and Wi-Fi control-- that it involved some reconfiguration.

I suspect the tooling isn’t quite there yet for desktop use cases.

If I were to try to replicate my current desktop in an immutable model, it would involve a lot of manual labour in scripting or checkpointing every time I installed or configured something, to save a few hours of labour in 2 years time when I get a new drive or do a full install.

The case is easier for defined workload servers and dev environments that are regularly spun up fresh.

No, this is a general practice-- I see it a lot with third-party vendors who want you to integrate with their services. They’ll expire the documentation portal password after 90 days, but the actual user facing service still accepts the same “password123” that’s been set since 2004.

I suspect the pattern is to protect the vendors from developer scrutiny: by the time you’ve jumped through enough hoops to read the docs and realize it’s trash, the execs have signed the contracts and the sunk costs are too high to bail out.

Also add another 6 months to actually get the credentials for the test environment.

ARM was designed because the 6502 was approaching end of viability, and Acorn (the maker of the BBC Microcomputer) needed a next-gen product. At the time, RISC was the trendy thing, and I suspect the 286 and 68000 were too expensive to adapt for their products; they weren’t pushing £5000+ workstations like IBM or Unix vendors.

It was light and small because they had a small team; low power was a happy accident.

I have a similar one, different seller and possibly submodel, but also a refurb HGST 12T enterprise drive. It sounds like I left a soda on my desk most of the time, subtly popping and ticking.

This is the firmware I’ve been working on. Basically I wrote it because at the time (early 2023) there wasn’t a “good” keyboard firmware like QMK or ZMK for the CH32V305. Now it supports keyboards, joysticks, and a rudimentary pointing device made out of a PS2-style analogue stick.

https://gitlab.com/hakfoo1/ch32v-keyboard/-/tree/fightstick?ref_type=heads

That branch has the mapping I used. Note this firmware has a keyboard-centric assumption that switches are wired as a matrix (between two sense lines), even if that matrix is 1x24, rather than just grounding a sense line individually.

The stick portion was one of those “Pandora Box” devices that was built into a cabinet and pre-wired to a crappy Android TV box.

I bought it because I figured it was probably cheaper than cutting a decent looking cabinet and buying the buttons off AliExpress. That also meant it came with a predefined cable harness to fit the Android box. In the hopes of making it tidy, and reversible, I ordered a little throwaway PCB that accepted the existing 40-pin plug and bridged it to a nanoCH32V305 breakout board. Of course, I made a design mistake, so the PCB had bodge wires, so not much was saved.

If you’re starting from scratch, you could direct-wire to the MCU breakout board.