It's not that long ago that I found myself with a box of 8" floppy discs from a legacy product and no way to read them. Yes, the software on them was long obsolete. But I would have liked to be able to preserve a bit of company heritage.

The product in question (Glenayre GL-3000) had been updated in the interim to use 3.5" floppies, though with a bespoke format. I figured out how to use Linux and creative parameters to dd to write disc images. We packaged this as a bootable CD for customers to write their own disc images. After a sharp drop in floppy quality around 2005 I discussed other storage options with my boss (e.g. USB) but the business case just wasn't there.

...laura

I presume it varies greatly based on your area and task.

If you are slapping together a thin generic webui over a milquetoast sql database in a boilerplate-heavy language/framework, then sure I could see massive speedups.

In my particular area, the most unobtrusively useful enhancement is letting it take a crack at a 'code review' before I push it for real. One time it did catch something that would have gone unnoticed that wouldn't have come up for a long time and then it would have been annoying. However earlier today it started going nuts highlighting code that I hadn't changed and insisting that all the variables were named 'dict' and that was a bad idea and should be renamed. Nothing was named dict, the word dict didn't even appear in the codebase it was looking at.

If getting started on something unfamiliar, I *might* do a prompt and then reference that for things to potentially look up. I first started trying to do that and fixing up the result, but ultimately decided that outcome from prompt was harder to salvage than to just throw out and maybe use it as a reference.

I have had moderately more success in letting it predict the next few lines, though it often gets very opinionated about something very wrong. It also tends to assume incorrect things about interfaces that I deal with, interfaces that *should* have been verbatim in their training material.

In my experience it is, how effective it is is directly proportional to preexisting project complexity when the commands are run. The bigger the project, and the more parts that are interfacing together, the worse it performs. But for small, simple projects and creating frameworks, it can be amazing.

I'm not sure what "Building the Metaverse" is supposed to even mean anymore. Is he still obsessed with Ready Player One fantasies?

I mean, if he's just talking about generating 3d assets and the like, then maybe? AI 3d model generation is pretty useful if you don't care about every tiny detail matching up to some specific form. For example, I used an AI tool to make an image of an ancient mug with cave-art scrawled around its edges. It got the broad shapes of the model right, but had trouble with the fine engravings, making a lot of them part of the texture rather than the shape, but overall it was good enough that I just left off the engravings, had it generate a mug without them, then re-applied them with a displacement map. It got all the cracks and weathering and such on the mug really nice, and the print came out great after post-processing (cold-cast bronze + patina & polishing).

(I ended up switching from cave art to Linear A, because I also plan to at some point make a Linear B mug so that I can randomly offer guests one of the two mugs, have them rate it, and thus conduct Linear A-B Testing)

It's a cute "attack", but not practical in the real world. The "poison" stands out like a sore thumb on your loss graphs.

It's so easy to get tempted into feature bloat these days. You need a microcontroller for some simple set of features, like doing PWM control on a fan and handling a rotary switch, so you get something like a Seeed Studio XIAO ESP32S3 that's the size of a thumbnail and costs like $10, but then all of the sudden you have way more processing, memory capacity, pins, etc than you need, and oh hey, you now have USB, Bluetooth, and WiFi, and surely you should at least do SOMETHING with them, right? But the hey, for just a little bit of extra cost you could upgrade to a XIAO ESP32S3 Sense, and now you have a camera, microphone, and SD card, so you can do live video streaming, voice activation, gesture recognition... .... it really creeps up on you, because there's so much functionality in cheap, small packages today.

The irony though is that nobody really seems to bundle together everything one needs. Like, could we maybe have such a controller that also has builtin MOSFETs, USB + USB PD charging, BMS (1S-6S) functionality, and maybe a couple thermocouple sensors? Because most small devices need all these basic features, and it's way more cost, space, weight and effort to integrate separate components for all of them. The best I've found is a (bit overbuilt) card that has USB + USB PD (actually 2 of each, and reverse charging support), BMS support (1-5S), one thermocouple sensor, and a small charging display - but no processor or MOSFETs.

You can add how they've recently reclassified transgenerism (ne gender dysphoria) to the reasons why people have lost trust in the mainstream medical/psychological profession, as well.

Psychology has always been prone to un-scientific activities, but it's become increasingly bad with the wanton politicization of diagnostic standards, on top of the un-scientific approach employed in making most individual diagnosis.

It seems it can (although whether it's economical to do it at scale is another question).

It's been mangled by culture.

Once upon a time, it was unambiguously a pretty debilitating mental state. If you had that diagnosis, everyone could see issues and it wasn't at all something that anyone would aspire to.

Then Asperger's came along and thus began the 'diagnosis as an excuse for selfish behavior'. The general impression was "a smart person who has a tendency to be a jerk", which sounded totally awesome to a lot of people. They didn't need to try not to be a jerk, they had a pass in the diagnosis. People *wanted* this diagnosis.

Then, at least in part, some felt that Asperger's had become a very coveted 'diagnosis', and self-diagnosis was popular. They said 'oh, you know what, maybe if we group it with general autism, maybe people would be more reluctant to want that association, and it can go to being an aid for those that needed it.

But no, bereft of their diagnosis, they would instead do the same with autism, really diluting it and making a lot of people end up not taking autism seriously.

Nowadays, Gen Z highly values 'neurodivergent' as a badge of honor, that anyone cool *must* be neurodivergent.

So we end up with everyone saying they have a diagnosis, that they are neurodivergent, and they absolutely are not anything so pedestrian as 'normal'. Meanwhile those that really need it are generally taken less seriously because it's been diluted so much.

Looks like the displays have something like a 128x78 'pixel' active LED display as a backlight, and then put an LCD on top of it.

So if a tiny region of the display is just dim reds, then it can get a backlight that is doing just that and the LCD doesn't have to block as much other stuff.

No, this is still backlit LCDs.

The LEDs are still 'just' a backlight, but now a colored backlight. You basically have an OLED-like characteristic of emissive lighting at some resolution. The problem is the resolution of these LEDs would be something like a 128x78 display. Impossibly low even by old fashioned 'SD' standards.

So you have a 128x78 active LED display, and then an LCD panel on top to give it resolution. So you get to pick a good tiny local backlight color and minimize how much extraneous unwanted color that tiny dimming zone needs to filter out.

None of the woo "hurt him in terms of accelerating cancer". The problem is that he did that instead of actual treatment, and let a highly treatable cancer turn into a nearly-untreatable one.

AI has invented a new variant of Pokemon Go. Gather points by taking virtual photos of imaginary destinations!

But less convenient than version numbers, particularly since Ubuntu uses very predictable versioning.

So I know that even numbered years are LTS and the version number is YY.MM, and the month is always April for LTS and October is the other possibility.

So with that all in mind, one says "ok, I know I need to add stuff for Ubuntu 24.04 to this configuration". Except some configurations don't do version number and take the codename. So now I've got to remember 'noble'. Canonical themselves in their web site sort of de-emphasizes the codename. The 'tag' results for the blog all fixate on the version number. The download page doesn't mention the codename. The release cycle page does, and the *original* blog announcement mentions it, but not the subsequent ISO refresh release announcements.

I'd say the likely scenario is that the person actually buys stuff but doesn't consider the stuff an 'investment'. I bought a house to live in, not to turn it around for a profit.

To the extent people are 'investors' in things like 401k, they may not be 'active' investors and would just as much prefer something like a massive expansion of social security instead of letting investment companies play with their money. Or to the extent they do want to 'invest', they actually want to contribute to the potential success of things they intrinsically want to succeed, rather than chasing the best percentage return without regard for anything intrinsic to the people using the money invested.

The sentiment I think is plain enough, that they don't like the thought of handing their money over to a group of folks that will mostly enrich themselves above all else while their money is used for who knows what without regard for his deeper consideration of what is going on.