addie

Imperial came about as a system of units by measuring "everyday" things, and it remains pretty good for that. When you step outside the everyday, then it absolutely sucks - science deals with a lot of things that are too small, and engineering deals with a lot of things that are too large.

When I used to work in the water industry, working out how much chlorine is required to dose a hundred million litres of water per day at 0.5 mg/l, and therefore when I'd have to place an order to refill our fifty tonne storage tank, is easy enough to do in my head. If we were working in imperial, I'd have converted it to metric first and then estimated it.

On the other hand, metric calculations for pressure suck. If I weight 160 lbs and my bike tires are at 80 psi, then I have about two square inches in contact with the ground. If my car weighs 2500 lbs and its tires are at 30 psi, then each tire has about 20 square inches in contact with the ground. If I wanted scientific accuracy, then sure, I'd do it in metric, but I'd check the end result in imperial.

There's near enough five thousand feet in a mile - if you need more accuracy than what you can do in your head, do it in metric with a calculator.

Abstraction is not very compatible with concurrency, so as well as your your beautiful abstract API, you also need some 'cut through the layers' functions to return the underlying classes you need to synchronise on. Now you have a right mess that's incredibly hard to understand, infuriating to debug, and impossible to refactor. Best you can do is put another layer of abstraction on top. Repeat every six months.

If you think you might want to leave early, more polite not to disturb the other patrons. If there were any other patrons, of course.

5G is for spreading the woke gay mind virus. Collecting all of your personal information is the Jewish space lasers. Fortunately, tinfoil hat stops both.

The Webb produces some beautiful pictures, as always, but identifying 800k galaxies in an area 2 1/2 times the size of the moon is hard to conceive. Both how good a telescope it is, and the scale of the universe.

Don't think it says it in the link, but if you assume that all galaxies are randomly oriented, then in the places when the distribution isn't quite average, you can assume that light has been pulled by gravity's 'hidden hand'. And with nearly a million galaxies to analyse, you get a very good picture of how sources of gravity are distributed.

Google Stadia wasn't exactly a responding success...

From a previous job in hydraulics, the computational fluid dynamics / finite element analysis that we used to do would eat all your compute resource and ask for more. Split your design into tiny cubes, simulate all the flow / mass balance / temperature exchange / material stress calculations for each one, gain an understanding of how the part would perform in the real world. Very easily parallelizable, a great fit for GPU calculation. However, it's a 'hundreds of millions of dollars' industry, and the AI bubble is currently 'tens of trillions' deep.

Yes, they can be used for other tasks. But we've just no use for the amount that's been purchased - there's tens of thousands of times as much as makes any sense.

We've had multiple instances of AI slop being automatically released to production without any human review, and some of our customers are very angry about broken workflows and downtime, and the execs are still all-in on it. Maybe the tune is changing to, "well, maybe we should have some guardrails", but very slowly.

Unfortunately, server RAM and GPUs aren't compatible with desktops. Also, NVidia have committed to releasing a new GPU every year, making the existing ones worth much less. So unless you're planning to build your own data centre with slightly out-of-date gear - which would be folly, the existing ones will be desperate to recoup any investment and selling cheap - then it's all just destined to become a mountain of e-waste.

100% of supercomputers, 80% of mobile devices (as Android), 4 or 5% of desktops depending on whether you count ChromeOS. Desktop share is a few percent higher if you just count gaming PCs, eg. the Steam survey, since it's more widely used at home than on business machines.

The rate of adoption is accelerating, too - slowly but steadily.

Oof. The second-hand market is full of stuff that businesses are throwing out since they won't run Win11, but which run Linux perfectly well. I've just recently replaced my NAS / home server with a £20 core i5 mini-PC that if anything is a bit overpowered for the job. Runs Mint desktop very nicely.

I'd imagine that if you're spending a hundred times as much, then you don't just have "web and office" in mind, though...

Just need a chutney that was developed in the 90s, and then we can make a delicious chutney+blue ciabatta sandwich that's all younger than you think.

Magnet link for the actual mod download? Then you just need some fairly standard forum software in front. Let the users host the decentralised data.