I once accidentally dropped a phone from several feet up so that it landed flat on its screen on a tile floor. No part of the phone itself broke (not even the screen, due to the way it landed), but the microSD card developed a crack across the middle (because the slot was positioned so there was a gap on each side of the card) & was unreadable. (Thankfully, I had just backed it up a few days before.)

I, for one, love my 28" 4K monitor. ~157 DPI was a vast improvement over the 96 DPI my previous monitor had (especially with text, but still visibly so with images & video), & even then, I can still see the pixels in large solid-color areas & especially diagonal lines (even if antialiased). (I think I sit around 2 feet from it on average.)

A multiple of 96 DPI would be a significant improvement for integer scaling of content (unless X & GTK decide to get their act together). But the overall screen size is also nice, so 8K at 288 DPI (about 30.6") would be good.

(I doubt anyone will make a 4:3 or even squarer display with high DPI & good color reproduction any time soon, but a girl can dream.)

I thought we have spent years training people to ignore those sorts of nag screens as malicious nonsense...did Microsoft not consider that maybe they should have gone with a different presentation?

The delay thing can actually cause problems. One of the times I tried to dial a number that had 911 as a substring as a child (that I regularly called due to it being for some of my relatives), I apparently paused slightly too long before the next (& final) digit...good way to freak out an unsuspecting kid. (I suppose the rationale behind substring matches is that someone in an emergency might be panicking & accidentally press other buttons before dialing 911.)

If the test is trying to detect being sufficiently non-repetitive, encrypted streams should have no problem passing, because they should appear to be maximum-entropy. Whereas people are notoriously bad at being random, so gibberish will tend to be less random than meaningful content in some real language.

My guesses (without knowing about all the systems it supports):

Having a resolution that is a multiple of every supported system's resolution

Alternatively, having a high enough resolution that non-integer scaling is unnoticeable

Imitating quirks of the original displays (like the scanlines on non-backlit GBAs)

Within a year or so after I first got a smartphone I tested how fast I could type on it, & I got around 30 WPM with all the predictive features disabled. At the time, I could get about 60 WPM on a keyboard, which has since risen to about 80 WPM (in a typing test, which did not allow things like cursor navigation & erasing entire words), so the claim that being (somewhat) fast on a smartphone prevents being fast on a keyboard also seems unfounded (or at least not universally valid).

I managed to type in a dream I had on a TI-89 Titanium in the dark in a tent in a thunderstorm. You just need muscle memory of the keys' locations. (Granted, it has physically separate buttons, which makes it closer to a PC than a smartphone in that respect.)

(Why did I have a TI-89 Titanium but not a source of illumination...priorities. Why was I in a tent in a thunderstorm...archaeological dig in Russia.)

Companies that dislike emulators being made of their devices (e.g. Nintendo) never managed to argue this with respect to the software interface presented by a device (as distinct from the BIOS code & such), so why should a software interface presented by software be any different?

In Nintendo's case, I seem to remember they also tried to use mandatory inclusion of the Nintendo logo to keep people from even using their interface without their permission, which then failed in court. Which is different from implementing the interface, but I would think if they could compel people not to release emulators they would have long since done so.

The reason is B, basically...notice how later rows have more columns than earlier rows. The first row has only 2 elements because of the s orbital, then the p orbital adds another 6, then the d orbital accounts for the transition metals' block, & finally the f orbital gives the lanthanides & actinides. The table would be really wide if they were put where they "belong," & it would get even worse for the g orbital that should appear in the next row after element 118.

I had that feeling at least once when I fell asleep on the way home from school, & my mom drove somewhere I had never been before. I distinctly remember telling her I felt like we had driven through an interdimensional portal when I woke up (& wondered what would happen if we did not go back through wherever the portal was).

So perhaps sleeping while someone else drives you somewhere random would work even better than driving yourself?

I did not expect to like Mortal Engines based on the trailer, but my parents got me it for some holiday (I forget which...maybe Easter?), & I ended up liking it.

Wikipedia says singular "they" dates to the 14th century, although that usage was discouraged (but prevalent) in the 19th century before once again becoming more widely acceptable (& even preferred in some cases) in the 20th.

I do still occasionally need 16-bit support (usually for running DOS programs that for whatever reason do not have Linux equivalents—e.g. to restore a backup in a discontinued format). I have ended up using QEMU because VM86 is not supported in long mode.

I was previously unaware of it, but apparently (I just checked sandpile.org) 16-bit protected-mode tasks are supported under long mode (which does not help with DOS).

Not only that, but Q(uick)BASIC & Visual Basic by default started numbering at 0...the important difference from C is not the lower bound but the upper one: in C, the number of elements, while in those BASICs, the last element's index. They also had a variant where you could specify both the lower & upper bound when declaring an array. E.g. DIM X(-7 TO 7), which was sometimes convenient for things like sprites.

I learned to program using various BASICs (mostly QBASIC, TRS-80 CoCo BASIC, & occasionally Applesoft BASIC.) I still occasionally use QBASIC, because I do not know of any good way to quickly test graphics code in the language I usually use these days (Haskell). But of course when I do (or when forced to use C), my coding style is influenced by functional programming.