I find it interesting that "knowing how to change a tire" is such a big deal for so many people. Nobody ever taught me how to do it. I just found myself in a situation where it needed to be done, opened the car manual, and kinda just figured it out. The hardest part was figuring out how to unpack the folded up jack that came with the car, because I think that was the only bit that wasn't covered by step-by-step instructions.

"Most" !~= "All", and there's always a long tail. Sometimes the most critical software for users is some niche application that can't be bothered to update to deal with new compatibility issues.

And sure, Microsoft might create-and-abandon libraries all the time. But all the old libraries still work.

Only Apple has the luxury of actually breaking compatibility with new versions of their OSes, and that still causes problems with people who use software from stodgier companies that aren't part of the Apple RDF.

I blame this entirely on the design philosophy of many turn-by-turn navigation systems. That is to keep you as ignorant of your surroundings as technically possible, until 5 seconds before you need to make the next turn... If you're lucky. That's why I can't stand many such systems.

But its also why I really like the system Tesla puts in their cars. In addition to the usual turn-by-turn stuff, they also show you two things everyone else tries really hate to hide: A list of your next navigation steps, and a big top-down map showing your route. It actually gives you some situational awareness.

(* Yes, I know there are people who will complain about the actual directions. There are people that will complain about any implementation of that if its not perfect for their specific area in comparison to some other implementation they like better. But it works fine for me, and that big map means that its useful even if its not giving me turn-by-turn instructions.)

Combined with "WHY WON'T YOU SHOW ME ALL THE STORES IN THE SHOPPING CENTER I'M FOCUSED ON?!?!?".

I know designers love to reduce information density to make things look cleaner, but sometimes it can have a massive negative impact on the actual utility of the product.

This happens with everything Google does. When they come out with a new product, they have enough of a technical advantage that they're far superior than any competition can ever hope to be. This is such a lead, that the competition doesn't even try unless they serve some niche market.

Then, years later, Google loses interest in that product and it starts to go stale. But its still good enough that the competition never really takes off.

Then it finally goes poof, and everyone is scrambling to find an alternative, but nobody is actually prepared to take over.

When I'm doing something new or interesting, I often end up building it three times:

1. Build the critical functionality as throwaway code just to see if I can make it work
2. Build out the full system to get a feel for how all the pieces fit together
3. Rebuild the full system now that I understand it well enough to properly architect it

Of course depending on the circumstances, I may wind up combining two these iterations with refactoring, or not doing all of them on every project.

Unless you point a spectrometer at them.
Then they often have a huge blue spike, followed by a big dip, and then a gentle curve.
Or if you're lucky, and find one of the few brands that can do better than this, you still have at least a small dip after the blue. And regardless, the spectrum curve is always a bit spiky and uneven.

So sure, for normal everyday lighting purposes, you're absolutely correct.

But if you actually need something whose output can be modeled by a blackbody radiation spectrum, then not so much. Sure, most of these cases fall in the realm of scientific experimentation and material analysis, but they do still exist.

I keep getting the impression that Canadian tech wages are already depressed enough, which is why anyone who can tends to flee Canada to come work in the US. At least for people originally from Canada.

Also, keep in mind that EVs have much more sophisticated battery management electronics than a laptop, all else being equal (and obviously all else isn't equal). This probably makes a huge difference in battery lifetime on its own.

The more frustrating part of road trips really isn't supercharging along the route. Its destination charging once you get there. With a big enough battery, one strategic supercharger stop near the end of the trip can overcome that problem.

Overcoming this whole sentiment is pretty much why Tesla won out over the first-gen Nissan Leaf and all of its peers. All of those cars had the range appropriate to make the OP happy. You know, "good enough" for some high percentage of driving needs, but nowhere near good enough to actually be your only car.

Meanwhile, Tesla had the range to actually be useful as a normal car. And so now everyone else sees that as important too. (Though the rest of 'em kinda ignored the charging network problem until very recently.)

You can't ignore the "pain in the arse" factor when it comes to rental cars.

Earthquake insurance in CA is an interesting case. The premiums are very high, for starters. But the deductible is also so high that its pretty much useless insurance unless the earthquake causes your home to be a total loss. And at that point, you'd expect it to be a wide enough catastrophe that the government would have to step in.

As a result, most people simply don't bother with earthquake insurance. Instead, they lean on modern building codes and construction techniques and hope their homes are designed to actually withstand most of the shaking that's actually likely to happen.

Of course this could also be seen as a double-edged sword. If you craft the policy such that nobody can justify getting it, then the pool of insured will be so small that it just feeds back and makes the policy even more useless.

But that's kinda missing the point... Which is that you still need Windows. Sure, it might be contained within a VM, or on a separate computer, but its still there, still needs to be maintained, still have to deal with updates and whatever gunk MS decides to add this week, etc.

In other words, just because you're using Linux as your primary OS, doesn't mean you can use it as your only OS.

Yeah, the above dump was from the Brother p-Touch software.
I have both Brother p-Touch label printers and a Zebra label printer. The software for the former runs on Windows/MacOS, whereas the latter is only Windows.

The Brother really needs dedicated software that understands its features and media sizes.
Fortunately the Zebra can also behave like a normal printer, which is great for printing things like shipping labels. But a lot of the time I actually do still want traditional label-design software to use with it, for all the features that provides.

I'm not aware of any good way to do this, for either printer, from Linux. The best you ever find is someone failing to understand the problem, detailing cryptic ways of configuring CUPS that don't actually work 75% of the time. But there's still nothing out there for actually designing the labels. No, a word processor template doesn't count. Neither does some half-built unmaintained project like gLabels.