Seriously though, it's a good format! It works!

I ended up doing this inadvertently earlier this week due to some medication side effects (yay akathisia!) making me want to go outside and just seriously run fast, including up the nasty hill nearby. Which I could, so I did. I'm amazed at how much it seems to have helped, for how simple it was. (Another one from the past, though for different situations: alternating hot and cold showers.)

Really, if you care beyond "blow the fuse if something shorts", you need active current limiting. For common cases like USB ports, you can buy chips that do it cheaply and efficiently. There are also some textbook circuits, though they each have their pros and cons. No matter what, if you've got a pass transistor or switching transistor that's about to go seriously overcurrent, you have to do something about it with active parts -- fuses aren't going to get it done.

But defense in depth is always a good strategy, and fuses play a key role there. The active circuitry saves the rest of the design; if it can't get the job done, the fuse makes sure nothing burns.

Even a fast fuse is very, very slow compared to semiconductors. I've seen transistors blow up to "protect" fuses. They're for stopping fires and preventing the slaughtering of batteries, nothing more, nothing less.

With the physical model you chose your software and it didn't stop working. (It might stop being useful, but that's a different story.) No option to patch later meant a company couldn't ship trash: word would get out and people just wouldn't buy it. (Windows ME, anyone?) Now you can get the rug pulled out from under you on all of your tools, and things "ship" as garbage with a massive "day 1 patch" (and a "day 2 patch" and...) and honestly often never get finished. At least not before they're "retired" and you're "updated" to some dumpster fire that's nothing like what you wanted, and still doesn't even work, and you can't do anything about it.

Economically and logistically, yes, absolutely. Would it have had positive ROI (stuff like that is expensive)? Who knows, there's a lot of factors to consider, like accessory costs, increased sales, brand goodwill, new markets, whatever. But they could do it if they wanted to.

Culturally, though, this is just not how Microsoft operates. Not for boring accessories like that, anyway. (The Dock team was not where the ambitious-for-promotion people wanted to be.) It absolutely is how Apple operates, though, which is why I mentioned them. Their relative positions in the hardware ecosystem are not an accident.

(However, the technical constraints at that time were severe, and it might not have been technically possible. As I said in a sibling comment, this was the era of Intel's Alpine Ridge (I think that's the right one), and it was a nightmare. I don't know that it was really feasible to do much better, on the timelines needed. Now, after that disaster of a chip was out in the market, everyone else knew what the correct solution looked like, and could build it. But Alpine Ridge had to... salt the earth?... first so we could all move on. Build one to throw it away, but ship it first and poison the market.)

Installing the specific driver like it's 1999 works well, but most people don't bother these days. And thus the world is a bit more crap.

Yep, that's the Surface Book (original). I've got one, and mine still works! Somehow.

That detach mechanism is insane. As far as I've been able to put together (I haven't done a patent search or anything, just heard bits from people who'd know), there's no motor involved... it's way weirder. I believe the thing actually heats up a nitinol shape-memory alloy latch in the base so it detaches from the tablet piece. That heating is why it takes a couple seconds. But then reattachment is instant, so it's just something clicking in to place. And you can't reattach immediately, because the base has to cool down (just a few seconds, short enough that you never notice unless you're deliberately messing with it). Black magic!

I'm not 100% sure of any of the above, except the use of nitinol somewhere. That's right, the weirdest piece of the conjecture is the only one I've got hard confirmation of. Like I said, black magic.

Docks were bad, bad products in those days. They were no longer the dedicated bulky-but-reliable things of years past, or the modern finally-debugged dongles we've got now.

This was Intel's Alpine Ridge and it was hell. (At least, I think that was the one. Certainly, it was hell!)

Docks are horrifying products. Thunderbolt docks are doubly horrifying. They ordered in every single competing dock they could find, from that era's products, and found that every last one was garbage in some way or other, usually fatally so. The Thunderbolt interface in particular, and the firmware that needed to run on that interface controller, was the source of a lot of issues. None of them were particularly intrinisc to the protocol, but the hardware available was junk and the software available was worse. They couldn't really order up a custom non-garbage IC just for a $100 accessory that sells in limited volume. (Apple, however, could and would; they'd also demand to control the whole stack. This shows.)

They were very proud they got the thing working as well as they did, even though they all knew it was still pretty much trash. It was still better than the competition. Which is sad, but what can you do?

(At least it wasn't the Wi-Fi chip. The Surface Book's Wi-Fi adapter was chosen by higher-ups as the same one used in the XBox, presumably for sourcing reasons. It is trash. Again, much blood, sweat, and tears were spilled making it work as well as it does.)

(I also have the exact circuit for the LED that lights up on the charger cable. Apparently it was a big deal, which I find hilarious.)

Just because alternatives exist for some people some of the time does not mean Office is worthless, or that buying it isn't rational.

(Though buying it starts to look a lot less rational when things like this happen.)

This is somewhat annoying. Please don't offer only one storefront as a place to buy your work.

So if it is unacceptable, it is unacceptable, and needs to be fixed. I said "trash" because it's going to become trash, and with luck just the power supply. Hope there's a fuse inline! Input rectifier failures tend to take down other stuff without one.

If you are truly close to the design failure point of the rectifier, it's not safe to run at all. (You are almost certainly not.)

If you are worried about the fact that you're only using one element of a multi-element package, again, it's a nonissue. We do this all the time. It's often cheaper to add a second bridge rectifier to get a single diode than it is to add another BOM line item for the "proper" part. As long as that diode isn't operating near absolute maximum ratings (it probably isn't), it doesn't matter that there are or aren't three more in the box.

(At least not on timescales relevant to individual humans.)

So hearing that makes me get suspicious that something else is going on.