SpaceGhostC2C

Not really. It's more like 3 posts of drama and 3 pages people fulfilling their urge to tell everyone how they don't see any problem.

At the end of the day, you are only looking at one screen at a time, so It all boils down to how costly is it for you to tilt your neck vs. alt-tabbing. Multiple monitors are likely to be worse ergonomically, and probably slower to execute, but they provide a less abstract form of focus switching. Alt-tabbing is physically more efficient, but also more disconnected from your instincts. It's similar to the trade-off between touchscreens (or the silly Minority Report interface) vs mouse&keyboard/keypad.

This may be the difference (from the manual): Your CPU adapter must be triggering the special latch, hence it works in PWM mode. All other headers are standard, and auto-detection must be based on the number of attached wires (4=PWM, less than 4 = DC). Sucks to not have a manual option, only auto.

OP has already answered that: having to apply a sticker by yourself vs. not having to do anything to obtain the final result. Whether the seller was misleading people or not into believing it was one or the other, it's something I haven't looked in detail to form an opinion. Still, reiterating questions with obvious, already provided answers is a waste of time in any case.

Also, check how it behaves with and without "hardware acceleration" in the browser settings.

Yes. you also want to delete any non-data partition (e.g., boot record, recovery partitions) and just merge it with the data one, if possible. Just make sure you go into the BIOS right after re-plugging your old drive and make sure there's no change in the boot order, and the Sata SSD is the last option, to prevent the old bootloader from being summoned. If you have any place to temporary store your relevant data (including a temp folder in your new NVMe, if big enough), you can just manually copy anything worth saving, then nuke the drive, then move the data to the now empty drive. I've done both, and generally recommend the second approach, but which one is more tedious depends on how much data worth saving you have and how it's organized, vs. how much has Windows and software installations abused your drive. It certainly can be done if no other choice is viable.

If you touch it repeatedly (say, 3 times in a row), does it shock you every time or just the first time?

The Cooler Master 590 had the same fan mount on the right panel. I believe there were a couple more. Plus, some cases whit left-panel side fans that you could interchange with the right side panel. However, none that I remember had a built-in motherboard tray fan. The Zalman Z9 had a ventilated SSD mount on the back, not sure if a fan could be mounted there - but probably wouldn't align with the socket anyway. Yes. I mean, not if your socket and VRM temps are under control. But if you do have a VRM temp issue, then yes, blowing air over the VRM area, whether from the front, from the back, or both, will make a difference. Case airflow won't be enough, because there can still be a pocket of hot air stagnating over the VRM area, below the rear exhaust, kind of caged between GPU, CPU cooler/RAM, and topmost motherboard components (including VRM heatsinks themselves). Again, overbuilt motherboards with low power CPUs will show no difference at all. But, for example, practically every AM3+ board with an 8-core chip (and unless you got an FX-9xxx, you should overclock it) would benefit drastically from specific, local VRM airflow. I'm talking 20ºC-like improvements (albeit with front-side cooling in my tests), I'm talking lower CPU temps (and lower socket vs die delta) because, without the fan, the VRMs where getting so hot they were heating the socket. I'm talking about preventing this when overclocking: (Notice the CPU was not supported by the board, and even for FX-8xx chips the manual would tell you to use top-blow CPU coolers instead of towers, precisely to provide VRM airflow. Same problem with AIOs -no Liquid Freezer II back then). Bottom line: it doesn't make a difference as long as there isn't a difference to be made.

Still much better than the landfill, polluting our food, and intoxicating the inevitable underage scavengers that roam the third-world dumps where the "recycled" (LOL) electronics end up...

False. A 10 years old PC is perfectly usable today. Even as far back as the 8086 you will find times in which essentially the same processor would remain ubiquitous for a decade or more. And that's just computers, which until very recently were in a very infant stage. If we take "technology" in all of its breadth, the case for your ignorant statement becomes far worse.

It's not an Ubuntu-exclusive feature; in fact, I probably installed Linux that way most of the times. Granted, they were mostly Debian derivatives, with a few RH-like exceptions. I prefer to avoid it these days, but, for example, my laptop is dual booting Win10 and Linux Mint from the same drive: I shrank the OS partition, then installed Mint to the unused space and let it set up GRUB to give me the choice every time. I believe Windows 10 and later are smarter about installing themselves alongside Linux too, but I never tried: I stick to the old advice, namely, make a partition, install Windows to it, then install Linux. The installer will tell you whether in needs to nuke Windows anyway, just make sure to choose carefully the install type option and don't confirm any deletion you are not sure of.

Yes, every game deserves preservation, when "preservation" boils down to a no-scam rule: you don't sell games that will stop to be playable with no reasonable exist strategy (and the level of "reasonable" proposed by Ross is really light on developers/publishers). They can still disappear forever because none of the people who bought it chooses to keep a copy, because no one chooses to keep working hardware that is compatible. Dead by oblivion is still very possible under that scenario. But prohibiting "death by killswitch", even for people who have the game, the hardware and the will, is all the level of preservation being demanded here. I think that's closer to the case of strictly multiplayer game, where's there's no game without other people, or even Pokemon Go, where the physical world may change so you may not be able to play in the same area you used to. In other words, when interfacing with something that's outside the game is an essential part of the game, especially when it's something out of the control of the game developer/publisher, preservation beyond the software itself cannot be expected. However, it would still be the case that if the Google Maps data was available then the game should run without artificial phone-home requirements failing, and to the extent that you sold that to anyone. The Google Maps data, in my view, is no different from a console or other legacy hardware: the software itself being preserve-able doesn't imply an obligation to presereve working hardware that can run it to anyone. Were the servers for sale, or are you just trying to force a fallacious argument? I'm pretty sure Ubisoft has no interest in that trade. Also, there's an obvious reason why setting a precedent so the industry stop scamming people by pretending to sell a product is in no way equivalent to keeping one specific game alive.

I'd dare to say that's the weirdest conclusion to draw from it all, but you do you ¯\_(ツ)_/¯ (as a side note, those two videos are 4 years apart..)

That video, plus the "Games as a service is fraud" video summarize my thoughts on the topic perfectly. So, rather than attempting a reply to OP, I'll just send him to those

It doesn't matter, as long as you have a couple programs actually doing something, some spikes to some degree in some cores could happen. The issue is how often, how high, how many... But that's harder to tell. If you are never able to spot it in real time, it makes it more likely that it's some background task (antivirus, windows updates, indexing, etc.) taking some of your activity as "idling", but deciding you are "active" when you open task manager or something else (technically, your computer is always running something, so there is some decision making involved in triggers based on "idling" or not. You can notice similarly inconsistent behavior with monitors sleeping, screensavers, etc.). Yes, even if it works, ideally you want to find out the culprit, so you can decide if you want to accept it or not. I don't know if there's any program (that won't be detected as task manager) that can log CPU usage by application in the background, so you can check ex post.

That's not idling, though. Don't close task manager. Keep it open, sorted by CPU usage, together with coretemp. Keep both visible. As soon as CPU usage increases, check whose causing the spike. If it never happens, use your computer with task manager open forever, but I doubt it'll be that simple

The screenshot shows the CPU is idle, though. Downclocking in such circumstances is expected behavior. Can you take another one while stressing the CPU a bit? Also, I would assume they fixed it later, but 1st gen Ryzen (i.e., 2xxx laptop chips) had a sleep bug, which caused some combinations of sleeping and plugging/unplugging (I believe it is sleeping while plugged, then unplugging and waking up, but not 100% sure - even though I got one myself ) to produced a stuck CPU at lowest clock. Just plugging it back would un-stuck it, though. Doesn't seem to be your case.

Visual Studio is essentially a glorified text editor. What kind of performance issues are you facing? If you're struggling with compiling whatever you develop there, then it will depend critically on what you are doing with it. I mean, a Python "Hello world" project done in VS will fly in potato-level hardware... Conversely, compiling an office suite from source will take a bit longer, with or without VS...

As far as I understand, you can expand your current pool with a second vdev. Not sure if there's any consequence from not having the same capacity across vdevs, but I seem to remember it's not that important. Having 2 vdevs in one pool is like having a "raid0" (stripe) over your two "raid5" (raidz1) vdevs. If you do that, each vdev will survive losing a drive, and the pool would survive losing 1 drive in each vdev, but losing two drives in the same vdev would kill the whole pool. I think raidz expansion has been announced but no yet available, or not universally. You can theoretically do a "migration-expansion" by creating a new pool in a degraded state right from the start, copying the data to it, then restore the pool integrity by adding the old drives as replacements for the "missing" drives - and you still would have the size mismatch problem in your case.

Or even uncomfortable. Like every time I go to a place heated up to 25ºC in winter, when we're all coming from the streets with warm clothes, and then find it cool down to 17ºC in the summer, when people arrive in shorts...

How much of the world's total power consumption is accounted by personal computers, though? People who make (a lot of) money from computers and pay the corresponding power bill care much more than consumers about efficiency and density, and they aren't exactly rushing away from Epycs and Quadros yet. There are reasons for that, as @igormp points out. That may change, but make no mistake, ARM isn't some form of magic when it comes to transforming power into useful outputs. Just like driving a pick-up truck for your daily suburban commute is stupidly inefficient, yet you can't replace it efficiently with a fleet of beetle-like cars when it comes to actual heavy lifting.