One of the things that a quantum computer needs to be able to do in order to function is to hold information at rest, no different from your classical computer. There are two things that make this tricky. First, the information is analog, rather than digital. Second, the environment likes to sneakily "measure" your data so that it decoheres and no longer behaves the way it should. Both kinds of problems are in practice dealt with by encoding the quantum information so that errors can be corrected.

If the word "decoheres" sounds really fancy, think about it this way: coherence versus decoherence is the difference between a rainbow and a grey cloud. In the former case the waves are able to interfere with each other in interesting ways, whereas in the latter case they scatter and do not interfere, producing boring results.

If you are able to read this, then you exist as a conscious being. Everything else is just a model, which you experience as thought projected into your consciousness, just as you experience other senses.

1. First, working in terms of decoherence is significantly simpler than worrying about whether something has been measured or not at every single step of the evolution of a system, because I have observed that when people do the latter they tend to get headaches contemplating the meaning of the "quantum eraser" when there is no need to. Second, you actually can observe Born's rule in action by modeling the evolution of a system with an experimenter performing measurements and watching it emerge from the calculation.

2. The only way that the two sides of the EPR pair know that they agree or disagree is by communicating with each other and comparing results, which can only happen through local interactions.

3. I have no idea what you even mean by this. What makes the (terribly named) Many Worlds Interpretation nice is precisely that you can just treat everything as a wave function, with parts that might be entangled in ways you don't know about (i.e., decoherence, modeled via density matrices).

4. The fact that you are even making this claim is why I have trouble taking the rest of your comment seriously at all, because I specifically said, "However, it is important to understand that the concept of branches is just a visualization; it is nothing inherent to the theory, and when things get even slightly more complicated than the situation I have described, they do not meaningfully exist at all."

I agree completely that that the Copenhagen interpretation makes an excellent phenomenological model in simple (albeit, very common!) settings. However, the problem is that it breaks down when you consider experiments such as the "quantum eraser" (mentioned in other comments here), which causes people to tie themselves into intellectual knots because they are thinking too hard about exactly what is going on with measurement; once one deprivileges measurement so that it becomes just another kind of interaction, the seeming paradoxes disappear.

See, this is why I prefer the (terribly named) "Many Worlds" interpretation. Unlike the Copenhagen interpretation, it does not privilege measurement over other types of interactions between systems. That is, the wave function never collapses, it only seems to because you, as the observer, are part of the system.

The easy way to see this is to imagine that you put some other experimenter inside of a box. When they perform a measurement, from your perspective the wave function has not yet collapsed, but from the experimenter's perspective the wave has collapsed. Essentially, it is as if the system in a box has branched so that there are multiple copies of the experimenter within, one who sees each possible measurement result, but because you are outside of it you could, in theory, reverse the measurement and unite the two branches. However, it is important to understand that the concept of branches is just a visualization; it is nothing inherent to the theory, and when things get even slightly more complicated than the situation I have described, they do not meaningfully exist at all.

(Also, if it seems implausible that a macroscopic system in a box could remain in a superposition of multiple states, you actually are not wrong! However, the reason is not theoretical but practical: any system inside the box will interact thermally with the box itself, so unless it is perfectly insulated you cannot help but interact with it and therefore measure it yourself. This keeps going until essentially the entire world cannot help but perform a measurement of your system. Preventing this tendency from screwing things up is one of the things that makes building quantum computers hard.)

Hey, now, just because I am an overly paranoid person does not mean that you have to be as well!

Thus dooming its fate.

Thanks, your comment is an antidote to my paranoia that it is impossible to do anything to address all threats. 😀

Given that your advice is very sound, I have a question: would I gain much by using OpenBSD? The conventional wisdom when I last checked is that it is the most secure unix-like operating system on the planet.

Right, but there is an entire spectrum of plumbing maintenance. I am perfectly capable of plunging toilets, but when a drain fails to work after several attack on my part then it is time to call in the plumber.

I mostly just like building and tinkering with things, and I really like the idea of setting up services that I control that host my own data that I can access from anywhere. I have no real interest in learning about more than the minimum amount needed to do that simply because that is not how I would like to spend my time.

(Lest you continue to have the wrong impression that I am afraid of learning new things: There was a period in my life where I was constantly learning new technologies, programming languages, etc. Eventually I realized that I had demonstrated that I was capable of learning anything that I wanted, and there were so many things out there to learn that I needed to start becoming more selective. At the moment my learning goals tend to be more math focused; currently I am trying to learn graduate-level category theory and measure theory.)

If I really need to master all of the steps that you've described before deploying my host on the Internet, then my conclusion is that it is more trouble than it is worth, because my concern is that if I screw up then I will make the Internet a worse place by contributing to botnets.

That does not sound so bad; the parent comment made it sound a lot worse than that.

I admit nothing.