Buying an XBox would still be redundant hardware, for the most part (excluding exclusive games). Also, at least for me, my office computer and my gaming PC aren't really redundant. For my office/productivity use, I'm content with an older computer with an integrated graphics chipset. It's not a gaming PC, but it's fast enough to open web pages.

I'm not sure I understand why this would make steam machines useless. The main value of a steam machine, as I see it, is that it allows you to have the convenience of a console in what is essentially a generic gaming PC. That is, it has a controller and a GUI aimed at connecting it to a TV and not using a keyboard or mouse, but it's not a locked-down console. It's just generic hardware that will play all of your PC games, and those games don't become obsolete and unplayable when you upgrade to the next generation.

The article says:

So what they're saying is, if you have a PC running Steam (which is really all a "Steam Machine" is) and an XBox One, then it's kind of the same as having a Steam Machine. Yes, it is... because you're starting with the scenario where you have a Steam machine. It's like saying, "There's no reason to buy a car, since if I already own a car and I buy a bicycle, it's like owning a car!"

Look, you shouldn't assume that I want an XBox. I can get a PC with better graphics and avoid being locked into Microsoft's ecosystem. I can install game mods, my games don't all go unnecessarily obsolete with every new generation of PC, Steam often has very good sales, Steam doesn't make me pay a monthly subscription for online services, and I can use that PC for other things if I like. To me, the only thing that would want me to buy a console at this point is if there were an exclusive game that I really wanted to play, and I've found that I can live without it. I don't want an XBox, so it doesn't make sense to me to say, "If you buy a Steam machine and an XBox, then it's like having a Steam machine!" I'll just buy a Steam machine, thank you, even if it's not a branded "Steam Machine".

From what I read back when the last time this was a story, a few people were saying, "It seems to be generating thrust, but on the other hand, the amount of thrust we're measuring is basically within the margin of error, so... we need to keep testing this."

So in your head, some people who got vaccinated should be included in the classification "those who cannot get vaccinated. Well that says a lot about the strength of your argument.

I didn't claim that measles would kill 3 million people. I was using simple math to point out that "a small percentage of the population" might still include a whole lot of people.

Yes, but it's distinguished from the steam engine, from what I understand, in that the power output is comparable enough to the margin of error that they're still in the process of verifying that it's actually doing something.

To be fair, it's not clear what you consider a "compromise", or even what features are desirable.

For example, I want a very thin, lightweight, but sturdy phone, and any additional hardware you pack in there runs the risk of adding weight, and any port or removable piece is a potential weakness in the structural integrity. If you give me a SD card port, I won't use it. If you let me remove the battery, I'll pretty much never do that unless the battery actually fails within 2 years. Front speakers? For what?

Now I'm not trying to argue here that these aren't good features or that you shouldn't want them. I'm just pointing out that when they say, "no compromises", it's inherently a claim without a specific meaning, and one man's "compromise" is another man's "that's exactly what I want".

Not true. It's true that there are some people who cannot be vaccinated, and those people are helped by other people being vaccinated. Aside from those people, vaccines are not 100% effective. Some portion of the people who are vaccinated may still get sick if exposed to the virus, and those people are also protected by other people getting vaccinated.

In both of these cases, you can say, "it's a small percentage of the population." Small percentages of the population, however, can still represent a lot of people. If something kills 1% of the US population, that's still about 3 million people.

It's worth noting that there's a big difference between those theories and this engine. With this engine, they're putting forward a piece of technology and saying, "We don't know how this works, but we're claiming it does." In the case of Newton and Einstein, they put forward a mathematical model that was internally consistent, and the question was whether it applied to reality.

So in one case, they're putting forward a technology without a real explanation as to how it works, and in the other case, they're putting forward a coherent theory that seems to explain phenomena that we have witnessed. Also, both Newton and Einstein's theories had the benefit of providing a clean explanation to phenomena that we were having a lot of trouble explaining.

I think some of the skepticism is not as to whether this might be an engine that produces some small amount of thrust. I mean, a little skepticism is a healthy response for any new scientific discovery, and it's not inappropriate to ask for proof. Since the thrust we're talking about is so small, the margin of error is large, and proving that it really works takes a bit of doing. I don't assume that it works, but I also don't really disbelieve it if NASA scientists say it does.

However, when this was reported, it was reported in many places as "OMG! NASA has created a warp drive. We can go faster than light now!" I'm skeptical about those claims.

I think that this is really part of a larger problem that eventually ties back to identity management and account management. That may sound like a strange thing to leap to, but hear me out.

One of the problems I've noticed for years is that it's not easy to keep track of all my accounts. Every time I sign up for a new account or trial, I have to create a new account, create a username, create a password, associate it with an email account, choose security questions, bla bla bla. Dual-factor authentication is supposed to help with some of the security problems associated with all this nonsense, but it also adds another complication to the whole thing. Once all that's done, I need to keep track of all that information that I used to sign up.

It's not so bad for individual accounts, but after a few decades of trying things out, abandoning accounts, signing up for trials that I end up not using, and all kinds of things, I really don't know what accounts I have available on which services, what the usernames are, or which email address they're associated with. When I answered security questions, I don't necessarily know what I answered with-- it asked for my favorite author, but was that my favorite author from 2 years ago or 10 years ago? Did I tell the truth when I answered it, or did I answer with a sarcastic joke answer? I honestly don't know for some accounts. I don't even know, for example, if I still have a MySpace account from roughly a decade ago, that I created, signed into a couple of times, and forgot about.

You're thinking this is completely off-topic, but here's the thing: as you have an "Internet of things", there's a good chance that each of those items are going to have their own account on their own service. You have some program to control your lights at home? That program will need an account. Someone invents a smart-vacuum, and it's internet connected? That'll have it's own account. These days, companies don't want to collaborate and develop standard APIs, common platforms, open protocols, or whatever else. Every company developing an app or a website wants to do it's own thing it's own way, while locking out the competition from interoperability. So now, every new Internet-connected thing is going to add complexity to your online life.

Asking to provide privacy controls to consumers is putting the cart before the horse. Even if you want to provide those controls, you're going to have different controls in different places in different UIs, all across different services with different accounts. Users won't be able to effectively manage those controls even if you provide them. What needs to happen first is that we need to develop some kind of identity management and SSO that begins to shrink the task of managing these various accounts. Once you have something like that, you could create APIs for managing those accounts, opening and shutting down accounts, viewing which private information is available in each account, and restricting/removing the private information as needed.

That's easy. Fitts' Law. Putting menus on the edge of the screen makes it easier and faster to select the menu.

I agree with your criticism about gestures. I had the same general problem with Windows 8 when it first came out: a lot of the features in the metro UI were hidden under some kind of obscured interaction. Hover over this area, and you get one menu. Hover in a different location, and something else happens. Right click when you're in this application, and it pops up with a menu from the bottom. Right click somewhere else, and it does something different. Drag down, swipe left, do a little dance, and some kind of other magic happens. What are all the features? Who knows what you'll find next!

It also reminds me of Apple's reluctance to have two-button mice. A lot of people made fun of it as pure stupidity, or as though it was a technological failure. It was a design choice. Apple designers didn't like context menus, since context menus mean that right-clicking in different places and in different contexts produced different menus, and the user had no real way of knowing what would be in a context menu ahead of time. The only way to learn context menus is to right-click in various places and try to discern what the pattern is, and hope that the developer was consistent. It's rumored that a big part of the reason Apple has stuck with one-button mice is that, if you're not relying on context menus, multiple buttons are largely unnecessary for normal productivity uses, and not having multiple buttons deters developers from putting important functions in context menus.

I don't think that's a baseless assumption. I've been working in IT for a couple of decades, and I've seen the inner workings of quite a few companies, and let me tell you: For most people in most fields, they're incompetent in their field.