I don't think that's enough. In addition to re-enabling it, I think we also need to re-pin it to taskbar, and pin two different versions of it (regular and 365) to the Start menu, make extra sure it's set to run at login, and add some periodic notifications. Hmm, anything else? Can we bind a key combination that gets accidentally hit a lot, to open gaudily-animated Copilot window that takes several seconds to render in? That'd be great.

Who the heck uses *any* version of Windows, within the first six months after its release? That's completely insane. Honestly, I think Seven is the only version of Windows that I ever *have* used in its first year of existence. At the time, my analysis predicted that Seven would have very few disruptive changes or major buggy new features compared to Vista, giving it roughly the same deployment desirability as a service pack. I started deploying new computers with Seven in 2010, just a few months after its release; but I think it was more than three months.

Seven was by at least an order of magnitude the *least* buggy and *most* stable initial release of an operating system, that Microsoft has ever managed[1]. XP needed two service packs just to achieve basic levels of non-horribleness. Eight and Ten never got there at all, and Eleven, let's just say it's not there yet. Win2K was not quite as bad in this regard, but even it was pretty shaky before the first SP. NT3 and 4 were strictly worse than Win2k in this regard, and it's not even fair to compare the 95/98/Me product line, that's just a different animal altogether. (No memory protection of any kind, means that if you run any third-party software at all, stability is completely at its mercy. Same thing with DOS. So you can't even really evaluate the stability of one of these system, outside the context of the complete list of all the software it's going to be running.)

Footnote 1: PC-DOS 3.3, which may arguably have been less buggy, was released by IBM. Also, as discussed, it's not a fair comparison.

That isn't really the big problem. I mean, yes, that's an issue too, but it's secondary.

The main problem is, people don't want to use these cars on their own separate system of roads; they want to use them on the _same_ roads as existing human-driven cars. Consequently, it's not enough for them to be better drivers than a human. They also have to be able to _interact_ reasonably with human drivers and not unnerve them or cause them to behave more idiotically than usual. Humans have a marked tendency to respond to unexpected situations by suddenly deciding to panic and flail about, and it's important that any new stimuli introduced into the system, don't trigger that sort of behavior. Additionally, human drivers rely fairly heavily on a number of poorly-documented informal non-verbal communications mechanisms to communicate with one another, to the point where people get *more* upset if these conventions are violated, than some of the actual rules (like using a blinker light to signal a lane change, for example, which human drivers fail to do pretty often). Some of these informal communications mechanisms are things a driverless car could theoretically be programmed to do, but others, like making eye contact, are really not. The design and implementation of the driverless cars need to take that stuff into account, and so far they mostly haven't.

When I think of long-haul trucking, I think of drivers being away from home for multiple days at a time. Dallas-to-Houston is a single-day round trip; you could be home every single night. That's a much less undesirable job for a human driver, and not really what "long-haul" means at all, IME.

Wow. I'm amazed anyone who's spent any time looking into the lives of people in Europe and Asia could believe this. Oh wait, you've never done that, have you?

Please keep sharing anything you observe as it happens.

The last thing any of us should want is for OpenAI to take over from Google.

I really don't understand this decision that Google should be broken up as though its 'monopoly' in search isn't entirely based on skill and talent. But if we *are* going to force companies to break up into components, can we make sure new monoliths aren't just created as a result?

This, entirely this.

> Even CVS was better for the masses.

You're out of your chair. CVS was literally more of a hassle to use, than just making a fresh full copy of the entire codebase periodically ("cp -r version control").

git isn't perfect, but comparing it to CVS is disingenuous.

> probably lightly-used service

It's not lightly used. Most American cellphone users get notifications from at least one sender via their carrier's email-to-SMS gateway. Common examples include library hold-available and overdue notices and pharmacy prescription-ready notices, not sure what all else. Pretty much any company or organization (other than your phone company) that is sending you SMS notifications in an automated fashion, is probably doing it via email. It's orders of magnitude cheaper and easier to set up, than a direct phone-company-type connection to the phone network itself. Realistically, the only really viable alternative for most organizations is to build apps for iOS and Android, and the user gets the notification via the app, rather than via SMS.

> USB-C is rapidly replacing DisplayPort, so within maybe a decade, that
> standard will be pretty much defunct, and arguably, it almost is already

I mean, I don't know about where you live, but I've only *started* to see substantial numbers of computers with DisplayPort output, within the last year or so. (I saw my first one quite a while ago, but then it was a long time before I saw another... Now it's most of the new computers I'm seeing.) DVI is the video output port standard that is currently in the process of going away, near as I can tell. HDMI and VGA are now losing market share (to DisplayPort) but will probably be around to some extent for a while, especially VGA (because, obsolete as it is, it has also been _extremely_ widely supported for decades, and that kind of ubiquity creates a certain amount of inertia; I just finished setting up a new, Windows 11 system, with 32GB of DDR4, and it has PS/2 keyboard and mouse ports, which have been technically obsolete since USB was introduced; but the tail on obviation of +standards like that can be lengthy).

Separately, HDMI is also losing market share to integration of formerly external functionality, into television sets themselves. A lot of folks don't feel that they need an external device to send video to the television any more, so now the only connector the TV has is for power, and everything else is via 802.11. But this is really neither here nor there for the computer monitor market. In any case, HDMI can't completely go away until video game consoles switch to something else (or get integrated into televisions, I suppose, but I don't currently see any reason to predict that particular development).

That's no problem. The company that makes those coin-flattening machines (which are present at basically every major tourist site in the world at this point) can just add to its line-up a little vending machine that sells penny-sized blanks, for a quarter or so. The stamper is going to put its own imprint on the thing anyway, so you don't need the surface patterns that a real penny would have, like the portrait of Lincoln and so on. All you need is an appropriately-sized cylinder of sufficiently malleable metal. If they're clever, they'll sell three different ones: basic zinc for a quarter, something with just enough bronze coating to "look like a penny", for fifty cents, and something nice and brassy looking for a buck fifty. And then they still get to charge you however many quarters it is these days, to run it through the machine. And you get your souvenir, which is *still* cheaper than anything in the gift shop. Everyone's happy.

Yeah, but any time it's proposed, the (relatively small number of) people who actually *want* the time change, deliberately agitate the argument about which way to go (always daylight savings, or never).

The obvious answer is, we clearly don't need to make that decision at the national level. Just let each state decide which time zone it wants to be in. States that are near the edge, or can't achieve a consensus, can even split themselves up into pieces, like Indiana has done in the past, so that areas with a specific reason to go one way or the other, can do that. We don't *need* to make a national ruling on whether always-DST or never-DST is better. We don't all have to be in the same zone.

But the pro-time-change lobbyists are very good at getting people arguing about which is better. It's easier than trying to convince people to actually like the time change, and just as effective at derailing any effort to get rid of the time change.

And 100 days in a year?

That won't cause any problems at all.

Eh, I'd just give them an incomplete and tell them I'll see them again next semester.

"If you want to pass this class, learning to write is not optional. It's a required element of the class. You can't just skip it."