"Applications" was in common enough usage as early as 1990 to where it was one of the default program groups in Windows 3.0. I remember seeing them occasionally referred to as apps as well, although mostly as an abbreviation on BBSes and things for the category rather than something we used verbally.

Why are you so fixated on 2020? Democrats have been screeching about stolen elections since 2000. But, you think those really were stolen, right?

Any service a company has to run is something that takes resources in management, upkeep, security patching, monitoring, etc etc. Even if the base cost for the hardware isn't much, there still has to be resources devoted to keeping it running. Whether that's still a significant amount for a company like Ubisoft is still up for debate, but it's not just something they can throw on a small AWS instance and forget about it. If they actually wanted to functional, and it's something like Forza Horizon where it's primarily a single player game with a lot of online functionality, being able to modify it to work offline even with some reduced functionality would likely be a lot cheaper in the long run than continuing to run barely-populated servers.

I'm also curious whether this was a situation of a game that had zero or nearly zero player base for years and people are more up in arms over the idea of a live services game that was paid for being made unavailable, or if it still had a decent player base being cut off and urged to pay for newer versions/other games. I'm also confused why this one seems to be the one that has triggered the backlash, I'm fairly sure there have been plenty of other online-only online service games that have shut down and been rendered unplayable even if the companies involved weren't quite as explicit about removing access to the client. I guess part of the backlash is there was talk about community server emulators and now Ubisoft has made the game completely unlaunchable or something? I'm not super familiar with the game or Ubisoft's launcher.

He mentions why it looks old on the home page, point number 12.

For a long time (and technically still for the time being, still got some standalone ones hooked up) Rokus were my favorite streaming devices. Simple and straightforward, clean UI, lots of decent apps, independent of big companies pushing their own streaming services (Firestick, Chromecast, etc) and even supporting older devices for a surprisingly long life. They seemed pretty developer-friendly as well including the ability to quasi-sideload unofficial/private channels too. But they've been gradually turning evil the last few years between locking the things down more, shitty T&C popups, questionable privacy practices, etc. I doubt I'll ever buy another Roku anything but I'm not even sure what the current ecosystem for non-evil streaming devices looks like anymore.

I'd really love for any of these to pan out and I genuinely hope they and similar things can make a difference. I feel like I've been seeing people talking about novel CO2 removal and/or sequestration methods for decades now and it never quite seems to pan out. Especially on the scale necessary to actually effect change. I sincerely hope they do work out.

A big part of the problem is just that there's basically no way to completely win. Every option will piss somebody off.

Release OSes too quickly that are too much alike? You get people complaining about there not being enough different to justify the upgrade. I remember people saying that about Windows 98 (especially compared to the last OSR version of Windows 95 with the USB add-on and IE4 for the Windows Desktop Update that could be hard to tell apart from Windows 98) at the time. Win2k to XP was criticized for just being Win2k with a Fisher Price theme. And even Vista and 7 a bit, though Vista was so poorly received they mostly escaped that fate even if 7 wasn't all that different from Vista in the end.

Release them too far apart or with too many changes, a different subset of people are upset they're too different and require too much relearning. 98 to XP perhaps, XP to Vista, 7 to 8 with its Start Screen and forced full screen metro apps might count.

Likewise with compatibility, where's the sweet spot? Too much backwards compatibility and they get criticized for "bloat" and "old code". Does the current 32-bit version of Windows 10 really need to still support running DOS and Windows 3.1 apps? Or even Windows 1.x and 2.x apps with some very minor resource tweaks? Probably not, but it can.

There's been a few attempts at an appliance computer with a somewhat minimalist OS that is never updated, and it gets left behind by technology very quickly. But then there's been other attempts at appliance computers with minimalist OSes that get updated at the whims of their creators and other people don't like that either.

But cut off too much compatibility and other people will be critical of that. 64-bit driver migration primarily with Vista was a big one, dropping NTVDM in 64-bit too even though I'd hazard to say a lot of the people complaining about it didn't use it anyway or had alternative methods to run their stuff (including the still-supported 32-bit versions of Windows until 11). Now we're seeing it with TPM 2.0 and other requirements for Windows 11.

Not to mention that at least until pretty recently, the speed of technological change meant that you couldn't always plan ahead for what would be efficient code and available APIs/services. Windows 95 had a number of compromises built in so it would run acceptably on low-end machines and maintain compatibility with DOS and a lot of other stuff. Do we really expect app developers to keep their apps running on Windows 95 and the compromises it made to run on 386s and 486s with single-digit megabytes of RAM when we have multicore, multigig RAM, multiTB disk space machines now?

Personally, I'm the kind of person that wants his OS to stay out of the way. I use a computer to use applications on it, not muck about in the OS and what it provides, so I'd prefer to be left alone. I feel like Win2k was probably the pinnacle of Windows UI design (although XP is fine too, and can be more colorful) and I really dislike the low contrast, flat color, hard to distinguish active buttons and controls from the background stuff that modern OSes in general are doing. But I also don't expect Microsoft or anybody else to maintain a single OS for decades based on a single purchase, or keep their applications running on decades old OSes that haven't been touched. Even for the ones that are "just" compiler flags every OS supported increases testing load significantly, as well as potentially code size. It sucks but sometimes progress leaves behind your favorite OSes and things. All that to say I'm agreeing with you very verbosely.

If I'm understanding the distinction being made now, it's that random individual home users will be able to take advantage of it. Corporations and other organizations have been able to do paid support contracts since at least XP, if not earlier, but Joe Smith with his 6 year old personal laptop he doesn't want to upgrade or mess with for whatever reason was always SOL. Now there's that option, whether it makes sense economically to do or not. By the time you've paid the $427 for three years of extended support you'd be able to cover a all or a lot of a cost of a new machine, and it also seems unlikely that someone dedicated enough to pay to keep an updated old version kicking around would change their mind three years later and upgrade then unless the hope is to skip a version (unlikely given current release schedules) or wait longer into the new version's lifecycle.

So I don't fully understand the use case there. I'm not quite conspiracy minded enough to believe it's testing or preparing the waters for a full Windows subscription future - I think that'd be a pretty big misstep even for a giant like Microsoft - but other than checking off a box and saying they technically offer it to maybe either drive upgrade adoption through annoying popups or a revenue stream from people who don't know better I'm puzzled.

As long as it was safe (which being batteries there may be chemicals involved, but perhaps they could be removed?) I'd certainly be at least asking about the possibility of getting it back if it hit my house. Not to sell, I'd stick it in a shadowbox and have it prominently on my mantle. Hell of a story to tell, not many people can claim their house was hit by space junk.

Yep. I got the emails to set up an account and buy shares but passed. To be fair, I predicted a great than 25% drop by now, so they've managed to do better than my low expectations.

You've obviously never been to Reddit. And you're obviously unaware of what was going on at Twitter pre-Musk. Lefties and righties live in echo chambers that they carefully curate lest anybody realize that the world is more complex.

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

I haven't tried it on the aforementioned 386 but you're almost certainly right about complex/big stuff. Although it would be interesting to try someday - one thing I did do a few years ago was taking my 486DX2/66 with Lots of Memory and Win98 Lite and Office 97 (Or maybe XP, I forget, I was able to get the Office 2007 compatibility pack installed) and see just how much I could theoretically daily drive it for a whole workday. Most of my job is sshing into servers and Doing Stuff and there were versions of putty with sufficient cyphers supported. I also converted a handful of mildly complex but security-safe docs from our Google Docs to the compatible format and was able to open and mess with them pretty much fine. I didn't actually do much because I'm pretty sure my company IT wouldn't appreciate me connecting up a 30 year old OS to the corporate VPN but I could have almost done it. Obviously a 66mhz 486DX with I think 192MB of RAM would have been orders of magnitude faster than the Toshiba for lots of reasons.

NT4 and the whole Cairo project ended up somewhat delayed and kind of a mess. I've always felt like NT 3.51 was a bit of a stopgap to get a handful of the Windows 95 common controls and DLLs and such out there in the wild while they were finishing up NT4. There was only just barely browsers released for NT 3.51 at all - Microsoft never released 32-bit versions of IE for it which means it was stuck with the same 16-bit IE straight from Windows 3.1, and a handful of other 32-bit browsers like Netscape did work on it but ended up behind the times. Technically you can get Firefox 2.x running on it but nobody was really trying to use NT 3.51 by then so it's mostly just a party trick.

NT 3.51 with the newshell preview is kinda neat and fun too.

I've done a lot of retro computing in the last few years and one of the surprising things I've found is that NT 3.51 runs surprisingly well on some pretty low end stuff. Most notably a 386SX/16 mhz with 11MB of RAM (Yes, 11, it has weird RAM setup) Toshiba T3200SX. It even runs Office 97 mostly decently. As it happens I actually ran OS.2 2.1 on the very same machine back in the late 90s during high school and it was... OK but the lack of software available, especially networking/internet stuff, really limited what I could do with it. I honestly can't remember now if I never got internet working in OS/2 or if I ended up doing something like Trumpet Winsock in WinOS/2 with Win3.x internet apps, but I ended up going back to WIn3.1 on it eventually. It was a fun little quirky experiment though. That said the thing was considered a high-end machine at the time of its release in 1989 (or as high-end as a 386SX could be, but it was a basic evolution of a high-end 286 machine) and cost a huge amount of money at the time, so I'm not sure it's a fair comparison. But even a few years later with 386s much cheaper and readily available.

I can't say how it'd fare for daily driver use BITD if it's what someone wanted to use or needed to have NT but I could see an alternate history where 32-bit software took over faster, DOS compatibility importance waned faster, and an effort was made to slim down NT (or OS/2) for lower class 386s and 486s. You're right though, the RAM was the main issue, though prices collapsed throughout the 80s and by 1996 or 1997 even 16MB was a solid mid-range affordable machine. A 12MB 386 like the T3200SX maxes out at would have been a monster machine in 1989 and still a lot of RAM through the early to mid 90s. The first PC I bought with my own money was in 1997 for decently under $1000 with an AMD K5-133 and 16MB of RAM. I'd always wondered what might have been if Microsoft had (or IBM?) put a little more development effort into a consumer version of NT knowing that RAM prices would continue to drop, maybe doing something to ensure good driver support a la WDM, and being a little more pushy with getting people to migrate to it. Mostly though it was DOS/Win31 being Good Enough for most uses until the internet era (and some stuff like gaming) started pushing software complexity upward.