Ask Paul: March 10 (Premium)

Good morning and Happy Friday from Roma Norte, Mexico City! Here’s another great set of reader questions to kick off the week a bit early.

Realized ambitions?

sabertooth920 asks:

Roughly 20 years since Longhorn, do you think the various iterations of Windows (Vista, 7, 10, 11) have realized the ambitious aspirations Microsoft initially envisioned?  Obviously, much more gradually.

In an interesting coincidence, I just finished updating the Longhorn section in Windows Everywhere (which has grown to a mammoth 830+ pages with lots still to go), and I was literally just thinking about this. Not just in the context of Longhorn, but also of Cairo, which I wrote about the other day. These two projects very closely parallel each other in that each was ambitious and future-leaning, and of course each failed. The difference is that Longhorn did so publicly, whereas we’ve never seen a public build of Cairo to this day. (You’d think something might have leaked given how long it was in internal testing.)

I miss this Microsoft because they were really shooting for the stars. But the problem with evaluating them vs. what we have today in Windows is that the stakes are so much lower: when Cairo and Longhorn were ongoing concerns, Microsoft dominated personal computing, and everything that did—everything they said—had such weight. A simple announcement from this company could set the industry down a different path.

The influence of Windows faded slowly but it has only kept fading. The start of it, in some ways, was Mac OS X, which was so more advanced than Windows 2000 in some ways, but was also more advanced than Windows XP, which not only shipped later but was then kept in-market longer than expected because of the delays in Longhorn. That was the beginning of Microsoft ceding its technical leadership in the PC space.

Did Microsoft catch up? And to your point, did they realize their ambitions? Yes, mostly. But it’s interesting how many of the initiatives that began with Cairo/Longhorn just don’t matter much anymore.

For example, Cairo was supposed to introduce an object-oriented file system based on SQL Server technology and virtual folders, and the former was eventually found to be too top-heavy to be viable and was abandoned, as was a lesser attempt at a similar file system in WinFS. Virtual folders did happen in Windows Vista, and stuck around through Windows 7 (and are still technically available today, though you have to dig around to see that), but were found to be too confusing to users. That’s a great example of an engineering mindset running into the real-world needs of normal people. (Microsoft is apparently resurrecting ReFS for Windows 11; that file system debuted in the Windows 8 timeframe, but I believe it was only really used in Server. This won’t achieve the goals of Cairo/WinFS, but then I don’t think we really need any of that anyway.)

The biggest visual innovation in Longhorn was the Aero Glass visual style, which appeared in Vista over 6 years after Apple had debuted similar technology in Mac OS X. Aside from the prettiness of it, the point there was to overcome a long-time issue with GUIs and floating windows, where it was easy to hide other application windows, something that used to confuse users. So with this glass-like effect, you could see through windows to some degree and get a sense of depth and that there were other windows back there. And while it lasted longer than the Luna UI from Windows XP (two product versions vs. just one), it was short-lived, and Windows has employed a flat, non-transparent base UI ever since. (Well, until WinUI material effects, I guess, so now we are starting to see some translucencies again, but not to the same degree.)

To me, the crowning moment of the Longhorn debacle was when Apple saw what Microsoft was trying to do with instant search and just made it happen in Tiger. I wrote about that in Programming Windows: Redmond, Start Your Photocopiers (Premium) and will soon be editing that bit for the book version again, but it was a brutal comeuppance. The good news? I guess Windows 11 Search pretty much meets this bar. Granted, it’s almost 20 years later.

I guess at a high level, we’re pretty much there. Longhorn focused on both core/foundational improvements and user interface/usability, and those gains are all present. It was also the last viable major developer platform in Windows, and I guess that’s what’s really missing today. (And will never happen.) Longhorn/Vista was Microsoft’s last attempt at revolutionizing desktop apps. Given my interests, I find that the saddest loss of all, but it’s understandable given how the world has changed and that most new app development is now on the web and mobile. Creating a native app on any platform, but especially on a desktop platform, is passé.

A new Era at Sonos

matsan asks:

Sonos Era 300 – what do you think? Personally, as a long time Sonos user, I think it’s the most ugly piece of tech I’ve seen in many years and I can’t think of any way to fit then in our home. Luckily we invested in an Arc+Sub+One SL combo so the Atmos needs are satisfied for us. I sure hope they at least sound good.

Based on my experience with the Sonos Five, I’m positive they will sound good. Sonos is (sort of) late to the spatial audio thing, so this is their first big push aside from a single soundbar (the Arc, which you have). Aesthetically, that’s suggestive. I appreciate that they are likely using a new design for spatial audio products, given that the Era 100 is just a traditional evolution of the Sonos One/Sonos:1 family and looks more familiar. I would need to see one in person, but I don’t mind the look. I feel like we get used to these things over time.

Naturally, I consider whether I should buy any of this stuff, but we’re in a particular place right now in which we just sold our house and will be moving into an apartment soon. The Sonos Five pair and Sonos Sub we already have will be overkill for this new place, so upgrading to two Era 500s is out of the question (aside from the cost of doing so). And I feel like the true value of spatial audio now is mostly with movies anyway, so your setup is ideal. (I’m curious about headphone-based spatial audio too, to some degree.)

I’ll see if we can make the Sonos speakers we already own work in the apartment. (We might need a Sub stand, at least, to lower the booming for the people downstairs.) And we would like to figure out something for our even smaller place in Mexico, where maybe a Sonos Ray and two smaller speakers (Sonos Ones, Era 100s, or maybe the IKEA Symfonik bookshelf speakers we currently have in the kitchen) will be more than adequate.

Speaking of which, Sonos needs some kind of mid-market speaker akin to the old Sonos:3. The Sonos Move comes the closest, and I would consider getting a second one of those instead of using the Fives in the apartment. We’ll see.

Windows 12?

andrew b. asks:

You and Brad discussed a Windows build on one of the rings being “Windows 12.” What, exactly, is Windows 12? What is its purpose?

So let’s step back for a second and ask those questions about Windows 11 first. What’s the point of Windows 11? And what is it?

I think the point there is to reenergize customers and boost new PC sales in the wake of a pandemic, a time at which PC sales were predicted to fall (but not as hard as they did). It’s just a new version of Windows 10 with a new UI. That latter bit is important: upgrading is no more difficult than moving between two Windows 10 versions, but the perceived value is higher.

So what is Windows 12? As I understand it, it’s a new version of Windows 11 that will almost certainly be as easy an upgrade as was Windows 11. Its selling point, its purpose, is to bring AI to the Windows desktop broadly, a move that Microsoft hopes will reenergize customers and boost new PC sales.

Put another way, Windows 12 and Windows 11 are essentially the same thing, the next version of Windows, and an arbitrary line to draw for support purposes. Windows 11 has arbitrary hardware requirements, but Windows 12 will have less in the way of arbitrary requirements (or at least recommendations) because that AI stuff will require a neural processing unit (NPU), hardware that is only just now beginning to appear in new PCs.

Given that you mentioned Windows 11 has no LTSC version, do you see Windows 11 ever having an LTSC version? Is its absence an acknowledgement by Microsoft that Windows 11 is not ready for business? If so, was it intended to ever be?

Yes. Apparently, Microsoft did say that an LTSC version of Windows 11 would arrive in a few years. This could be tied to the EOL of Windows 10, or not. It may only happen with Windows 12, given the timing. I think I mentioned that there might be parallels with how .NET ships in an LTSC version every other version, and maybe Microsoft is looking at that for Windows, too. So Windows 10 was LTSC, Windows 11 was not, and Windows 12 will be. Maybe.

But yes, there will be some new LTSC version of Windows in the future either way.

Microsoft’s Windows strategy makes no sense to me (outside of “make sure we force Bing/MSN on them”)

See, it does make sense to you. 🙂

Less cynically, one of the challenges today with Windows is that it’s a free upgrade, so monetizing it is difficult. And people upgrade PCs much more slowly, exacerbating the issue. And we don’t talk about this too much now—it was a big deal with Windows 10—but Microsoft clearly still wants Windows to fit in with the whole cloud service/subscription model that is so successful elsewhere in the company. They know they can’t just charge subscription fees for Windows directly, no matter the benefits. So they nickel and dime us with these other monetization efforts, among other things with sponsored ads in Start, forced usage of Microsoft online services and ads, and with secondary subscriptions (and the resulting ads for them; Microsoft just advertised OneDrive to me today in Word, for example). I would personally pay for a less disruptive experience. But they seem disinterested in that, even as an option.

Related to this, spacecamel asks:

Any early indications on where Microsoft will add advertising in Windows 12 based on what they are updating?

I’m away, so I haven’t had a chance to even look at a Canary build yet, but I’m also curious to see what people like Rafael Rivera find hidden in these builds as we move forward. Nothing yet. But maybe AI is where Microsoft “figures out” a subscription fee for Windows: it could charge for AI functionality somehow and/or make it require a Microsoft 365 subscription. This is the type of thing I expect.

Floppy disk games

j5 asks:

What’s your favorite floppy disk PC game?

By the time I moved to the PC—roughly 1994—the Internet had happened, and I was downloading games like Jill of the Jungle, Commander Keen, and Wolfenstein 3D to my first PC from local BBSs. I did have floppy drives, of course, but hard drives were common by then, and I only used them to install things like MS-DOS and, later, to boot into a DOS that was customized to help DOOM run as well as possible. (My first PC was a 386SX that I built myself.)

I was an Amiga user from about 1990 until 1994, and floppy disks were a big part of that experience. All games came on floppies, and few could be installed on a hard drive. My favorites included some familiar Psygnosis games like Shadow of the Beast, but also lesser-known titles like Leander and Agony (all three of those were made by), and many, many Cinemaware titles like Defender of the Crown, It Came from the Desert, and especially Wings. I spent so much time with these games. And many others. Arkanoid. Turrican. It’s hard to remember them all.

Mine was Pool of Radiance. Horrible graphics and terribly slow. But I’ve such fond memories of staying up late playing that game. Writing notes about places and things I figured out, lol what a dork!

I did this all the time. I had a mouse pad that had a see-through cover, and I would notes under there for Shadow of the Beast and other games. The Beast one had a hand-drawn map so I’d know when things would happen so I could try and get through it. That game was brutally hard.

The games of today are just amazing! Graphics, gameplay, and story are better than the movies a lot of the times. This has nothing to do with old school PC gaming really. But I do miss the pick up and go play of older games. I was playing Golden Eye on the Xbox with my sons last weekend. The game is great as I remember it. But man it was annoying to me to have to wait for the Xbox to go through all the menus and then load up the game. A few months ago I had to help one of our son’s create an EA account to play some games. It was so frustrating. I didn’t want to use my personal email or his. So I created a throw away. And had to create an EA account before he could just play the game.

I would add the Half-Life games to that list, and the original game is now available in a remastered version called Black Mesa that was blessed by Valve. All the original DOOM games. (And offshoots like Heretic and Hexen.) Quake, of course. I have replayed many older Call of Duty campaigns repeatedly, and I would do so for the original Medal of Honor and its offshoots (one of which is Call of Duty, go figure). There’s nothing better than discovering that an old classic is still as good today as it was back when. Love that.

Designers vs. engineers

AnOldAmigaUser asks:

Do you think that the problems the Windows development can be attributed to the fact that they have put a designer in charge instead of an engineer? It seems with the inconsistencies from machine to machine that we are all part of the insider program whether we want to be or not.

I think you need to have both. The issue in the past—and this sort of came up above with regards to Longhorn—is that in the early days of personal computing, whoever developed an app or whatever would create the UI. And these people aren’t necessarily the right person for that job. One of the innovations in Avalon (which became Windows Presentation Foundation) is that Microsoft separated the UI from the back end, using XAML for the former and C#/VB for the latter. This allowed teams of people to work on different parts of a project, with a designer creating the UI in a graphical designer (Blend now) and the developers working on code in Visual Studio. This was a forward-leaning idea 20 years ago.

The problem I have with some of the modern design stuff in Windows 11, as noted in Care About Windows? Do Not Watch This Video, is that the people they chose don’t see to use or understand Windows at all. Thinking back to Joe Belfiore’s contributions to Windows 95 or early IE versions, or Hillel Cooperman and Tjeerd Hoek with Longhorn, they clearly cared about those products very deeply. But I don’t get the sense that anyone working on Windows 11 has any idea what they’re doing. This thing has been out for less than 1.5 years, and the UI has changed repeatedly. There doesn’t seem to be any strategy at all. It’s just lunatics running the asylum.

Anyway, the developer/designer mix is correct. It’s just not being used well by Microsoft right now. And putting stuff out publicly that hasn’t been tested? There is no excuse for that.