Happy Friday! Let’s kick off the end of another busy and momentous week a bit early with another great set of diverse reader questions.

Windows vNext

wright_is asks:

Is there any news of when the next versions of Windows are due for release?

There’s no literal news, but we know that there will be a major new release of the Windows client this year. There are questions about the timing—many are reporting that it will land earlier than usual instead of the usual September/October timeframe, but other reports disagree—and the name/brand. I suspect it was supposed to be called Windows 11, but Microsoft has reconsidered that given the renewed attention it’s giving Windows 10 and because it would be confusing to customers.

Our supplier has said that now that Microsoft has dropped support for Windows 2019, we have to upgrade all of our 2019 servers by May and when we go to Server 2022, we will have to upgrade them again in 2026. We would like to go direct to 2025, if possible, as we don’t want to go through the process of upgrading a whole farm of servers every 2 years!

(Yes, I know that Server 2019 gets support until 2028, but the supplier is only supporting versions of Windows under active support (new features), not long term support…

Windows Server is not really on my radar at this point, and I don’t see any official roadmap or similar. Microsoft is obviously testing future versions through the Insider Program, and it’s not surprisingly moving at a much slower pace than the client versions, based on what little I could find out in the world. The Tech Community site has a Windows Server Insiders blog that suggests the last build was in December 2023, and that was for the LTSC version. But mapping these things to traditional product versions is just as hard (impossible?) as it is for client, from what I can tell.

I can ask Mary Jo about this, but perhaps someone at Petri would have a better idea.

UPDATE: Mary Jo just tweeted about this, lol. –Paul

Epic v. Apple

jrzoomer asks:

Paul what do you think of the whole Apple vs Epic thing? Specifically the part where apparently Apple will now allow outside links for purchases–but Apple’s new rule is that they will still charge developers a 27% fee to developers who now have to report to Apple. I think though it may be technically legal to do this, the phone is such a large and important market that there should be a set of rules or laws around this

I’m going to write about this separately, so let me just hit some high points here.

The two major App Stores charge unreasonably high fees, and they should be legally required to match the range that credit card processing services charge.

The “overhead” of the App Store is illusory as it is fully subsidized by Apple’s unusually high (40-50 percent) margins on the iPhone, the best-selling phone in the world, and its other hardware.

Apple’s security story is a fantasy: This happens entirely at the OS level, and has nothing to do with the Store, and there’s no reason side-loaded apps from any source can’t be as secure as apps delivered from the store.

Apple is so belligerent that it is going to the trouble of making iOS and the App Store work differently in different parts of the world so that it can retain as much of the revenues from its unfair App Store fees for as long as possible.

When I look at Big Tech and the regulation it so badly needs worldwide, fixing app stores is Job One.

There’s a lot more. But again, I’m literally writing about this now.

Rabbit, rabbit

JHeredia asks:

Hi Paul, I was wondering if you had seen (and thus have any thoughts) the Rabbit R1 from CES? There are a few videos on YouTube about it and one from them demonstrating what it is/does. It seems like what the next goal for AI should be and what the next major step toward “Ambient Computing”, which I believe you have said you see as the future.

Yes. But the ambient computing future I once wrote about is not about adding yet another device. It’s about there being sensors everywhere—in homes, public spaces, businesses, airplanes, whatever—that identify us biometrically and then know to respond when we ask questions or whatever.

This is perhaps a device that can happen on the way to that future, but it’s also just for the one person and is yet another device. And we already have personal devices—smartphones, especially, but also wearables and, soon, embeddables—that are with us all the time. And some ambient devices—smart speakers and screens, etc.—that are in some places, but usually just constrained to individual or family usage.

When I Google this to find the website, I got an AI-generated summary that is perhaps useful to make the point. “The Rabbit R1 is a handheld AI assistant device. It is a standalone device that is not a smartphone. The R1 is designed to help people complete tasks that typically require using a smartphone or computer, such as booking flights, making reservations, answering questions, ordering pizza, and launching a Spotify playlist.” Which are all tasks we already do with the smartphone we all already own. And that reminds me of Steve Jobs’ initial marketing spiel for the iPad, which was that it wasn’t enough for this new device to do the same things as existing devices, it had to do them better.

Looking at the R1 objectively, what I see is a device that offers a simplified experience with functionality that straddles the smartphone and smart home device worlds. The most interesting bit, however, is the AI stuff. And there we have a problem …

The Pixel 8 Pro and Samsung Galaxy S24 series both have onboard NPUs and support onboard LLMs, and Apple has had NPUs for years (but lacks the LLMs for now), and this will become the norm in the coming year. So I’m not sure what hole the R1 fills. To get to truly ambient computing, we need to solve the smart home/personal/family bit first, and I don’t see an interim world in which existing devices—phones, wearables—and then embeddables can’t move us towards that.

It also struck me that it is a bit like Windows Phone’s “Hubs”… you go one place for what you want but then the “OS” interacts with the service/app. I think I remember you commenting that while this was great for the consumer, most app makers didn’t like it because it reduced the user’s engagement with them. Could this, and really ambient computing in general, suffer from the same problem? If “Rabbit” is viewing the apps/websites instead of the user, does that reduce the value of the advertising? Just wondering if you had any thoughts on this.

So this bit is interesting, of course, but this is also something one could solve on existing devices. There is a cottage industry around digital wellness now, with features built into device OSes, self-help advice of all kinds, and the like, but the trick here is that people don’t typically give up something they love so they can use something that does less. Brands are too powerful to allow any kind of heterogeneous hub solution to succeed. Look at the Apple TV app, for example: Even Apple can’t get Netflix and other services to integrate into that because it would benefit users, not its business. And people get sucked up in these brands, too. People literally identify with Spotify or Apple Music (or whatever). Overcoming these forces is maybe impossible.

For every person who tries a minimalist black and white smartphone home screen with four icons in a bid to end their doomscrolling addictions, there are billions who don’t care, don’t see the problem, and are saving up right now to buy the next phone. I feel cynical even writing this, and I spend a lot of time and effort trying to overcome this exact kind of (lack of) thinking. But we’re human, it’s hard, and even if I figure out something for myself, convincing others to do similarly is probably impossible.

God, this is depressing. 🙂

Blockers

Akis asks:

I followed your article and replaced Edge with Brave both on Windows and iOS. It mostly works greatly. Two issues I’ve observed:

The comments on the Thurrott site don’t display at all. Page loads and the comments’ section appears to be loading but this won’t happen. Unsure why.

There is nothing inherent to Brave in its default configuration that blocks the comments section on this site. But there could be a setting change that would do that. Or an extension or other configuration change. For example, if you use NextDNS, a Pi Hole, or a similar device- or network-wide ad/tracking blocking solution, you won’t see the comments on Thurrott.com until you add the domain “*.spot.im” (no quotes) to the allow list.

If you enabled Brave’s strict fingerprinting protection (Settings > Shields > “Block fingerprinting”), that might be the culprit. I haven’t done this, but this feature is going away because it breaks too many sites.

I have been using MS Authenticator app and haven’t found a way to make them work together. Whenever I am prompted to put a password, authenticator won’t pop up and I use the Brave password manager. Which doesn’t sync with Authenticator. Not sure I am doing something wrong and wondering whether you’ve experienced that.

Brave has one major issue, and it’s the reason I use a standalone password manager (Bitwarden, but that’s beside the point): It can’t do autofill on mobile, which is really surprising. (They’ve been working on this for several years, not sure what the hold-up is.) But it shouldn’t matter which app/service you use for password management and autofill, it should work fine with Brave.

For example, if I go to Adobe.com on Brave on mobile or desktop, click “Sign in,” and select the Email address field, I get a pop-up from Bitwarden with my sign-in credentials. I know it’s from Bitwarden because it shows the site icon next to the credential choice, and because I don’t have passwords stored in Brave anymore. So that does work.

But you use Microsoft Authenticator for password management. My understanding is that you can install the Microsoft Autofill extension to use those passwords with Chrome (or any other Chromium browser). (This is just built into Edge.) I haven’t tried this, sorry.

If Brave did autofill on mobile, I would just use its password manager.

Office is not Bing

TheJoeFin asks:

Your article on how Copilot in Office was helpful to you was very interesting. Microsoft seems to be moving very fast with Copilot and is rolling it out faster than most expected. Do you have any idea how Microsoft is judging the product as successful or not? Bing Chat didn’t significantly improve Bing usage; so is Microsoft expecting Copilot to increase Office 3 subscriptions, or do you think executives are more focused on Copilot becoming financially profitable on its own?

Copilot is an interesting service on any number of levels, but from a monetization perspective, it mirrors what we see with other GPT-style services like ChatGPT, where there is a free tier and a paid tier. This is basically the Spotify model, really: Most people will use the free version, which makes money from ads, but the company will make much more money from the much smaller audience that pays on a per-monthly basis. For Spotify, this is a reality. For Microsoft right now, it’s a theory.

But there’s a strong foundation of experience to support this strategy. Microsoft used to sell standalone Office versions to individuals in retail stores and via bundles on new PCs, but it found that most customers just kept using the one version they bought until they acquired a new PC, and as PCs got more reliable, the upgrade time frames stretched out. And as we see now with smartphones, that required a new strategy based on subscription offerings, similar to how it has worked with businesses for a long time. The result was Office 365, now Microsoft 365. And I’ve spent the past several years charting, quarter by quarter, how those two sides of the businesses have intersected, with “perpetual” (standalone) Office versions declining, revenue-wise, while the subscription revenues keep growing.

The interesting thing here, to me, is that there’s a strange gap in the numbers where there are roughly 300 to 400 million customers/seats on subscriptions but over one billion people still using the same old perpetual products. But those subscriptions, while smaller in number, are far more valuable to Microsoft because they are all recurring, annual revenues. Meanwhile, many of the people on perpetual Office may never upgrade. So it’s a lot like Spotify. Many “free” users, some subscription users, and the latter are worth so much more.

My theory is that Copilot Pro and Copilot for Microsoft 365 essentially represent a third choice in this matrix, a second subscription tier that a) requires you to first pay for an existing subscription and b) costs almost exponentially more than that existing subscription. And thus it should land in a familiar place where it is a far smaller audience than the other two choices, but be much more lucrative as well. That is, just as each Microsoft 365 subscriber is worth so much more to Microsoft than each standalone Office user, each Copilot subscriber is worth so much more than each Microsoft 365 subscriber. The number of users is much smaller, but the revenues—per user and overall—can be much higher.

Currently, I pay about $100 per year for Microsoft 365 Family. The actual cost is less because I have found ways over the years to pay less, but whatever, let’s just say that’s the price. And now I pay $20 per month for Copilot Pro, so instead of getting $100 per year from me, Microsoft is getting $340 per year from me. I know that AI is expensive for them, too, but those costs will come down, and ideally, they’re still making money or breaking even on day one. But it’s a smart play either way.

Fewer but more valuable customers sounds a bit strange. But that’s what Thurrott Premium is, too, when you think about it. You guys are way more valuable to me than those that don’t pay, even though you’re a much smaller audience. Focusing on you makes sense for the business, just as focusing on Microsoft 365 and now the paid Copilot subscriptions make sense for Microsoft. (Our businesses are so alike, lol.)

Speed dating

helix2301 asks:

Paul, What is the Byte cover I noticed behind you when you do Hands-on Windows?

That’s a poster-sized print of the July 1977 issue of Byte magazine. I received it as a thank-you gift from the person whom I gave my Byte magazine collection. He lives in Pennsylvania as well and knows I love trains, so it was the perfect gifts, and now that I have more space, I can finally hang it up in a place that others can see. (There was no room in our previous apartment.)

I pay for Club Twit, but will the Hands-on Windows videos ever be included with the book? I remember when tech books came with a CD. I know for years you mentioned to Brad about wanting to have videos with the book.

No, Hands-On Windows is owned by TWiT and that content is behind the Club TWiT paywall.

In a weird coincidence, I had just started to explore what it would take for me to include videos with the book when Lisa Laporte contacted me to see whether I’d be interested in doing that podcast. So that was kind of a no-brainer: I don’t have to do any editing or publishing, and I get paid for the podcast. Putting this type of thing in the book would have been a lot of work with little payoff.

Leo mentioned on MacBreak Weekly which I am sure you listen to 😉 that the GUI on Macintosh did not take off till Windows started doing it? Macintosh was released 40 years ago this past Wednesday. Just wondering about your take on this.

He is correct. The Mac was a failure, and while the advent of desktop publishing gave it a lifeline in the late 1980s, that was always a niche market. Microsoft popularized the GUI with Windows 3.x, and it went on to dominate the PC industry and defeated the Mac by a wide margin. Both companies stole their GUI ideas from Xerox Parc, but only Microsoft and Windows were wildly successful. And few seem to remember this, but many of the GUI innovations in the early Mac systems came from Microsoft, not Apple, as part of their partnership at the time.

Tech nostalgia

Kenneth_Burns asks:

Love the tech nostalgia posts. Do you know of books or articles that dive deep into early productivity applications? I’m fascinated by how people used microcomputers/home computers to solve problems back in the day, like this awesome woman on a 1988 “Computer Chronicles” who wrote a knitting program for her Commodore 64.

I’m reminded of this image I made years ago, the original being an Apple IIe user in the kitchen, because that’s where we were always going to use computers.

A great example of what I’m thinking of is the wonderful book “Track Changes,” in which literature scholar Matthew Kirschenbaum examines how word processing transformed fiction writing. I’d love to read the equivalent for spreadsheets, databases etc.

Interesting.

The only book I can think of that is semi-related is The Gutenberg Parenthesis by Jeff Jarvis, it’s a kind of history of print media with some ideas about how this form of communication will continue to evolve. I do read a lot of tech history books, of course, and while this isn’t exactly what you’re looking for, Sweating Bullets: Notes about Inventing PowerPoint by Robert Gaskins is an interesting history of a similar type of product.