
Happy Friday! Here’s an epic set of reader questions to kick off the weekend and the coming start of summer a bit early.
madthinus asks:
I have been thinking about Windows 23H2. When will we test it? Where is it being build? What would be in it. Initially you speculated that the Canary builds will be early Windows 12, but looking at what is being exposed there, is that actually 23H2? Will the key features be what was showed off at Build 2023, key among them the never combine on the taskbar? Is it too early to ship Windows Co-Pilot in the 23H2 time frame or Developer Drive? For all the communication and transparity we actually know very little.
The dynamic here has changed dramatically, and this is something I still struggle with. You can see that in my endless discussions about how Microsoft tests (or doesn’t test) and ships new features, and how this system has evolved into one in which we can literally get individual new features at almost any time. My brain still thinks in terms of product releases, which are increasingly a thing of the past, at least from the perspective of new feature boundaries. Now, releases—like 23H2—are just milestones for purposes of support.
This is a long way of saying that we’re already testing much of what will constitute 23H2 and that, for businesses, 23H2 will largely be composed of features that individuals have already received from Microsoft in the form of quarterly Moment releases, monthly quality updates, and interim random (CFR) feature additions. That is, in any given quarter, Windows 11 users see some number of new features, and 23H2, being the culmination of one of those quarters, will see both some number of new features plus (for businesses) all the new features that arrived in the previous three quarters.
I find this … confusing. But it’s not just Microsoft: when Google discussed Android 14 at Google I/O it was asked why this release seemed kind of minor and various publications compared this relatively boring approach to Apple’s annual “big bang” approach where they still focus on a big milestone releases (iOS 17, etc.) that each come with some long list of new features. And the answer was interesting and bears on what Microsoft is doing: Apple still relies on these annual releases because they sell new iPhones (and whatever other devices), and the two are timed to occur with each other, with the software often including some features that are specific to the new version of the hardware. But Android is treated like Windows, it gets updates all the time, in the form of quarterly Pixel Feature Drops (for Pixels only), monthly Android updates, and routine feature updates through the Play Store (not just apps). Google is using the same “continuous innovation” strategy as Microsoft and that just doesn’t lend itself to big bang annual releases.
To be clear, I have to remind myself of this all the time. I’m so old school that I have a hard time dealing with the fact that each channel in the Windows Insider Program doesn’t map to some Windows version, as they did in the beginning. But if we look across the various channels today, and at the features each adds in new builds, we see some collection of things that will be added at different times to stable. Some will be part of 23H2 (which I think maps to the next Moment release). Some will be added in a monthly CU. Some will just appear as if out of nowhere, and only on some PCs at first, as CFRs. This is the new system.
Regarding Windows CoPilot, Microsoft said we can expect it this month. Whether they keep to that schedule or not, whatever, but it will be interesting to see where it lands (i.e. which Channel) and how quickly it comes to market (and, then, in what form). My guess is that Copilot is the marquee 23H2/Moment 4 feature. But maybe it’s one of those things that rolled out like a CFR, over time. Or ships in preview form. Or is delayed until later. It’s hard to say right now, obviously. But the way the system is now, Microsoft can insert this thing however and whenever they want. That’s the beauty or curse of continuous innovation, depending on your perspective.
The biggest weakness in the part of Microsoft we’re discussing remains communication. I don’t understand why it can’t be clearer about how and when it will deliver the features it does discuss. But … that’s been true for a long time. I don’t see it changing.
madthinus asks:
ReFS, is that coming back as a mainstream feature or why is DevDrive build on it?
It appears that ReFS is making a comeback of sorts after years of silence on Microsoft’s part. When they announced that ReFS would be used for DevDrive, I assumed that the reason was related to the fact that ReFS (for whatever) reason was more efficient for large collections of small files, like the typical GitHub repository and/or software development project, and that NTFS, likewise, might be more efficient for a smaller number of larger files. My brain is simple.
If you dig deep enough, you can find the features that NTFS supports and ReFS does not. On Microsoft Learn, for example, there is a list of features not supported by ReFS (albeit in Windows Server) that includes file system encryption, transactions, and disk quotas, and Windows cannot boot from an ReFS volume either. (There are several others.) I’m still not sure why DevDrive is better suited to ReFS, but Microsoft provided the performance numbers. I guess it’s real.
But based on what doesn’t work with ReFS, it looks like we won’t be switching to this file system across the board anytime soon. One wonders if the development of ReFS lapsed after the regime change in 2012 for political reasons. Maybe now that enough time has passed, that work is continuing and if it makes sense, maybe we will be transitioning from NTFS to ReFS in the coming years. We’ll see.
helix2301 asks:
I don’t want to get political but COD and Nickmercs are under fire from both sides. Do you think this has any effect on Microsoft Blizzard purchase? Maybe moving up that court date was not good idea maybe they should waited till the heat died down. Politics and or political pressure going into an election year sometimes can play a part in things all I am saying lol.
I’m not a big fan of game streamers and the silly personalities in that world—it reminds me of poker players on TV—so I had to look this one up. I guess I would say generally that one of the promises of Microsoft taking over Activision Blizzard is that they can/will fix what many see as a broken, bro-centric toxic culture there, and that what happened with his particular personality would have absolutely happened under Microsoft, which is about as politically correct as a company can be (for better or worse). I’m not sure that it has any impact on the acquisition.
SherlockHolmes asks:
I am testing the new Outlook on Windows for a few weeks now and just wondered why Microsoft is developing it so slowly. Do you have any news where this project is right now?
So I mentioned Microsoft’s inability to communicate above, and in that case (Windows), it’s not done maliciously, it’s just mismanagement. With Outlook, which is part of Microsoft 365, I see something a little different. Generally speaking, Microsoft 365 as an organization does a terrific job of communicating what they’re doing. (What I think it gets wrong is the haphazard way it rolls out new product features on different platforms.) And I think its quietness on the new Outlook is purposeful. It’s purposeful because there is a huge customer base that uses and relies on the classic Outlook application and they will not be moving to this new version, possibly ever. And so this situation is a bit delicate.
The new Outlook will appeal mostly to younger people who don’t have the deep-set legacy needs of specific features that are only available in the classic app. These people are more likely to quickly adopt products like Teams and Loop. I’m generalizing.
The transition from the old Outlook to the new will likely mirror the adoption of Teams, not in pure user numbers, but rather in hours spent getting particular tasks done. For so many people, the classic Outlook is their day, the central dashboard for the work they do. The new Outlook probably more appropriate for people who view Teams as the center of their day. This transition will take time. And some people just won’t come along for the ride. So they’ll retire or move on.
Honestly, Microsoft might have been better off coming up with a new brand for the new app. Too many people will look at it and say, that’s not Outlook. It’s the Outlook web app. It’s … Outlook Express.
Anyway, it’s definitely on a slow boil. That team add multiple account support in November 2022 and then finally added Gmail account support this past April. I suppose it could suddenly just pop out of preview at any time (Ignite?). But it will never satisfy the Outlook diehards regardless. And maybe it doesn’t have to: by switching to a web-based extensibility model that works in both apps, Microsoft can just keep both in market and let user choose which is more appropriate for them.
UPDATE: Microsoft just revealed that it will replace Mail and Calendar in Windows with the new Outlook. So perhaps the goal (for now?) isn’t to replace classic Outlook in Office, but rather to create a new Outlook Express. That is, you get a basic Outlook with Windows and a more expansive product with Microsoft 365.
SherlockHolmes asks:
Im using XBox Ultimate Game Pass on a PC and noticed that it is still required to use an XBox controller. Do you have any idea when cloud gaming will fully support keyboard & mouse? Thanks.
I was just (re)playing DOOM 3 via Game Pass Ultimate on PC and was curious that it didn’t support an Xbox Wireless Controller: that game would feel quite natural in that configuration. (And it did ship on Xbox back in the day. ) But DOOM 3 is also a classic PC game (on PC, through Game Pass) and so switching back to mouse and keyboard actually felt pretty natural.
But Xbox Cloud Gaming titles are not PC games, they’re Xbox console games. And so the experience is reversed: these titles were only ever designed for the controller. On mobile, Microsoft has added touch controls for some/many of these games, which is neat. And so it makes sense that it may similarly added keyboard and mouse controls to them on PC. My guess is that it plans to do so, but that it focused on mobile first because it’s the bigger audience.
Coming from console, I prefer using the controller. But I get that a PC gamer would want keyboard/mouse. I bet it happens.
jrzoomer asks:
Paul do you use the Command prompt and/or Powershell, or do you find that the Windows GUI provides everything you need to operate Windows efficiently?
For the most part, the GUI does do everything I need, and I suspect that’s true of most people. But there are specific use cases where the command line makes more sense. I use a command line every day to update the book files in GitHub, for example. And winget, the Windows Package Manager, is only available (natively) from a command line. That said, I use WingetUI (a GUI) to automatically keep apps up-to-date. That’s a much better experience, not because it’s a GUI but because it’s automatic.
This was a good time to ask this question. I spent the past several months reacquainting myself with the command line for the book, and chapters on Terminal, Command Prompt, Linux, and Windows Package Manager are all available now. (Only PowerShell isn’t finished, but that will happen soon.) And this has been an interesting way to flex muscles I’ve not used in a while, but I haven’t come across too many use cases beyond my GitHub stuff. The System File Checker is command line-only, for example. And I will occasionally use it to change a file extension since they’re not visible normally in Explorer. But not too much.
andrew b. asks:
The desktop PC next to me proudly displays a custom AMD Ryzen sticker designed to resemble AMD’s great Athlon stickers from the early 2000s.
Hah! Nice.
Are there any CPU, GPU or Windows stickers over the years that have stood out to you? Or are you an anti-sticker kind of guy?
No, I’m anti-sticker. Related, I don’t like seeing laptops with stickers all over the lid, it looks like a car with bumper stickers (“This car climbed Mount Washington,” etc.) all over them. I used to have a Commodore logo sticker (really kind of an embroidered thing with a sticky back) on the power strip under my monitor, but that’s gone now. I did like that one for whatever reason.
smartin asks:
I’m having problems keeping up with what time zone you’re in. Currently in the US Friday will be June 16…
(He’s referring to the fact that the original headline for this was Ask Paul for Friday, June 15 when Friday is June 16.) What can I say, dates are hard. I write that forum post, look at the date in the tray, and then unconsciously just transcribe what it says instead of adding a 1 to the date. I feel like this is what used to happen at the start of a new year and the first few checks you wrote, back when we used to write lots of checks, had the wrong year on them. Hopefully, this is just normal and not some early indicator of dementia.
christianwilson asks:
You talked a bit about advertising on the web on Windows Weekly and how troubled it is these days. I can imagine this is concerning for a site like this where, despite not wanting all the ads on the site, you need them to thrive. Is there an idea of what, if anything, Thurrott.com/Petri.com can do to address this? I’d like to think subscriptions alone are enough to keep this going but I’m sure you need more than that.
So I don’t want to get ahead of myself per se, but I mentioned some weeks ago that I’d have a major announcement to make about the site. That announcement is still pending, but we’ve made some structural changes and I’m more involved in the business side of things. And here’s what I can tell you: the ad-supported side of the business generates enough money to keep the site up, functioning, and paying the bills, but it’s not enough to pay my salary. And despite being a much, much smaller audience by volume, the Thurrott Premium side can pay my salary. So that’s the business in a nutshell: if we didn’t have the subscription, I wouldn’t be able to do this.
I’ve used this example before, but it’s a good one and quite relevant. Each quarter, Spotify releases its quarterly financial results and one of the things it communicates is how its ad-supported and paid subscription businesses compare. In the most recent quarter, Spotify had 317 ad-supported (non-paying) subscribers and 210 million paid subscribers. Ad-supported subscribers generated €329 million in revenue in the quarter, while subscribers generated €2.7 billion in revenue. So 40 percent of its customer base generated 89 percent of its revenues. This demonstrates, sharply, what a difference paying subscribers can make vs. free, ad-supported users.
Of course, 40 percent of Thurrott.com readers do not pay for a subscription. It’s probably not even 4 percent. (I honestly don’t know.) So I’ll be working for the foreseeable future. 🙂 Which is fine, I like what I do, and I like the honest relationship there. But I hate the ads. I wish there was some version of this story where some large enough audience paid just to support my work and we could forego the ads entirely or at least have fewer ads. But that’s not realistic, and it’s certainly not happening, and here we are.
As to what we can do, I feel like we’re doing what we can do. I would like us to be a bit more aggressive about pushing Thurrott Premium, perhaps, and we have some big newsletter changes coming up that will hopefully turn that into more of a money maker and less of a cost center. We’ll get to that as soon as possible, we just have a few things to work out. And yeah, I will be communicating the changes we’ve made and are making as soon as I can too. But it’s a net positive for you and me both, I think.
jimchamplin asks:
With Intel’s new horrible branding announcement -the branding is horrible while the announcement is simply meh – what are some of your favorite gawd-awful product names?
In the tech industry specifically, there is a weird dynamic where brands that seem terrible at first—Pentium is the classic example, or maybe iPad—are just accepted as normal over time and are then no longer controversial. But Zune was a bad name, I think. Wii U was pretty bad. Xbox is a good brand, by comparison.
In the Microsoft space, we’ve also had our share of long product names like Windows Phone 7 Series, Windows XP Service Pack 2 with Advanced Security Technologies, or anything with “R2” in the title. In many cases, these products had cool codenames that were lost when the product was released, like Avalon (Windows Presentation Foundation), Springboard (XP SP2), or Palladium (Trusted Platform Module). But every once in a while, the codename gets through, like Cortana. It’s too rare.
And let’s not forget the version numbering silliness. Windows 3.1 was a major Windows release. So was Windows for Workgroups 3.11. So was Window XP Service Pack 2. And so on.
There have to be a thousand bad Microsoft brands and product names.
Daninbusiness asks:
Have you heard anything about Windows on Arm lately? Still seems like a strategic benefit for MS to keep pursuing this though it also feels like progress has been glacial.
Honestly, I feel like Windows on Arm is in a good place. What’s holding it back is the hardware: we’re all waiting for the results of Qualcomm’s Nuvia acquisition to magically catapult WOA into relevance (if not semi-even competitiveness with Apple Silicon), and of course the Arm Holdings lawsuit over that threatens to scuttle any momentum.
But WOA is otherwise right there. WOA was, until this week, the only Windows hardware platform to offer Neural Processing Unit (NPU)/AI engine capabilities, and that was a big part of Microsoft Build last month. Microsoft normalized Arm a bit by offering Intel and Arm versions of its most recent Surface Pro product, and the ThinkPad X13s makes a reasonable case for WOA even on today’s hardware. But from a technical (software) perspective, WOA is pretty much ready: it’s gone from being purely 32-bit to soon being purely 64-bit, and aside from drivers, compatibility now is pretty much automatic across the board.
And so the big remaining piece is still hardware, as always. And that will/should be solved by that Nuvia generation of hardware. Qualcomm typically announces new chipsets in December, so that’s the expected time frame. And if they pull it off, we should see more new Arm-based PCs at CES than has been the case in recent years.
Way back in 2017, I wrote that for Windows on Arm to be successful, it must be boring, the idea being that success means that everything just works normally. We’re very close to realizing that now.
With technology shaping our everyday lives, how could we not dig deeper?
Thurrott Premium delivers an honest and thorough perspective about the technologies we use and rely on everyday. Discover deeper content as a Premium member.