Will there be a Windows 11?

That’s been the rumor circulating for the past few weeks, and it’s looking like the Windows event on June 24th will answer all the questions. Since I left Microsoft before conversations about this release happened, I can play the role of uninformed industry pundit and offer my expectations on what we’ll see on the 24th.

First off, yes, it’s looking increasingly likely that Windows 10 will become Windows 11. But this really doesn’t change much, beyond the marketing message that you would put behind such a change. I wouldn’t expect this to change anything related to Windows as a Service (meaning yes, you’ll still get free upgrades, and unless someone decides to make more unannounced changes to WaaS, you’ll again see those every six months), licensing, or SKUs (editions).

Why change the name at all? It’s really about marketing: It’s really hard to market something as “new and improved” if the name doesn’t change. In the past few years, that’s led to such fun names as the “Windows 10 Creators Update” (and even more “creative” names like the “Fall Creators Update”), because you need a name of some sort to be used as an “anchor point” for new features and capabilities.

Will there still be the “technical” names like 21H1, 20H2, 20H1, etc.? Yes, most likely, as it’s just a shorthand to tell you which specific release you are running. I would fully expect to see a 22H1 (or maybe just a 22H2, as there might not be any compelling reason, or even time to develop, for a 22H1) release on top of Windows 11, which would itself still be 21H2. But those names aren’t intended, and don’t really work well, for marketing purposes. You’ll only see them in WINVER so you can tell at a glance what you’re using.

So what will we see released? Here’s what I expect:

• Windows 11 in all the same SKUs (editions) that you see today. Everyone will be able to upgrade for free (subject to standard licensing restrictions, e.g. you can’t upgrade to a later Enterprise release if you stopped paying for your E3/E5 subscription — but you could move down to Pro).
• A “feature pack” that installs on top of Windows 11 to provide a new shell and related UI components (e.g. rounded corners). This will likely have the ability to be turned off, for enterprises that can’t handle change and want to continue to use the Windows 10 UI (for a period of time at least — eventually there will be no desire to continuing maintaining both UIs, so you’ll be pushed forward to the new one). This feature pack can be updated independently of the OS release, so it could evolve over time.
• An LTSC release that can’t get the new shell and related UI components. These LTSC releases are designed to stay exactly the same for their entire supported duration (now down to five years instead of 10, as Microsoft previously announced), so they are fundamentally incompatible with this “serviced and updated independently” concept. (So will this LTSC release still be called Windows 10 because it won’t have the Windows 11 UI? That is quite possible — it also helps reinforce the case to enterprises that this isn’t a general-purpose OS release that you should be considering.)

Yes, my expectations for the new functionality in the Windows 11 release match what others have suggested (in too many links to mention): A new coat of paint. Lots of eye candy. A new Start menu (without live tiles). Improvements to the clipboard and copy/paste. Updates to in-box apps (but not all of them). Updated settings pages (but not all of them). Fewer cases where you need to fall back to Control Panel (but you’ll still need to). A new store app (which might actually have stuff in it that you want, e.g. via links to winget). Some new fonts. Updated sounds (will the startup sound return?). A better out-of-box experience (OOBE) UI (which most enterprises never see either because of unattend.xml or Autopilot).

There is one feature that I would love to see: stub apps. Right now, the in-box apps are preinstalled (hence “in box”) and update themselves frequently, which generates plenty of pointless network traffic if you aren’t using those apps. Imagine if those apps were instead stubs that don’t actually install until you launch the app the first time. The savings in network traffic (along with some minor disk space reductions) would be nice. It’s not clear from the rumors if the apps would be stubs by default, or if they could be “stubified” based on non-use. I’m hoping for the former.

Interestingly, the only other “non-UI” improvement that I’ve seen discussed is linked to Bluetooth. There will probably be a few other features that “come along for the ride” because they were in development anyway — with the 20H2 and 21H1 releases being mostly “no new feature” releases, there is a backlog from teams who have wanted to actually ship new stuff, so there will likely be a hodgepodge of other stuff. Not necessarily exciting, just along the lines of what we’ve been used to with Windows 10 feature updates.

What about these “Windows 10 support ends in October 2025” discussions that have been going on over the past several days? That’s not really related — you just never noticed that the support lifecycle date of Windows 10 always listed a 10-year support lifecycle. That’s just a side effect of Windows 10 being covered by the “fixed lifecycle policy” and amounts to more of a “display glitch” that would have to be resolved before 2025 — plenty of ways to handle that without a bump to Windows 11, but certainly the bump gives an opportunity to shift that date range (e.g. 2021 to 2026 — with five years and no extended support, it wouldn’t change it by much, adding just a year).

What about the stories about the build numbers jumping from 21nnn to 22nnn? That’s just a sign of a new fork being created for the “next, next” release of Windows 10. The build numbers are increased to allow a little room for a few more builds in the current fork (Cobalt), and hence it’s a sign that a “release to manufacturing”‘ (RTM) sign-off is imminent. Not that it means much in this age of “anything can be fixed in the next cumulative update” structure.

Will there be a new ADK? Most likely, yes, as https://oofhours.com/2021/06/04/new-adk-for-windows-server-2022-released/indicated by the notes on the recently-released Windows Server 2022 version of the ADK. But I wouldn’t expect much new in this kit beyond what’s already in the Windows Server kit. Will there be a new version of MDT? Most likely not, as the ADK and unattend.xml are unlikely to change to require that.

I’ll be tuned in on the 24th to listen to the official announcement, which will likely be followed by a flurry of additional blogs and documentation posts that explain everything that the event won’t directly cover — it will focus on the splashy UI features, not the nitty-gritty stuff that IT pros want to know. Until then, enjoy all the rumors…