The release earlier this month of a preview Windows 10 build that isn’t due until 2020 was a little strange. At the time, Microsoft said vaguely that it was because of features that “require a longer lead time,” with no indication of what those features are.
The well-connected Mary Jo Foley tells a different story: the release is actually a consequence of parts of Windows’ development moving to the Azure group.
The core parts of Windows—the kernel, file system, networking stack, hypervisor, security subsystem, and so on—underpin a wide range of Windows variants, including Windows 10, Windows Server 2019, HoloLens, Xbox One, and Azure. According to Foley, Microsoft makes two releases of these core parts each year, in June and December. The various Windows variants build on these dual releases.
Recent releases of that core have focused on the demands of Windows Server, Xbox, and Windows 10, and as a result, the Azure team hasn’t felt the need to aggressively track and switch to each new internal release. Instead, the team is running an older version. But it’s the development of these same core parts of Windows that have been moved to the Azure group. As a consequence, Azure’s needs are going to become higher priorities, and it will be in Azure’s interest to keep track with each new June and December release.
That means, however, that Azure needs to catch up, and apparently this has shaken up the release timetable a bit. The Windows 10 April 2019 Update is almost here, built on the December 2018 internal release. Normally, that work would then be stabilized and baked into a June 2019 release, which would have been codenamed Vanadium, with that release used as the basis for Xbox, Windows, and Windows Server development in the second half of the year. But this year, the June 2019 release is being skipped. We’re still going to get an October 2019 feature update, but that too will be based on the December 2018 release. The Windows core platform won’t see an internal release (codenamed Vibranium) until December 2019.
This would explain why the 2020 work has started early—because that’s where the next Windows core release will come from, so that’s what needs to be tested. Exactly how it helps the Azure team is less clear, but I suppose scrapping the June release would give the Azure team more time to sync with the December 2018 release and align their platform with the rest of Windows. From then, things should be back on track, with a June 2020 internal release codenamed Manganese.
https://arstechnica.com/?p=1466679