Microsoft this week reworked its Windows 10 development model, severing links between features and specific releases so that it can deliver the former “when they are ready.”
“While features in the active development branch may be slated for a future Windows 10 release, they are no longer matched to a specific Windows 10 release,” Brandon LeBlanc, a senior program manager on the Windows Insider team, wrote in a Dec. 16 post to a company blog. “New features and OS improvements done in this branch during these development cycles will show up in future Windows 10 releases when they are ready.”
Since Windows 10’s mid-2015 launch, Microsoft has publicly linked features to their intended releases, usually by introducing them in a blog post connected to the preview build in which they debuted. Microsoft has reinforced those connections by repeatedly trumpeting major new features each time it describes the coming release. On occasion, the linkages have had to be broken as the Redmond, Wash. company retracted a feature previously scheduled for a specific release. But that has been rare.
(One of the most prominent features backed out of a release was Windows Sets, which had been touted in 2017 and slated for a 2018 upgrade but ended up as a no-show.)
The difference between the old way and LeBlanc’s description of the new may be too subtle to be noticeable by outsiders. In the past, Microsoft almost certainly did not share everything related to scheduling the contents of a then-upcoming-now-past upgrade. Just as certainly, that will be the case going forward. What Microsoft doesn’t tell customers are, to quote a former Secretary of Defense, the “unknown unknowns — the ones we don’t know we don’t know.”
But by claiming “when they are ready” as the guide for future features, Microsoft implied that at least some had not been so when they were baked into past upgrades. That’s a puzzling tack, even risky, if customers pause a moment to absorb the phrase and inevitably wonder, “Wait, weren’t they ready before?”
Microsoft’s newfound stance on “not until it’s ready” was reminiscent of browser makers, who have long used that model as they crank out seven or more new versions annually. For browser developers, it’s nothing for a feature to miss the intended “train” or version; the feature can catch the next six or eight weeks later. That was the whole point of a more-is-better release strategy.
The analogy doesn’t track exactly, though. That’s because new Windows 10 features can catch only two trains each year, not seven or eight (as with Chrome) or 12 or 13 (Firefox). Miss one and there’s at least a six-month wait.
Or double that.
In 2019, Microsoft launched a major release in the spring and a minor one in the fall, making the latter little more than a rerun of the former, a “service pack” retread with few new features, none of significance. If Microsoft maintains that cadence in 2020 and beyond, delayed features of substance could be postponed a full year.
Speaking of major and minor….
Microsoft has yet to answer one of the questions about Windows 10 from 2019: Is the major (spring) and minor (fall) upgrade slate a one-off or the new normal?
More than a month ago, Computerworld predicted that Microsoft would tip its hand when it started previewing the version of Windows that was to follow 2004, the Spring 2020 upgrade that by accounts is finished or nearly so. If Microsoft began seeding Insider participants with early code for 21H1 (Microsoft’s code name for the upgrade eventually released in Spring 2021), 2020 would repeat the cadence of 2019, a major, then a minor release. But if Microsoft instead slotted next year’s second upgrade, 21H2, into Insider, it would be evidence that it was reverting to the 2017-2018 model of two relatively equal feature upgrades.
What Computerworld failed to consider – like someone asked to predict a coin flip not accounting for it to end on its edge – was neither. By separating features from specific releases, Microsoft is able to dodge the question of what comes after Windows 10 2004. What it will preview via Insider could be destined for 20H2, 21H1 or neither.
Nor did Le Blanc give any clues. “We may deliver these new features and OS improvements as full OS build updates or servicing releases,” he wrote, covering both major (full OS) and minor (servicing) releases.
Cynics might be tempted to see the whole instance – the talk of “when they are ready” – as a smokescreen that lets Microsoft put off deciding which upgrade cadence to adopt for 2020. There may be something to that: Microsoft has said it is “closely monitoring feedback” from the major-minor “pilot” of 2019. What with the year’s minor refresh, 1909, having been available for just a month and seeing as how businesses have likely done next to nothing with the service pack since, there probably is little to no feedback to monitor at this point. The firm may want more time to evaluate the 2019 cadence and whether it worked for the most important customers, enterprises.
By claiming that no feature will be assigned a specific upgrade before its time, Microsoft can continue to issue new code for Insiders to test without committing to a release model.
Although it can be dangerous to ascribe rationales to decisions others make – it risks organizing something without coherence – that’s not stopped Computerworld.
Separating features from releases so that the former is added to the latter only “when they are ready” may signal that Microsoft expects a slow, if not necessarily slower, upgrade pace – one that is, if nothing else, unhurried.
(If so, the debacle of Windows 10 1809 must still sting in Redmond more so than it has let on publicly. By slowing down the release tempo, Microsoft and its Insiders get more time to test.)
That, in turn, hints at the slower of the two cadences in question: 2019’s major-minor releases.
But if that is to be the case, why bother changing the feature-release relationship? Assuming the tempo of this year is retained next, what’s the purpose of Microsoft saying “((features)) are no longer matched to a specific Windows 10 release”?
Is it because the linkage will be increasingly unimportant, and because there will be really just one feature upgrade annually? What, exactly, was the purpose of this year’s 1909, an update that was in no way an upgrade, what with so few new features, enhancements or improvements over 1903? Was it only to have a release that qualified for 30 months of support for Windows 10 Enterprise users?
That could be arranged without the hassle of building a Potemkin upgrade for the fall: Simply announce that the spring refresh, the one that actually includes new functionality and features, will be supported for 30 months on Windows 10 Enterprise and Windows 10 Education.
And then boot the fall upgrade so that there’s only one each year.