I’m Joseph Conway, Senior Program Manager on the Windows Fundamentals team at Microsoft. The Windows Fundamentals team focuses on the underlying technologies that are used to install feature updates as well as a host of other things. We’ve heard your feedback about the lengthy amount of time your PC is unusable during a feature update installation, and we’ve been working on ways to decrease this time. Today, I am excited to share more details about the improvements we’ve made to the feature update experience.

Feature update cadence

As many of you know, the feature updates for Windows 10 occur twice a year. A feature update consists of upgrading Windows to the latest version with enhanced features, security improvements, etc. A feature update is different from the quality updates that are released on “Patch Tuesday.” Quality updates are mostly comprised of security fixes and are installed after a single reboot, whereas feature updates are installed in a phased approach that requires multiple reboots to complete.

There are four phases in a feature update installation. Each phase is done in either an “online” or “offline” manner. Online phases occur while the device is being used and the operating system is running. This phase is not disruptive to the user as tasks happen in the background. Offline phases, on the other hand, cause disruption to the user because the device cannot be used since the operating system is not running. For the Creators Update released in April of 2017, the average offline time for a user was approximately 82 minutes.

Let’s get to the exciting news — we’ve made a lot of progress in reducing the offline time

To achieve this, we moved portions of the work done during the offline phases and placed it in the online phase. Because of these changes, the average offline time for the Fall Creator’s Update released last October has dropped to 51 minutes, a 38% improvement! But we didn’t stop there. We’ve done additional work in the upcoming Windows release to move portions of migration operations to the online phase as well. This has resulted in an overall reduction of offline time when installing builds in the Insiders Program to an average of 30 minutes. That’s a reduction of 63% from the Creators Update!

Here is a more detailed comparison of the old feature update installation process versus the new installation model and how things have changed since the Fall Creator’s Update was released:

	OLD Feature Update model	NEW Feature Update model
Online	PC checks for available feature updates (manually or automatically) Feature update payload is downloaded PC waits for a required reboot to begin update installation	PC checks for available feature updates (manually or automatically) Feature update payload is downloaded User content is prepared for migration New operating system is placed into a temporary working directory PC waits for a required reboot to begin update installation
Offline	PC reboots to begin update installation process User content is prepared for migration New operating system is placed into a temporary working directory Drivers and other required operating system files are migrated User content is migrated PC reboots and completes the update OOBE begins Average Offline time of 82 minutes	PC reboots to begin update installation process Drivers and other required operating system files are migrated User content is migrated PC reboots and completes the update OOBE begins Average Offline time of 30 minutes (Insiders Program)

Because of these changes, the online phase for the feature update will take longer to complete. However, this should not be noticeable to most users, as the setup processes run at a low priority, so they won’t have a large impact on a device’s battery life or system performance.

We’ll continue to pursue further improvements to the feature update installation process. If you would like to give us your input on these changes, please share your thoughts via the Feedback Hub.

Source: We're listening to you - feature update improvements - Windows Insider

Tweet

— Twitter API (@user) View on Twitter

Tweet

— Twitter API (@user) View on Twitter

Posted By: Brink
16 Mar 2018