The update of Windows October 10, 2018 has been bothered by problems. In September 2017, Microsoft said that it had…
In September 2017, Microsoft said that it had just released the “best version of Windows 1
0 ever.” A year later, as Windows engineers struggle with the latest version of the company’s flagship operating system, there is a convincing case that the update in October 2018 is the worst version of Windows 10 ever.
The month began almost triumphant for Microsoft, with the announcement on October 2 that its second Windows 10 version of the year, version 1809, was ready for delivery to the public, immediately on schedule. Then, just a few days later, the company took the unprecedented effort to withdraw its October 2018 update from its servers while investigating a serious data-destructive bug.
Also: Microsoft Serves 40 New Windows 10 Bugfixes
An embarrassing drop-drop of additional high-profile bug reports has continued throughout the month. Embedded Zip file support does not work properly. A keyboard driver caused some HP devices to crash with a blue death screen. Some system fonts are broken. Intel powered the wrong audio driver through Windows Update, which makes some systems suddenly silent. Your laptops’ brightness can be reset in any way.
And with November approaching, the feature update has still not been omitted.
What went wrong? My ZDNet colleague Mary Jo Foley suggests that Microsoft is so focused on new features that it loses the track of reliability and basic. At Ars Technica, Peter Bright claims that the Windows development process is fundamentally deficient.
Or maybe there is an even easier explanation.
I suspect that a large part of the debt comes down to Microsoft’s overriding relationship with one of the greatest management principles of the past six century or so: “What’s being measured is being done.” It’s really a good guide for any organization, but it also leads to a trap for any manager who does not even consider what is not measured.
Also: It’s time for Microsoft to bring Windows 10 Mobile back from the dead
For Windows 10, a large number of performance and reliability events are constantly being measured on all Windows 10 computers. These flows of diagnostic data come from the Connected User Experience and the Telemetry component, including the Universal Telemetry Client. And there is no doubt that Microsoft uses the telemetric data to improve the foundation for Windows 10.
In September 2017, Microsoft tells us about the 17 percent improvement in battery life in Microsoft Edge, which made start time 13 percent faster and saw a decrease in 18 percent of users who hit “some system stability issues”. All that information was translated into increased reliability, as measured by a greatly reduced volume of calls to Microsoft support lines:
Our internal customer support team reports significant reductions in number of calls and online support requests since the anniversary. During this time, we have seen a healthy decline in monthly support volumes, especially with installation and troubleshooting update requests that take the biggest dip.
Microsoft has carefully focused on things it can see in its telemetry panel, monitoring measurement values such as installation speeds, start times, and number of crashes. In terms of reliability and performance, Windows 10 is undoubtedly better than any of its predecessors.
Unfortunately, the focus has been so intense that the company lacked what I call “soft mistakes” where everything looks good on the telemetry dashboard and every action gives a success even though the result is anything but successful.
Telemetry is most effective for collecting data to diagnose crashes and hangs. It provides good feedback for developers who want to fine-tune the performance of Windows apps and features. It can do a great job to identify third-party drivers that do not behave properly.
Also: Microsoft Windows Zero Back on Twitter again
But telemetry fails to detect anything that’s not a crash or a unique failure. In theory, the high-power low-volume problems will be flagged by members of the Windows Insider Program in the Feedback Hub. There were actually multiple error reports from the Windows Insider program members over a period of several months, which flagged the problem that caused data to be lost during some upgrades. There were also several reports that should have taken up the Zip file before it was released.
Why were these reports missed? If you have spent some time in the Feedback Hub, you know that the quality of reporting varies wildly. As a Windows engineer, I complained to me: “We have so many problems reported every day that are variations of” dark theme sucks, you should die “that it’s hard to find the six issues on a real problem that we can not repro- house. “
In response to these missed alarms, Microsoft has added a new field for its problem reporting tool in Feedback Hub, to provide an indication of the severity of a problem.
The time will tell if this add-on helps or if the tester automatically exceeds each bug report frustrated. Even with that change, the latest issues highlight a fundamental error in the Windows Insider Program: Members are not educated in the field of software testing.
The real value of Insider Preview builds is not surprising to capture telemetry data from a much greater amount of hardware than Microsoft can test internally. As for the manual feedback reports, I’m skeptical that even an additional layer of filtering will suffice to make them manageable data.
If Microsoft will require most of its non-Enterprise customers to install feature updates twice a year, responsibility for testing changes in these features begins in Redmond. The two most serious bugs in this bike, both settled in a published product, were caused by a change in the basic work of a function.
Also: Top Ten Features in Windows October 10, 2018 Update TechRepublic  An experienced program test may have and should have captured these issues. A good test knows that the test edge falls. A developer who rushes in to check the code to meet a six-year shipping period will almost certainly not test all these cases and may not even consider the possibility that customers will use that feature in an accidental way.
Sometimes In the next few days, Microsoft will update the update in October 2018, and everything in the Windows-as-a-service world will return to normal. But coming next April, when the 19H1 version is approaching the publication, many people will hold the breath.
Windows 10 Telemetry Secrets: Where, when and why Microsoft collects your data
How does Windows 10 telemetry really work? It’s not a state secret. I have reviewed the documentation and sorted out where, when and why. If you are worried about private documents that accidentally leave your network, you might want to close the telemetry setting.
Two Windows 10 feature updates a year are too many
Opinion: The idea of delivering two full Windows 10 upgrades each year sounds good on paper. In practice, the Windows 10 upgrade cycle has been unnecessarily disturbing, especially for home users who do not have the technical skills to handle those updates.
Windows 10 1809 bungle: We will not miss early trouble reports, says Microsoft
Microsoft makes changes to its Feedback Hub after failing to detect early reports that flag up losses caused by the update of Windows October 10 2108.
Microsoft Stops Windows Unrolling October 10, 2018 Update: What Happens Next?
Just a few days after releasing its latest Windows 10 update feature, Microsoft suddenly stopped unleashing and pulled the new version from its download servers, as it explored “isolated reports” of a data-destructive bug. What are you going to do now?