Windows 10 system updates are like Microsoft’s “bear kid”, almost every monthly cumulative updates and feature updates will occur in one way or another problem, resulting in some users not being able to use it properly. So why does Microsoft always manage these updates badly? Media Windows Latest recently wrote a detailed article on the reasons behind it.
Frequent updates of BUGS
The update to Windows Update began with the Windows 10 October 2018 update, which incorrectly deleted the user’s documents, pictures, and other file information after it went live. So just four days after going live, it was taken off the shelves.
After a major embarrassment for October 2018, Microsoft has taken a more cautious approach to the May 2019 feature update to avoid a repeat of the disastrous deployment. Although the May 2019 update is not a mess in itself, Microsoft has released a cumulative update to fix some long-standing minor errors, but the monthly update introduces a new error that results in high CPU usage.
Subsequently, in the release of the cumulative update to fix Cortana, the Start menu and taskbar are not working properly. It also prevents some devices from connecting to the network and causes audio problems. These problems were not properly resolved by Microsoft until the end of last year.
It is because of the many problems that led to the second half of last year’s feature update November 2019 (Version 1909) directly into a large cumulative update, not introducing new features, but focused on fixing various BUGS. However, this feature update still causes the file manager to crash.
So why is the Windows 10 update a mess?
Last September, Jerry Berger, who has worked for Microsoft for 15 years, released a video detailing the testing process for previous builds by the Microsoft operating system team.
Microsoft has changed its Windows Update testprogram, according to former employees, which may be one of the reasons for the confusion. As a former microsoft senior software engineer has said, Microsoft has an entire team dedicated to testing Windows updates. Unlike the driver and interface departments, the team member’s daily job is to discuss various failures.
Jerry berger also mentioned that the original Microsoft testteam set up special labs for such as Intel, AMD, Nvidia to test CPU/GPU. These specialized labs are used to test new builds or functional modules for compatibility or performance issues with important hardware such as processors and graphics cards.
The test team responsible for these labs will also interface with the manufacturer, so if the test team finds any problems, it can quickly identify and develop a solution. The test team passes the test and the code that fixes the scenario is merged into the main thread after the development team fixes the problem.
Instead of using virtual machines, Microsoft’s team of engineers uses automated testing and real-world devices to test Windows updates. In 2014, Microsoft fired the Windows testing team, which has spent most of its time stopping test updates to its actual configuration.
In addition to virtual machines, Microsoft now relies on Windows Insiders, a group of testers with enthusiasts and fans. Many users have already signed up for insider programs to test new features, and although Insider members have reported some problems, many of the problems have been ignored due to the large amount of feedback.
In the video review, Jerry Berger also discussed the Windows 10 test project, which simply doesn’t help Microsoft solve too many problems. The main reason is that most beta users don’t actively give feedback to Microsoft, although even feedback to Microsoft may end up being noone to do so.
The main reason for this is the dump log, which constantly records the situation while the system is running and generates an extremely large dump log for analysis. The reality, however, is that the dump log will only record its details if the system crashes, and other “small problems” systems do not log the dump log.
The full dump log is quite large, probably dozens of GB, hundreds of GB may also be at the TB level, and obviously most users do not have that large amount of space to store. That is, even if the user actively feedbacks to Microsoft and provides the dump log, the actual provision is only a partial fragment rather than the entire operating system full run log.