In a previous post, I looked at the situation facing Windows 7 users in spring 2016, as Microsoft exerted increasing pressure on them to upgrade to Windows 10. That post observed that many people had expressed many concerns about Windows 10, to the point that some were considering a switch to some other operating system.
The key problem seemed to be that Microsoft was still taking desktop and laptop computing for granted — was repeating its Windows 8 mistake, that is, in insisting that a Windows 10 designed for smartphones, tablets, and other devices could simultaneously be optimized for mouse-and-keyboard computing. It appeared that the situation might just get worse — that Microsoft would make Windows 7 increasingly difficult to use, and yet that Windows 10 would continue to offer an unappealing alternative.
Looking at Some Windows Updates
One concern identified in that previous post had to do with delivery of Windows updates that would nag, push, or force an upgrade to Windows 10. There was a fear that some such updates might deliberately sabotage Windows 7. The concern about sabotage arose from user experiences (including my own) in which a previously stable Windows 7 system became unstable following an installation of Windows updates. Such concerns had reached a point where, for example, an Overclock.net post (archived) identified approximately 40 updates as “bad” and/or “not needed.”
I became aware of that Overclock.net post when I decided to check up on a list of Windows 7 updates that Microsoft proposed to install on my computer on March 21, 2016. The first item on that list was “Cumulative Security Update for Internet Explorer 11 for Windows 7 for x64-based Systems (KB3139929).” Unfortunately, that particular item did not appear in the Overclock.net list — which was not surprising, as that list had been posted back in January. Since I had been keeping current on my Windows updates and had hidden or ignored few, it was probable that I had already installed most of the updates in that list. As described in one section of the previous post, however, I had taken some steps to defeat some bothersome effects from those previous updates.
I wondered whether I should research each of the 17 important and optional updates that Windows update was presenting to me at this point, and should do the same with all future Windows 7 updates. I started with a search for that KB3139929 item. This search led to an InfoWorld (Leonhard, 2016) article, quoting Microsoft for the conclusion that this so-called “security” patch included admittedly “non-security related” changes to Internet Explorer (IE). Specifically, the patch modified IE to include “functionality . . . that lets users learn about Windows 10 or start an upgrade to Windows 10.” It seemed clear that this sort of thing did not belong in a security patch. Users were reasonably concerned that Microsoft could no longer be trusted to insure that its so-called “security” patches were indeed focused solely on security.
I ran searches on three Microsoft .NET Framework updates (KB3135983, KB3135988, and KB3136000), on four optional updates (KB3133977, KB3137061, KB3138901, and KB31339923), and on security update KB3138910. It looked like some of these upgrades could cause problems for some software on some systems, but none of these searches seemed to point toward immediate and significant concerns.
It seemed I could probably go ahead with the Windows updates that Microsoft considered “important,” as I had been doing up to this point. If any of them turned out to be bad, I could uninstall it, use System Restore, or restore a previous drive image. I made an Acronis image of my drive C, at this point, as backup protection.
But these Windows updates raised a larger question: at what point would I get tired of this sort of thing — of making safety backups and researching Microsoft updates — and finally decide to abandon Windows 7 in favor of Linux?
Microsoft’s Progress with Developers
While running the foregoing searches, I saw that Leonhard seemed to be serving as an informal watchdog on Windows 7 updates. For instance, a look at his home page yielded a post on a problem, apparently experienced by many users, in which Windows updates would take hours to complete, when there was nothing actually happening in the update process. Leonhard attributed this to an “intentional” effort, by Microsoft, “to aggravate or penalize people who do updates manually.” As discussed in the previous post, the suspicion was that Microsoft would use such aggravation to push Windows 7 users to make the switch to Windows 10.
With or without deliberate sabotage and irritation, Microsoft certainly was pushing Windows 10 pretty hard. And that was interesting. The foregoing searches had led me to a Reddit thread containing this exchange:
PARTICIPANT 1: If you’re concerned about Microsoft installing things on your computer via updates, now is the perfect time to switch to Linux.
PARTICIPANT 2: i want to. really really badly. i did an upgrade from w8 to w10 and the performancy is really horrible. . . . but the problem is: I CANT SWITCH. i have lots of software for development (git, node, sourcetree, wamp, winscp, virtualbox, sublime, etc, etc) which i either have to reinstall (with an insane amount of custom settings) or isn’t available / works differently on linux.
PARTICIPANT 3: Sorry, but that’s not an argument. When I started reading your message, I thought you were working with .NET Framework and a stack of Microsoft technologies. In that case you would be locked to that vendor and could not switch even if you wanted to, although I have seen my share of .NET developers who moved their development environment to a virtual machine . . . . Node.JS development is so much more fun on Linux, and I am saying this as someone who actually tried it on both systems. Sublime Text is working great on Linux (there is a native port, no need for Wine), git — well, it is pretty obvious, and for every other program in [your] list there is a (better) alternative.
PARTICIPANT 4: As a fellow Windows user myself, I always felt like the whole “I can’t switch” argument is pretty empty. The whole privacy vs convenience dichotomy is pretty well-known. You “sell” information about yourself in exchange for more convenience, some small benefit, a little freebie, etc.
PARTICIPANT 5: Nothing you are using is preventing you from switching, some of which is actually easier on Linux.
I read that exchange in light of the conclusion, in the previous post, that Microsoft was pushing Windows 10 in order to get developers to write programs for that new operating system. It seemed ironic that, by pushing Windows 10, Microsoft seemed to be adding to the reasons why so many power users and developers remained alienated from Microsoft. It was as if the untrustworthy corporate mentality examined in one section of the previous post was destined to poison Microsoft’s dealings with informed portions of its customer base, even when the company under new leadership was trying to depart from some bad aspects of its past. The story illustrated in the foregoing exchange was this: Microsoft was going to push, some developers were going to come along, but many others would not.
A search suggested that, gaming aside, the tide was running against Microsoft. For instance, the annual StackOverflow.com survey indicated that the percentage of developers using either Mac OS X or Linux as their primary desktop OS rose from 39% in 2013 to 48% in 2016, while the Windows share dropped from 60% to 52%. The arrival of Windows 10 may have hastened the exodus from Windows 7 and 8, but not all of those people went to Windows 10: the overall Windows share still dropped from 55% in 2015 to 52% in 2016. (But perhaps that would be reversed by Microsoft’s announcement, a week or so after this post, that Windows 10 would run the Ubuntu command line.)
The previous post cited former Microsoft executive Sam Ramji for the view that Microsoft’s latest steps “may have come too late.” At this writing, that appeared to be the best interpretation — that Microsoft would struggle to hold onto its advantages, but would see them slip away as it continued with poorly conceived solutions like Windows 10. It continued to seem that the company failed to grasp the magnitude of change needed to make Windows a compelling choice.
At this point, it appeared that long-term Windows loyalty, from developers and power users, would materialize only if Microsoft took radical measures. It was appealing, but it was probably not enough, to observe that teen programmers “don’t hate Windows 10” (Bort, 2015).
One radical measure, occasionally rumored, was that Windows could someday become open-source. Microsoft’s decision, reported in the previous post, to treat Windows 10 users as guinea pigs for the benefit of the company’s enterprise customers, did resemble the Red Hat strategy of creating Fedora “as something of a testing ground for the enterprise features delivered to Red Hat’s paying customers” (Brodkin, 2012). But as PCWorld (Hachman, 2015) noted, Microsoft was not about to give up billions of dollars in annual revenue from Windows licenses.
It appeared, then, that Microsoft would continue to develop solutions that would contribute to corporate profitability if they worked. It appeared, indeed, that the company was so committed to that route that it would tread upon the users of its most popular operating system ever. For users who did not wish to become locked into Apple hardware, the question remained: at what point would it be easier and/or more productive to switch from Windows to Linux?
And yet that might not be the key question. If Linux did have potential to derail Windows, and thus Microsoft, why had none of the big players made an investment in Linux? There was the example of Open Office, developed by Sun Microsystems as a free alternative to Microsoft Office. TechRepublic (Wallen, 2014) suggested that Google had done something similar, in using Linux as the base for its Chrome OS, and Wikipedia pointed out that Google had adopted Goobuntu internally (see also ZDNet, 2012). Yet as Abishek (2015) pointed out, Google remained noticeably short of outright support for Linux.
The explanation was presumably that Google and others wished to use Linux for their own purposes, but not to unleash its potential to the general computing public. In that case, the 2016 StackOverflow survey (above) could provoke some reconsideration. If Google wished to impede developers’ growing affection for OS X, it seemed the time for Google to acquire something like Red Hat could be nearing.