Migrating Back To Windows

Migrating people from Windows to a Linux based system is an admirable goal. For those people with no special needs or software, migration often is painless, possibly joyful. For users with special software needs, getting there is an uphill journey, possibly Sisyphean. For some users, the journey is impractical.

Some typical toe stubbers:

  • MS Office compatibility.
  • Adobe Photoshop.
  • Quicken/QuickBooks.
  • Tax software.
  • MS Exchange/Outlook/Live Mail.
  • Skype.
  • Fillable PDF forms.
  • AutoCAD.
  • Vertical and custom made software.
  • Seamless integration with online services.
  • Presumed reliance on the terminal.

Oh yes, let’s not forget games.

There are solutions but mostly only for the technically inclined. Such as dual booting, WINE, running Windows in a virtual machine (VM), running Windows on a VLAN isolated machine, or running a Windows machine in a network.

Discussing these work-around solutions results in a common response, “Why bother with Linux? Why not just keep running Windows?”

The majority of computer users are not technically inclined. They do not care about software ideology. They do not care about vendor lock-in. They are interested only in using the computer as a tool — a means to an end.

Many users are dependent upon vertical or custom software. For such users migrating to a Linux based system is impossible or impractical. Accounting and tax software are examples affecting home users as well as small business owners. Custom specialty software used in the medical profession is another example. HIPAA and PCI compliance and other regulatory requirements are mandatory for people using such software.

Recently I migrated a person back to Windows from a Linux based system. This particular person had three work flow problems in Linux: Quicken, tax software, and fillable PDF forms. He had been dual booting into Windows XP to use Quicken and the tax software. At his request I had disabled the XP network connection.

This person could have replaced Quicken with GNUCash but did not. To paraphrase his words, he did not want to leave his comfort zone. The GNUCash migration idea became meaningless because there are no meaningful tax software packages in the Linux world.

I offered to create a Windows virtual machine (VM). He could run XP in a VM using raw disk access. That sounded fine but tax software vendors no longer support XP. He would need to update to Windows 7 or 10.

The next idea was installing Windows 7 in a VM. Although not end-of-life (EOL) until 2020, that is only three years away. Retail versions of Windows 7 are not for sale. Buying a retail license second-hand is possible as well as an OEM license. Yet the Microsoft folks have infested Windows 7 with much of the same telemetry nonsense of Windows 10. They now only provide all-in-one, no questions asked, cumulative updates.

This person is not a technical person. Running a VM is standard for many people, but for this person is pushing his comfort zone.

When told there were no easy Linux based solutions he asked why use Linux?

I agreed with him.

I agreed to install Windows 10 and be done with the work flow obstacles. While mildly intrigued with concurrently running two operating systems, he decided the learning curve was too much and he wanted to just run one operating system.

He is not tickled about Windows 10.

I disabled as much of the privacy intrusions and telemetry as I reasonably could. I unpinned all Live Tiles from the Windows 10 Start menu. I unpinned the Edge and Store icons from the task bar. I remove the Cortana search bar. I unpinned the workspace switcher. I removed apps from the Start menu. At that point Windows 10 looked rather ordinary, which is what the customer wants. These simple tweaks do nothing to prevent telemetry or tracking, but does quiet the otherwise horribly noisy Windows 10 desktop.

Oh, by the way, while configuring this system I did get to witness advertisements in the Start menu. The user is the product.

Converting the Firefox and Thunderbird profiles and data files to Windows was more or less straightforward. While he decided to remain with Firefox and Thunderbird, this person wanted MS Office 2003 installed rather than continue with LibreOffice. Exchanging LibreOffice documents with MS Office users is not always a smooth experience.

This person has a deep memory muscle footprint with how he expects to use a spreadsheet. He wants to return to using Excel. He does not use spreadsheets for anything more complicated than lists. Spreadsheets are spreadsheets but he is convinced Excel is easier to use.

Likewise with word processing.

The lack of tax software, which rendered moot the idea of migrating Quicken, along with his inability to easily use fillable PDF forms, pretty much decided the outcome.

I do not have simple answers.

MS Office compatibility? While many people do not create complex documents, which should allow for shuffling files between LibreOffice and MS Office, sometimes the problem is as simple as habits or unfamiliarity. Most users do not know what to do with an odt or ods document. This is the world of non technical users.

Adobe Photoshop? Yes, there is GIMP. Now try convincing non technical users to migrate from Photoshop to GIMP. Habits tend to discourage change, especially with complex interfaces. Professionals and hobbyists might make an effort, but not non technical users.

Quicken and QuickBooks? Forget about migrating without tax software support.

Email dependencies remain problematic for migration. Thunderbird suffices for home and small business users but is not palatable for the enterprise.

Fillable PDFs? Adobe pretty much has a lock on that kind of document.

AutoCAD? I have yet to read of anybody who uses that app speak highly of any CAD app in the free/libre ecosystem. What exists is acceptable for home and hobbyist users but not professionals.

Seamless integration with online services? Still dreaming.

A presumed reliance on the terminal often ends any hope of non technical users even being interested. They are not going down that road.

Throughout the Linux environment, adding the final GUI polish necessary for non technical users seldom happens.

Rapid release cycles exhaust and frighten non technical users. Updating a computer operating system every 6 or 12 months is insane. Something always breaks. Few people seem to really care that something is always breaking.

The Not Invented Here syndrome does not help.

A sad part of this story is the folks working at Red Hat, Canonical, or SUSE possess most if not all of the parts of the puzzle. They offer Long Term Support releases. Yet they do not want to spend time supporting desktop users. They believe the profits are in servers and the cloud.

Free/libre software has much to admire and enjoy. The toe stubbing migration points are too much for non technical users to overcome. I do not believe the overall picture will change as long as geeks design for geeks.

Would pre-installing a Linux distro on consumer computers solve the problem? I do not believe the overall picture will change until managers at one of the big three Linux companies embrace desktop users. That is, pre-installing is not the sole solution. Big name business support must be behind desktop computers.

Don’t forget those vertical software needs. Either many of those apps need to be ported or WINE works perfectly with no exceptions and no glitches.

I have read many stories how somebody migrated grandparents to a Linux system. Reading those stories reveals grandma and grandpa do little more than surf the web and check email. The stories reveal that geek grandchild is the tech support for grandma and grandpa.

I know how these discussions go. People reading this article will advocate all kinds of geek work-arounds. They only affirm my foundations for this story.

Is Windows without warts and blemishes? Hardly. I find Windows frustrating because of the way so much is hidden from users, making debugging and troubleshooting a serious challenge. Yet Windows is all most non technical users know or are familiar. There is a huge big business company behind the platform. Considering the massive scale of Windows usage, my observation is, despite the frustrations, Windows more or less “just works.” I cannot offer the same claim with any Linux distro.

Many people have migrated from Windows to a Linux based system. Most do not have to deal with vendor lock-in or vertical software dependencies. More importantly, the typical people who migrate tend to be intrigued by technology. Most non technical users are not. They see technology only as a tool — a means to an end.

I am unlikely to again migrate a non technical person with vertical software dependencies. The experience is much like a wearing a size 9 shoe on a size 10 foot with a pebble inside. Possible yes, but not fun or rewarding.

I wonder whether I will migrate anybody again. An elderly couple who I helped migrate a couple years ago no longer use their Linux based system. They bought a tablet. They did little more than surf the web and check email with an online provider.

Lessons learned? People want to use the right tool for the task. They want convenience. They want familiarity. There is a lesson in humility too. A Linux based system is not always the right tool.

I enjoy the Linux model and approach despite the usability warts and blemishes — and there are many. I enjoy the basic ideology. I foresee myself using Linux based systems for a long time. Yet I am fortunate not to have any of the aforementioned dependencies. I am not naive or willfully ignorant. I have thought some about how I would handle vertical dependencies. Using Windows is not the end of the world.

The Year of the Linux Desktop awaits.

Posted: Category: Usability Tagged: Windows

Next: Thunderbird and Minimizing to the Tray

Previous: 90 Percent of Everything