Claiming that Linux users are different reminds me of F. Scott Fitzgerald's comment that "the rich are different from you and I" and Ernest Hemingway's alleged reply, "Yes, they have more money."
After all, computer users are computer users. A few geeks may argue over the differences in operating systems, but aren't average users more interested in simply getting work done?
Superficially, yes. But operating systems and applications are far from neutral. Behind the code and the interfaces are assumptions about how users should use an application and what they want and expect from an application – even about the relationship between users and an application and its builders.
Use an operating system long enough, and the assumptions behind it start to shape your expectations -- so much so that another operating system may seem hostile and bizarre.
You can hear the differences any time Linux users mingle with Windows and Mac users. The three groups have very different ideas about their relationship to their software, and communication is regularly confounded by differences in expectations.
So what do Linux users expect from their operating system of choice? I can think of at least seven replies:
Linux and its open source ecosystem have their roots in impromptu groups in which everyone is only an email or chat channel away. Even now, they have not quite outgrown this origin.
Finding Linus Torvalds' or Richard Stallman's email is not difficult, and, although the average user may not expect to talk directly to Jim Whitehurst, the CEO of Red Hat, they do expect that someone at Red Hat and similar corporations will listen to their feedback.
This attitude comes out very clearly in the reactions to Ubuntu's Unity or the GNOME 3 release series. Windows users might grumble about changes like the upcoming Windows 8 interface, but after venting they conclude that they can do nothing, and learn to live with the changes. However, Ubuntu and GNOME users reacted with outrage, many switching to alternatives -- and none of the scornful invitations to code for themselves have lessened their expectation that they should be listened to.
Linux users soon learn that their installation is backed by software repositories containing thousands of packages. Since many of those packages are redundant, often some well-meaning observer will remark that progress would be much quicker if projects working on the same general category of application would only pool their efforts.
But such remarks miss the point. Amarok and Rhythmbox, for instance, may both be music players, but they are also opposites in design. Amarok is designed for those who want every possible feature, and Rhythmbox is for those who prefer minimal functionality.
In just about every category of applications, a similar variety of choices is available, up to and including desktop environments and distributions.
Until a few years ago, the only exception was OpenOffice.org. Released as a complete office suite, for years OpenOffice.org had no comparable rivals. But with the fork of LibreOffice and the growing maturity of alternatives like Calligra Suite, even office apps are less of an exception than they used to be.
As a result, while most users have favorites, anyone who has used Linux for more than a year or so has probably experimented with alternatives. If an application develops in a way that users dislike, or a seg fault disables it until a bug is patched, many users know how to switch to an alternative and be up and running in ten minutes.
If users are loyal to an application, it’s because they like it -- not because they are locked in to whatever a manufacturer chooses to give them.
In the user revolts of the last four years, KDE managed to minimize protests by restoring the configuration options that users expected. At the opposite extreme, GNOME 3 has continued to be unpopular because it not only limits configuration, but is designed so that users have no choice in how they work on the desktop.
When Linux Mint produced Cinnamon, a reconstruction of GNOME 2 on top of GNOME 3, its success was attributable to both the restoration of customization options, and the fact that it allowed users to choose which aspects of GNOME 2 to enable.
The moral is clear: Linux users expect software to conform to their work habits, not the other way around.