Off the Beat: Bruce Byfield's Blog

Last week, I suggested that the continued interest in GNOME 2 handicaps desktop innovation.  Since then, a proposal has been made that Fedora's next release should default to Cinnamon, Linux Mint's GNOME 2-like shell. My guess is that the proposal will find many in favor.

The popularity of GNOME 2 remains widespread and growing. In the social sense, it qualifies as a reaction -- an attempt to return to an earlier state of affairs, like the Counter-Reformation in the Catholic Church. It's a topic that has drawn more responses than most I've written about -- enough that, at the cost of being too meta, I want to answer back.

However, before I do, I should emphasize that I'm far from a fan of GNOME 3 and Unity myself. Both seem to me a triumph of usability theory over firsthand experience, and I have said as much many times. My concern about the GNOME 2 reaction has little to do with GNOME 2 itself, and nothing at all to do with change for change sake, much less with a marketing culture that always demands something new, contrary to what one commenter suggested. But I do maintain that the renewed popularity of GNOME 2 amounts to a rejection of experiment and innovation for no greater reason than one or two examples of them are flawed.

Arguments for reaction

People who defend the reaction often dislike the idea that their preference is conservative, so they find another name for it. GNOME 2, I have heard several times in the last week is "classic," not technophobic, but "familiar." If the reaction is a form of fundamentalism that is not the negative anyone might imagine, because fundamentalism, as one commenter expressed it is "grounded in what works best."

Such arguments sound familiar from politics. They are a quibble over semantics, designed to make the return to GNOME 2 seem logical rather than reactive. Their logic, however, is circular. How do we know that a desktop is classic, or "grounded in what works best"? Because many people want to use it. Why do people want to use it? Because it is classic and works best. By the same standards of logic, Justin Bieber is a great musician.

Moreover, the idea that the reaction is due to an appreciation of classic or functional design is hard to reconcile with the history of the last two years. Until Linux Mint's release of Mint and Cinnamon focused the reaction, people were considering other alternatives, such as Xfce and KDE. If an appreciation of classic design were the main reason for the reaction, desktop environments like these would have continued to increase in popularity as much as GNOME 2. Yet, except briefly before Mint and Cinnamon, they have not, gaining only a few percentage points at the most. 

Nor was the reaction simply a preference for GNOME apps. Both Xfce and KD run GNOME apps perfectly well. In face, Xfce depends on running tools for other desktop environments, having only a handful of its own.

Rather, like all reactions, the return to GNOME 2 seems to have begun with a response to GNOME 3. Users are not just rejecting the most radical features of GNOME 3, such as the overview mode, but the GNOME Shell in its entirety.

If you spend any time in GNOME 3, there are some features worth borrowing, even if you reject the design as a whole. One such feature is the automatic management of virtual workspaces. Users should be able to toggle it off, but this automatic management might be said to be a benefit in that it introduces users to the advantages of multiple desktops, teaching them to work efficiently.  Another useful GNOME 3 features include less intrusive notifications, and chat that can be used without losing window focus.

So far, however, no re-creation of GNOME 2 has made any attempt to borrow or re-invent the useful aspects of GNOME 3. GNOME 3 as a whole has been rejected.
Instead, users have opted to return to a number of design flaws that seriously undermine the claim that GNOME 2 is a classic desktop. The classic menu, in particular, is so unwieldy as to be unusable on a modern computer with hundreds of applications. The same is true of the taskbar and virtual workspace pager. GNOME 3's redesigns leave plenty of room for improvement, but at least they try to address such problems. By contrast, returning to GNOME 2 means enduring design weaknesses that are only tolerable because they are familiar. 

As far as GNOME 2-like desktop environments innovate at all, they stick to minor things. The largest innovation that comes to mind is Linux Mint's Nemo, which is a fork of GNOME's Nautilus to restore the design that GNOME 3 has abandoned -- a design that was itself considered controversial when introduced half a dozen years ago, yet which is now considered classic. By definition, you can't expect a reaction to introduce much that is new, even if developers weren't hyper-vigilant these days about introducing changes.

In the end, the reaction does not seem to have been about features, or even convenience. Rather, it seems an expression of anger. The GNOME project took eighteen months to respond publicly to user dis-satisfaction, and even then the response was the indirect one of announcing support for a GNOME 2 experience via plugins. Faced with apparent indifference, users responded by rejecting GNOME 3 in favor of what it was meant to replace.

Searching for the Excluded Middle

Don't get me wrong -- I see nothing greatly objectionable about GNOME 2, aside from the fact that its Best By date is past. I used it for six or seven years myself, and I use several GNOME 2-like environments semi-regularly. Mate and Cinnamon remain, respectively my third and fourth choices of desktop after KDE and Xfce.

What I have trouble with is the either-or quality of the choices people are making. In the last year, the free deskup has swung from a period of irresponsible innovation to one in which effort is spent re-creating what already existed and was straining at its limits several years ago. Although innovations made without user consultation seem misguided, there is not much of which to approve in the throttling of innovation, either.

What the free desktop needs is innovation with user consultation, a controlled growth that allows new solutions to be explored, yet permits users to opt out of change without being penalized by reduced functionality and efficiency. 

Currently, KDE development seems to be attempting to keep this balance. However, with the GNOME 2 reaction, the GNOME world is nowhere near it.

Comments