GNOME 3.4: A Question Searching for an Answer

To anticipate the first question: No, the newly released GNOME 3.4 does nothing to change the general structure of GNOME 3. Nor does it install by default with any of the extensions that reproduce GNOME 2, although users can download many of them from the GNOME Extensions page.

Instead, GNOME 3.4 is an incremental release, focusing on polishing the GNOME 3 release series without introducing any major structural or conceptual changes. Specifically, the new release begins the implementation of the design concepts summarized two months ago by Allan Day.

Depending on your perspective, this is either an incredibly gutsy decision that shows GNOME’s developers refusing to compromise their vision, or an irresponsible ignoring of users’ feedback.

However, if you download the Live CD of the release, you’ll find more than cosmetic changes. You’ll also find a scattering of improvements in old apps, and the introduction of a few new apps as well.

But are such changes random? Or do they indicate trends in GNOME development? Such questions are much harder to answer than the changes are to describe.

The Redesign

With the GNOME 3 series a year old, the project is gradually redefining its Human Interface Guidelines. The 3.4 release marks a transition in the guidelines, with some applications conforming to the new design principles, with others still to come.

You can tell which applications have been re-designed to the new standards at a glance. Like all but a few small-windows utilities, they open full-screen, and the title bar has only a close button in order to reduce clutter.

However, one of the most noticeable features of the redesigned apps is the scroll bars. They are thinner than the old scroll bars, and omit the arrows at each end, which are usually so small that some users could have trouble clicking them with the mouse. Instead, users are expected to do what many have done for years anyway, and drag the much larger position indicator up and down.

The indicator, too, is easier to find with the default theme, being blue against a light gray background. By contrast, the older design made them light gray on dark gray, which makes the indicator harder to find.

The other characteristic of the redesign is the transfer of menus from the title bar to a single menu in the panel. But while the redesigned scroll bars seems sensible, this change seems arbitrary — at best, the difference between the titlebar and the panel seems minimal.

In fact, given that most apps open maximized so that users are less likely to lose track of them, there seems no theoretical reason for the move. After all, users don’t need to search for the title bar. The change would actually have made more sense in GNOME 2 than GNOME 3.

Nor does the change work well in practice. In GNOME’s web browser, the single menu on the panel means that items once clearly organized by a half dozen sub-menus are now tossed into a single menu, and are one more mouse-click away. In the text editor, the single menu means that only Quit is in the panel menu, and all the everyday options remain in the traditional menu, either because the editor is unconverted or because the panel menu can’t handle the complexity.

Add the fact that some apps are untransitioned, and some third party apps might conceivably never be squeezed into the new standard, and the change quickly moves from arbitrary to actively annoying. It doesn’t help, either, than items in the menu are separated by wide spaces, creating the appearance that some are missing.

Several higher-level decisions in the redesign also raise questions. For example, is the separation of mounted disks and documents from the file manager really an improvement? From one perspective, by making three apps where only one existed before, GNOME is adding to the complexity without getting many benefits. From another, GNOME is making functionality easier to understand by giving each function into its own application.

Similarly, when the Removable Devices icon is on the overview (in the lower right corner), why do the displays of mounted and virtual disks open in the workspace? Placing all disk displays on the same screen would only make sense.

Of course, the redesign will be clearer and easier to assess when complete. Yet, for now, it seems to be still evolving. Since natural selection is still going on, the results are not only mixed, but, in some cases, more complicated than the design standards that are being replaced.

Apps Old and New

Exploring the 3.4 release reveals dozens of random changes in different applications. For instance, the email reader Evolution now auto-detects common Internet service providers. Cheese, the camera operator, now defaults to WebM when saving videos. The Image Viewer now includes an optional side pane for meta-data. The Contacts app includes more fields for information. All these changes are welcome, but, considered together, don’t seem to suggest any particular direction.

By contrast, GNOME has taken some pains to renovate Epiphany, its Mozilla-based browser. Now rebranded Web, its minimalist design now resembles the Chromium browser more than Firefox, both in redesign and speed.

However, in adding a panel menu, Web has not opted for a traditional menu array for other items. Instead, they get crammed into another menu, this one on the top right of the browser window — complicating matters even further by giving two jumbled lists of features to search through instead of one. The result does nothing to make Web a contender against Chromium and Firefox, and you have to wonder why GNOME insists on keeping this also-run alive.

With this release, too, the GNOME 3 series has spent time on accessibility, which was largely ignored in last year’s 3.0 release. In 3.4, the Orca reader is integrated into the desktop, and improved contrast modes and an adjustable zoom help to make computing easier for those with weak vision. Given that GNOME 2 was the undisputed leader in accessibility on the free desktop, it seems inevitable that the project should have returned to it at last. If anything, more attention to accessibility seems overdue.

Other new apps continue GNOME’s trend of simplifying the display. The GNOME Disk Utility, renamed Disk, now both lists disks and serves as a partition editor. A separate utility called Document finds, displays and organizes various document formats, while Boxes is a virtualization image wizard and manager.

It’s probable that Boxes will become the major feature of the 3.4 release, since it not only fits into the design philosophy, but also makes virtualization much easier to set up. In the last few years, with all the talk about GNOME’s usability and interface principles, Boxes is one of the few additions that is simply an improvement that users will appreciate.

Back to Basics, or Frivolous Fiddling?

The jury is still out on GNOME’s redesign, as well as on many of the improvements and new apps. However, whatever the public verdicts on these things, the features that best summarize the overall approach in GNOME 3.4 are two of the most frivolous.

First, according to the release notes, Wanda the Fish, a silly but time-honored applet of GNOME 2 that wriggles across the screen, is back. This time, it’s as an Easter egg (and, no, I don’t know where it is, and haven’t bothered looking.). Perhaps the intention is to encourage exploration of the desktop environment?

Then there is the wallpaper, which lightens or darkens according to the time of day. This feature may have been inspired by a similar wallpaper introduced in an early Fedora release, or perhaps by the live (that is, animated) wallpapers available for Android.

These features are notable in that they are the first indications of triviality in the GNOME 3 series. Together, they might seem to deliver a message. However, what the message in their triviality is anybody’s guess.

One possibility is that they mark the moment when GNOME put aside major changes for a few years, leaving controversy and concentrating on incremental releases. They might signal that, instead of addressing the major issues of desktop environments, GNOME developers are back doing what they do best.

Another interpretation is that GNOME has no intention of addressing the questions about its intent and design that have been raised in the last year. Instead, its developers reserve the right to do whatever interests them — and, if what interests them is trivial, that in no way reduces their right to do as they please.

In itself, GNOME 3.4 is a release made up of tweaks and small innovations. But, like Wanda and the animated wallpaper, exactly what message – if any — the content of 3.4 sends about GNOME as a whole remains obscure.

Although the project has publicized the release almost as heavily as it did GNOME 3.0, and rebranded a number of apps, I can’t help wondering if even the GNOME project knows exactly what message it wants to deliver.

At a time when GNOME is fragmenting and possibly losing market share, this seems a question badly in need of an answer.