What if Linux Took a Vista 'Mojave' Test?

Wednesday Aug 6th 2008 by Matt Hartley

Microsoft gave a "blind test" to Windows users to (supposedly) find out their true attitudes toward Vista. How would Linux perform under this condition?

With the perceived success by Microsoft regarding their Mojave experiment (at least in the eyes of Microsoft PR reps), I found myself wondering how today's popular desktop Linux distributions might do under identical circumstances.

Without actually duplicating what Microsoft put together with their Mojave experiment, some basic Linux usability factors can be examined by looking at the most common factors as seen by average users. And yes, because of the target audience of such an experiment, I left out security – clearly, based on current malware infection numbers, it’s not on the minds of the casual computer user.

So instead of yammering on about theory, I’ll present some of the scenarios provided in the Mojave experiment and see where Linux options would have put participants.

Backing up your system:

On Vista’s system, backing up is as simple as heading on over to the Backup and Restore center. It's very straight forward. For desktop Linux distributions such as Ubuntu however, there is as much "choice" as there is likely confusion on how to accomplish the same sort of task. First, Ubuntu (and most other popular distros) provide command line tools by default with anything GUI-related on a need-to-install basis. "Joe User" is not going to fool with this, unfortunately.

So it would be safe to say that in a Mojave-like experiment with many desktop Linux distributions, the user would not be a happy camper trying to use DD to backup their hard drive. Here's the direct comparison, using a GUI solution:

Vista: Go to Backup and Restore, choose whether to go with a full system backup or a partial.

Linux: One option that’s best for directory specific backups, works with most desktop environments, and provides sane/default folder selections as it’s installed, would be sbackup. This backup utility is not installed by default on distributions such as Ubuntu.

Another alternative would be either FlyBack or, someday after a little more work behind it, TimeVault. Both are promising.

Winner? Vista clearly would win in a grassroots comparison here based on the fact that most Linux distros to not come with GUI tools for backing up installed by default.

Connecting to the Internet:

A big selling point for participants in a Mojave-like experiment is the ability to connect to the Internet as easily as possible. And as broadband has become more prevalent, gone for most people are the days of struggling with DUN (dial-up networking).

Vista: Network and Sharing Center, create a new network and select the option that is relevant to your connection – wireless, PPPOE, always on, etc.

Linux: Both the GNOME and KDE versions of Network Manager can handle nearly anything you throw at it, hardware compatibility withstanding. What’s nice about both network managers is that all you generally need to do is use the pull down menu from the applet to get the connection setup.

Winner? With compatible hardware, Linux both wins and loses at the same time. If the user is working with an OEM machine designed for use with Linux, there is a clear win here, both with most broadband and wireless connections. This comes to light as I have seen issues with Vista's wireless compatibility despite providing Vista-ready hardware. Allegedly, most of this issue has been rectified, but some third party wireless vendors are still not providing the best drivers for the Vista OS.

At the same time, this is also where Linux tends to lose out as well. Despite the existence of vendors that provide PCMCIA, USB and integrated wireless solutions that work perfectly out of the box, most distributions put all of their effort into supporting wireless devices designed for Windows. This leads to the ever-frequent statement "Linux wireless sucks" which, comically enough, is only a half-truth uttered by Linux newbies.

One area Vista currently blows the doors off of desktop Linux distributions is with PPPOE. This is what many DSL providers use to get you connected to the Internet. This means entering a user name and password for basic broadband connectivity. So while those people using a router can simply allow the box to do the PPPOE work for them, many other users still need their OS to tackle this – using a simple GUI, provided by default. To my knowledge, without extra application installation, I cannot think of a single distro that provides a GUI tool for this.

Besides, when the latest distributions are still relying on apps like WvDial, something is drastically wrong.

Luckily for GNOME users however, salvation is on the way. NetworkManager 0.7 – everything from improved VPN, GSM and PPPOE support is coming...finally.

I suspect that Mojave-like Linux experiment participants would find that, with the right hardware, under most circumstances, Linux suits them just fine for Internet connectivity.

Peripherals for their computers:

When a casual user is looking to purchase a computer or simply better accessorize an existing one, they expect to be able to go to their big box store and buy whatever fits within the scope of their personal needs. For some, this is based on functionality, while to others, this means locating the best price possible. Generally, the two most common peripherals outside of a mouse/keyboard are the printer/scanner.

Vista: Unless the user is presented with an extremely dated or extremely new printer or scanner, the chances are excellent that it will work out of the box. If it is a newer model, then it may mean using the included driver CD, which offers extra software to provide maximum functionality for the end user.

Linux: Due to the fact that you will not find a printer or scanner with a CD containing Linux drivers for that device, it’s up to the vendor and/or community to ensure that these drivers are provided via CUPS/SANE.

In both the case of printers and scanners, support is pretty good. This said, if you are thinking all-in-one printer support, think HP.

Winner? Vista wins with the sheer number of compatible devices, while Linux wins with a better peripheral installation experience. With Linux, it will either be supported or it won’t. Despite power users being able to compile their own drivers into their Linux installation, Joe User is not going to be interested in anything short of plug and play.

Both Vista and Linux will provide a good UI for printing, while Linux tends to lean with extra software for OCR scanning support. On the Vista front, I prefer Windows Fax and Scan over anything else for my scanner, while in the Linux world, Kooka has never let me down.

To Vista's credit, Fax and Scan are provided by default while some Linux distributions give you... XSane by default. Xsane is the single most unusable, ugly, poorly designed application I have ever used on any platform – period. So including this as a comparable option to Vista's Windows Fax and Scan is laughable.

Software value provided by default:

Windows is a joke with regard to the software provided out of the box. It's true that Microsoft has improved somewhat with their various Windows marketplace efforts along with Vista, yet this does little good when the user is on a slower Internet connection or unfamiliar with software installation in the first place.

With Linux, in some instances the software is about as smooth as it could be. But many Linux apps are a bit rough around the edges with regard to overall feel and usability.

Vista: This Windows OS provides security features, a Web browser and an email client, among other accessibility applications. Connectivity applications are also well serviced with Vista - from wireless to bluetooth. Vista also provides access to software for video editing/producing, a photo gallery and the infamous sidebar gadgets.

Linux: Lacking the pre-installed video editing, photo gallery and sidebar gadgets, most distributions provide you with a Microsoft compatible office suite, powerful image editor, and MS Exchange compatible PIM (personal information manager).

Winner? Out of the box, I think Linux clearly wins due to its immediate availability of software that people really need right away. Both the PIM and office suite (Open Office) provided with popular distributions such as Ubuntu may happen to be less than glamorous with their outward appearance, but once you get used to them most people participating in a grassroots taste test will find them to be perfectly usable considering the cost – zero.

OS Familiarity:

One thing Microsoft knows is that OS familiarity is a zero sum game – if the user feels totally comfortable with the interface, it’s what they expect; users give no points for this. Unfortunately this begun to shift as Microsoft completely lost its focus with Vista, in my opinion.

Vista: For better or worse, Vista has a new workflow, and its menus and overall GUI layout have people either loving or hating it. Settings integration is generally good in Vista, despite the initial hardware compatibility issues early in the release cycle that dwarfed any visual benefit to how the UI was being presented to control hardware.

Today, outside of older PCs and select legacy applications, Vista is a perfectly usable operating system for those looking to upgrade from XP.

Linux: Desktop Linux is difficult to label as "familiar" for a couple of reasons.

1. People migrating from Windows for the first time are obviously going to be unfamiliar with the new interface.

2. There are multiple desktops to choose from, most commonly KDE or GNOME. Each has its benefits and drawbacks with regard to usability.

This does not mean someone who has just sat down in front of a Linux desktop for the first time in a Mojave-like experiment would be unable to use the operating system, nor does it mean that they wouldn’t be receptive to it. But for someone using XP at home, experiment with Linux is going to feel different.

Winner? If the user is able to get past the fact that GNOME, by default, is rather bland in appearance, I believe that (with a little time spent with it) they might find that Linux using GNOME is vastly more straight forward in application layout and desktop navigation. But again, I’m afraid that Vista would win regarding its settings unification. KDE overwhelms with its settings while GNOME's settings are all over the place. Sadly, much of this could have been fixed by installing the Gnome Control Center by default.

So, would desktop Linux do well in a Mojave-like experiment?

After taking everything above into consideration, I believe that a desktop Linux distribution such as Ubuntu or PCLinuxOS would do well by people taking on a Mojave type of challenge under controlled circumstances. Today's desktop Linux distributions are powerful, but for newbies, more often than not they’re too "geeky" for out of the box use without a little tweaking from your friendly neighborhood Linux geek.

With a helping hand to make the right hardware choices, to learn where to select the best software from download repositories, to drill into the minds of these new users that off-the-shelf software is not really an option, I could see desktop Linux becoming a real hit with 85-90 percent of people tested in a social experiment like the Microsoft Mojave project.

Many people believe that the success of Windows is due exclusively to market share. Speaking as someone who happens to be a full time desktop Linux user and a person who has used various distributions over the years, I would point out that people being asked to try Linux for the first time will expect the following things that currently are not being delivered.

1.) Consistency. For a Linux geek, choice is king. But to the casual user, they want a consistent interface. This means desktops, folks.

2.) Hardware compatibility. Despite the fact that to some degree Linux outshines Windows by leaps and bounds with installation of hardware that is compatible, locating compatible hardware means the user must know to look to "hardware compatibility lists," which is simply a foreign concept at this point, not to mention terribly inconvenient for most people.

3.) A single marketplace to purchase compatible hardware. This is one that I suspect will be happening sooner than later, but by simply pushing the OEM and designing distros to work with peripherals would do wonders for adoption among those who would tried Linux in a Mojave-like experiment.

4.) Clarification. One of my biggest pet peeves with desktop Linux is how the distribution supporters always seem to neglect to disclose early on that installing the distro onto a Windows machine is a "best of luck to you" proposition. Sure, more often than not it will be fine. But the marketing materials for distros such as Ubuntu make it seems as if everything is going to be just fine. It is only later that the user discovers the wikis and forums illustrating fact that some hardware does not play as well with Linux as others.

5.) Include the obvious extras! Honestly, it kills me that such obvious items as sbackup are not included with distributions such as Ubuntu 8.04 by default. Considering the fact that the competing operating systems have their best foot forward in this area, the very least Ubuntu devs could do is include something so painfully obvious.

Obviously most Linux users will exclaim that it is the user's responsibility to do their homework before trying a distro. And this is definitely true – when treating an operating system as a beta product or a hobby.

In the end, if the stage is set properly and the user's needs are met before show time with a public Linux tryout event, I’d say that Linux is indeed ready for the desktop for most users. I believe it’s simply a matter now of tackling the bugs and locking down what works, bringing together stronger unification with GNOME and providing clearer lines between typical and advanced settings views with KDE.

Bundle those improvements with a unified means of hardware selection for those who wish to build their own PCs or purchase peripherals that work out of the box, and offering a Mojave-like experience with distros like Ubuntu or PCLinuxOS would go over very well. The key is to be straight about any shortcomings and demonstrate the strengths of desktop Linux for the common user.

Mobile Site | Full Site
Copyright 2017 © QuinStreet Inc. All Rights Reserved