May 05, 2010, 12:39 PM —
I will 'fess up: I am a branding nerd. There's something appealing about a slick piece of branding... the very essence of a 1,000 words, reflecting corporate identity with a side order of commercialism.
"Branding" is one of those commercial buzzwords that has gained a bit more traction these days, as various Linux distributions try to differentiate themselves from each other.
Branding first appeared on the splash screens of Linux distributions: cleverly designed logos that briefly showed the world what kind of Linux you were using, then rolled away to reveal the stock GNOME, KDE, or other window environment/manager underneath. Maybe there was a stock background included with the distro's logo, but lately that seems to be regarded as gauche, as default backgrounds now are filmy, arty abstract paintings that designed to soothe the user with pleasant eye candy.
There were other, more subtle clues for the trained observer. When I would attend LinuxWorld conferences, I could usually tell what distro someone was running by the color scheme. Green was SUSE Linux or openSUSE. Blue usually indicated Fedora, with red reflecting, well, Red Hat Enterprise Linux. And if was brown, you knew it was Ubuntu.
Lately, branding has become a bit more pronounced. Canonical, in particular, has pushed a new Light theme for its default GNOME interface in Ubuntu 10.04, which not only shifts the color scheme from brown to eggplant, but also inserts (what Canonical believes) key interface enhancements into the software. Setting aside the argument of whether they were good changes, the key take away is that these changes will eventually reflect the Ubuntu brand just as much as their holding-hands logo does now.
And Canonical is very focused on that brand. Take a look at the branding pages for Ubuntu, Fedora, and openSUSE. The differences in approach are pronounced: Fedora and openSUSE seem more interested in providing technical details on how downstream and upstream providers can plug into their existing branding efforts, while Ubuntu's page seems to say "look how cool we are" and only a pledge to provide co-branding resources for the Ubuntu community--no resources to assist, say, a downstream developer who would like to make a Ubuntu-based distro.
Does this make Canonical a bad player in the open source community? If it were me, with the bias towards the consistency of a generally good set of branding, I'd say no. But today I read something from KDE developer/guru Aaron Seigo that made me stop and think about the negative consequences of branding.
In his blog, Seigo describes the market effects of a branded distribution versus one that rolls out with a stock desktop environment, which--in deference to Seigo--we'll say is KDE. At first, Seigo echoes the thoughts of what people like me might think when they see a "plain" desktop environment: that no effort is made by the distribution developers into look and feel so they must not care about differentiating itself over other distros.
But such positioning, Seigo elaborates, is basically setting Linux distros up to compete directly with each other, instead of cooperating more, as open source projects should be:
"The distinct brands choice goes something like this: each distribution can attempt to build their own brand, hoping to achieve enough consistency, enough market share and enough of a compelling story over time on their own to create significant value. This has the side effect of increasing the tension between distributions since it puts them in direct competition over our own 'heartland': the minds and loyalty of people already using F/OSS as well as those who might in the future choose to do so. So not only does it mean a smaller, more confusing and ultimately innefectual [sic] brand footprint in the general market, it also means higher tension between those who would benefit most by cooperating to grow the overall shared market position of F/OSS."
Instead of using distinct branding, Seigo advocates aggregate branding instead: contributing branding changes back upstream to the desktop environment, so everyone can share in the advances in design and technology. In short, be excellent to each other.
On paper, Seigo is spot-on in his assessment. Any time Linux distributions work towards differentiation, there's a certain bit of reinventing the wheel to be found. If KDE and GNOME are the Big Two Linux desktops, why trick them out for your distro alone? The problem is that as sound as Seigo's arguments are, I am not sure they will be fully implemented. The "problem" of branding is just a symptom of the larger issue of differentiation in the Linux community: the Law of Developer Ego.
This development law goes something like this "all problems have multiple solutions, and even if a solution is already present, I, as the developer, can likely build a better solution. It doesn't matter if I can build on the existing code; it's my project and I will build it my way, because my way is better."
The Law of Developer Ego is exactly why we have so many redundant tools for Linux available today. PackageKit works very well in Fedora, as does yum--so why do I need a Synaptic and apt for Ubuntu? On a fundamental level, they do the same things. Scale that redundancy up a bit, and then the need for a KDE and a GNOME tends to look odd. For that matter, why are there so many distributions, when we could share all advances in one super-distro?
Of course, I am being a bit facetious, which Seigo does not deserve. There are strong historical, technical, and legal reasons why Linux development has come to this point that are too complex to blame on one big "Law." But ego does remain a key motivator in decisions about product creation, because human beings strive to win. There are millions of years of evolutionary pushback on the notion of open sharing. We will share, but usually only with our tribe. So, even in free and open source software, we make or join tribes with which to share.
That gives us a KDE community and a GNOME community and a Fedora community... and on and on. And, with tribal thinking comes creating a sense of tribal identity. How do we make the things we create and share different than all the rest? How can they stand out?
You guessed it: branding.
I realize that I have compressed hundreds of years of sociological research into a few paragraphs of smart-alecky prose, but I do believe that in order to create the better world of aggregated branding that Seigo describes, there will have to be some core rethinking of values, even amongst FOSS community members who are pre-disposed to openness from the start. The long-term benefits of sharing may fight with our instincts, but they are too great to ignore.
We will need to realize that, in the immortal words of Charlie Brown, "winning isn't everything, but losing isn't anything."