Skip to content

Rant: the elusive open source/free software desktop

Every once in a while there is this happy, chirpy announcement of an organisation (usually a public body) that it will start a project to replace its current closed desktop with an open source one.  And while I hate Windows XP (any edition, and let it be known there is nothing ‘professional’ about Windows XP Professional) with the hatred of a few thousands suns, I strongly feel the vast majority of these attempts to be misguided. 2010 is not going to be the year of Linux on the desktop, nor is 2011 likely to be. I used to think that the year of the open desktop was about three years away, but that was in 1998 and we’re twelve years down the line now.Basically there are several major obstacles that have to be overcome and quite a few of those are strictly non-technical and therefore unlikely to be overcome by sheer geekery. Those are, in no particular order:

  1. Organisational inertia;
  2. Interdependence of systems;
  3. Lack of immediate benefits;
  4. Lack of mature desktop applications;
  5. Stockholm syndrome.

Organisational inertia

Basically, people rarely want change. For several reasons, it all depends on what are your drivers in life. Even if you value your own expertise, like most geeks do, you will dislike any change that will result in a write-off of previously hard-won expertise. The vi versus emacs was in a nutshell (either editor is awful in my opinion, just use joe). Others are just conservative for the sake of maintaining harmony. Opportunity driven people don not like to pay attention to IT because it distracts the organisation from business opportunities in the outside world (yes, there really is one). Or just from fighting challenges from that outside world. Change also may influence the balance of power, so any change that empowers the end-user will be fought tooth-and-nail by IT, generally speaking.

Basically, you will need a clear incentive to implement any change in IT in order to overcome organisational inertia. And the less incremental your changes are, or the more visible they are, the more likely you are to face this phenomenon. No matter how much shinier the new toys are from a new perspective, most people won’t care.

Interdependence of systems

Face it, most organisation IT-architectures look like a game of mikado. Pull something and something that should be completely unrelated falls over. Basically, the only architecture that does not suffer from this is one that strictly adheres to the ‘many small pieces loosely coupled’ approach. Which is a really good one to take. Sadly enough, there is this phenomenon called ‘network effects’ that has given every proprietary vendor of the past (and quite likely the future) a hefty incentive to consistently pick tight coupling over loose coupling. Which, given that they are not charities, have consistently done.

Lack of immediate benefits

So far no open source desktop provides a compelling immediate benefit. And by that I mean a quantum leap in functionality and/or user experience. At best a reasonably snug transition is provided (see Mint and Ubuntu). Basically, Windows XP and later are reasonably good (for quite a few values of ‘reasonably good’) desktop. And on those desktops where it falls short (laptops come to mind), the open alternative tends to fall equally short or even moreso. And that is from an end-user perspective. I am not qualified to judge it from a systems administration perspective.

Lack of mature desktop applications

This could be lumped together with the ‘lack of immediate benefits’ if it weren’t for the fact that there are immediate drawbacks in various fields. And no, OpenOffice.org is not a proper alternative to Microsoft Office, even though the latter sucks more than a Dyson vacuum cleaner when it comes to handling complex documents, such as large contracts. And we’re not even talking about its weak spreadsheet and database capabilities. Likewise for specialised accounting and finance applications. In the graphics department does the Gimp still not hold a candle to Adobe Photoshop, QCad is no Autocad. Generally speaking you can do a lot on a Linux desktop (I’m an almost daily Ubuntu user), but when it comes to that last 10% you need, you quite often have to resort to other systems. Any time spend on arguing that I am wrong here is better spend on actually fixing stuff, like the most popular feature request in OpenOffice.org that has been around for over five years now.

Stockholm syndrome

What? Yes, the Stockholm syndrome, an affliction suffered by people who feel helpless towards their captors and therefore develop sympathy for them as a cooping mechanism. It is the same with Microsoft, most people feel that they can’t do without them and therefore (at least subconsciously) are scared of the idea of having to do without Microsoft. The devil you know versus the devil you don’t know. In marketing literature this is also called ‘latent pain’, a problem you don’t see a solution to and therefore feel no need to address it. Especially in the corporate world Microsoft is seen as a problem without a feasible solution and therefore as a force of nature. It happens to us, and we’re helpless. This is a very strong psychological phenomenon and should not be underestimated. But one can dream.

So basically, are we screwed? Yes, we are, for the time being. Not that we should stop trying, but we should stop pretending that Linux will conquer the desktop anytime soon and that the only reasons for it being held back are general stupidity and cluelessness. Because that is arrogant and stupid.