The trouble with open source and the road to its widespread mainstream adoption is its inability to shake off the IT expert mindset, namely to think like a regular computer user. Although the beauty of open source is its variety and flexibility, this could also be its Achilles heel.
There are open source projects which compare on an equal footing to commercially produced 'closed' software.
Usability and deployment are key to increasing adoption especially in the desktop and SME (small or medium-size enterprises) server environment. But how far will a non-expert go to install and configure software?
Some users just want to run a piece of software without understanding or caring about library dependencies. They are easily put off and frustrated.
I recommended to a friend the windows version of the excellent image editor GIMP (GNU Image Manipulation Program). In the end my friend elected to install a piece of purchased software as he'd experienced file dependency issues during setup.
Had he read the website where the installer file originated from he would have seen some prerequisites for installation, but when was the last time you read the small print?
From my own personal experience, after months of perseverance I managed to configure and use my wireless PC card with my laptop, running Red Hat Fedora Core 3. That was after much trawling the net, compiling, un-installing and gnashing of teeth.
I observed in the news groups, bulletin boards and wikis a tendency to say RTM (read the manual) to naïve new user postings. All well and good, but where is the manual, and in what directory? When I finally succeeded to configure my wireless card (it was more by luck than judgement!), I was unable to contribute anything meaningful back to these forums.
With the backing of big established companies in the enterprise server space, there is no doubt that open source offerings will continue to contend with established proprietary software.
Indeed companies like O'Reilly Publishing fill the gap with excellent books covering open source topics. But for the smaller organisations and individuals is there a desire to go out and spend additional money on a piece of software they've downloaded for free? I'm not so sure.
Recently a fellow ITer, a devotee of Linux, described Redmond software as 'Fisher Price' computing. Mockery of the use of wizards, GUIs and help files is common place amongst the ranks of technically gifted.
But why would these companies invest time and money in devising and creating these aides to usability?
Perhaps to increase adoption, minimise support calls and seize market share. A casual user of open source may notice the occasional scarcity of help files and user friendly documentation. Life is too short for the command line, MAN and readme file - give me a wizard every time!
Should all this matter? In an ideal world people wouldn't care about the purveyor of a word processor, spreadsheet or operating system.
We should care about human-computer interface design and simple readable documentation. These are ideals to which any software project should aspire. Teach your kids to be vendor agnostic.
Many of the more astute readers are perhaps, quite rightly, murmuring 'Stop complaining! Why don't you make a contribution to writing a user manual?'
Well that's exactly what I intend to do this weekend: put off tackling the Serengeti that is now my garden and start writing a user guide to a lovely little open source Java based application I use.
