Regular readers of this column will know that I'm a fan of education and positive experience as an advocacy tool in place of shouting from rooftops. Winning the mindset of an average computer user -- particularly home users -- is never going to be a quick process but a recent experience showed me we still have some old and familiar hills to climb. How do we combat legacy reputations of GNU/Linux that are no longer valid?

Average users

A friend of mine has been using GNU/Linux -- Debian to be precise -- for about three years. Prior to that he had used Windows for several years. It was his choice to switch, but I helped him do it. He's taken to Debian with great gusto; in fact, he's often mentioned how much easier he finds his computer to use when he's using it. He has a "just in case" dual-boot set up with Windows XP but he rarely uses it because he finds it "slow" and "less intuitive". He'll ask me for advice on software choice because he's keen not to have to go back to Windows at all. He's by no means an expert but he's also no longer a beginner. He's quite happy to engage with the shell if required but prefers not to. I support him when called upon but rarely have to. He is analogous to a car driver who is confident in changing a lamp bulb. Sounds like an ideal conversion story doesn't it? And it probably is - yet even with him there are parts of the "reputation" of GNU/Linux that keep resurfacing.

He's had an ongoing issue with his printer: a classic case of a single flashing light which could mean a million things. He's browsed help sites and read PDF manuals and followed help guides with no success. Assuming the printer was dead his son-in-law donated a slightly newer one. It worked for a while but then developed a slightly similar fault. He went through the same process again with similar results. At this point he figured the common point must be his computer so he removed and reinstalled both printers in Debian (wizards make life so much easier for some people you know). Still no joy. Neither printer was playing nicely any longer. At this stage he called me.

Just for your information: the problem with the first printer was age: it had just plain worn out. The second printer had been sitting idle for about a year and the ink well had developed a dry blob which would refused to shift. Both printers incorrectly reported that there was an error with the ink cartridge. After spending several hours on them I came to the conclusion that he would have to buy a new printer. This was fine with him and I was about to leave when he said something which considering his GNU/Linux enthusiasm surprised me. He said "I thought it was me going mad, I tried everything -- I even tried them both from Windows 'cos it would probably work from there".

Legacy reputations

So there it was, a reasonably seasoned GNU/Linux user who had experienced great success with it, fell back to the old, old reputation that it's harder to get peripherals working with GNU/Linux than Windows. If you read back through my description above you'll note he had successfully installed both printers and then un-installed and reinstalled them later. Those procedures had "just worked" -- but the printers hadn't -- yet he still thought that perhaps using them from Windows would fix it. As it happened just installing the second one in Windows had been a pain -- he had no driver CD you see.

What bothers me is that someone who has had so much success with GNU/Linux can still believe that Windows would automatically succeed where GNU/Linux hadn't

I don't blame him for his assumption. I've had plenty of cases where printer manufacturers were doing special -- undocumented -- things with the Windows drivers. It's also fairly unfortunate to have two printers experience very similar problems one after the other so his assumption that it might be the PC was a pretty good guess. But what bothers me is that someone who has had so much success with GNU/Linux can still believe that Windows would automatically succeed where GNU/Linux hadn't.

I suppose I'll get some comments saying that peripherals don't "just work" in GNU/Linux but that's not my experience -- nor that of my friend. The truth is I've not had an issue installing a printer under GNU/Linux for years and I know of nobody who has. I know of a lot -- a lot -- of people who have installed and use GNU/Linux successfully. They take part in the community and by-and-large they are enjoying their computer experience again. Yet despite this the -- now largely false -- reputations that GNU/Linux is harder to use, more long-winded, only for experts and feature-poor still endure. In the battle to get people to accept free software in general and GNU/Linux in particular, this -- 17 years after Linus Torvalds' usenet post -- is what we are still fighting.

Fixing it

So how do we combat these legacy reputations? Here's what I did: reassuring him that assuming the PC was at fault was a pretty good guess I then asked him why he thought it might have worked in Windows. When he said "well sometimes things just work in that don't they", I reminded him of the trouble he'd had installing the second printer in Windows, and of the time his old-ish digital camera refused to work in Windows because the driver thought XP-SP2 was not the same as "XP or higher". When I asked him if he'd had any difficulty using the camera or printers in Debian before now he got my point -- I hope. His experience told him something different to any reputation he had heard.

Sometimes the only way to combat an old reputation (especially ones that weren't entirely true to start with) is with some -- positive -- experience of the reality. If these legacy reputations of GNU/Linux are going to die, if we are going to change people's minds, then we need to do it by helping people to install it themselves -- not always do it for them. We need to do it by walking them through solutions, not just pointing them to a how-to page and we need to listen to their views and correct them where required -- gently.