Open Source: Where is the Innovation?

In recent years, we've seen the Canadian government pass all manner of laws to protect children. They outlawed certain chairs because of 1-in-a-million accidents, outlawed pit bulls (but not other aggressive dog breeds), and outlawed three-wheeled motorcycles. They didn't outlaw neglectful or foolish parents taking unnecessary risks. They outlawed the tools contributing risk over incidents with less odds of occurring than being struck by lightning. On the Michael Coren Show, a Canadian political roundtable, one of the guests remarked that society has moved from polices that avoid lawsuits to polices that avoid risk. It's as if the country believes that the pursuit of excellence is less important than avoiding pain--physical, financial or emotional.

News Flash: real life is a complex environment and in any complex environment there is always risk, especially in pursuit of excellence.

There is no innovation without risk. Most expert programmers are members of the IEEE or the ACM (Association of Computing Machinery). I'm a member of the ACM, and like most members I have a pile of unread journals and newsletters on my living room table. Where once the ACM was on the cutting edge, an organization that encouraged and inspired, now most of the articles are space filler. In the December issue of the Communications of the ACM, the president, David A. Patterson, laments the declining quality in the ACM's publications. The reason? Papers submitted to the ACM committees are graded on an average score: since innovative ideas will be unrecognized or threatened by some members of a review committee, truly important papers get lower average scores. They don't make it to publication.

People hiring software contractors often don't understand the price of creating something new. In the January issue of the Communications of the ACM, Philip Armour (often times the only part worth reading) talked about the confused perception over the risk of software project. Armour points out that clients will often agree to a risky project...and then demand their money back if the project fails. If you tell the client there is 15% risk, the client is likely to say, "That sounds reasonable." But if you compute the dollar value of the risk, $37,500 could be lost on a $250,000 project at 15% risk, the client may say, "$38,000! We can't afford to lose that!" Armour's argument is that people don't understand the price of risk.

Risk is not necessarily the same as reliability. Consultant Tom Dejarco in his book Slack argues that a software project that crashes may not be worthless...provided that the client can work past the bugs and get the results they are looking for. If the software provides a unique functionality, the fine tuning can be done later.

When it comes to innovation, open source is paradoxical. One of the reasons Linux was able to be developed so informally was that it was a copy of an existing product. UNIX has been around since the 70s and there were already many commercial UNIX-based operating systems. It was easy to pick-and-chose the best features from these systems to reverse-engineer them into a compatible operating system. As a result, Linux is revolutionary as a free challenger to the dominance of overpriced, poor quality alternatives, but from a developer standpoint, can any product based on 1970s ideas be truly called innovative?

As a friend of mine likes to say, if a committee choses what's for lunch, everyone will be eating hot dogs. To create innovative software, developers need to band behind and support visionaries creating cutting edge ideas rather than living in isolation and doing their version of That 70's Show. Innovation must be protected, supported and given a forum.

With Linux, open source developers have done the easy part: they've created a well-understood, widely available foundation for new projects. That was the easy part. The question is, will Linux developers band together and share a new vision, one to create a better world for the new century? Commercial operating systems have failed: open source is the hope for the future. Whether or not Linux developers will take the risk, it remains to be seen.

Talk back on the Linux Cafe.

March 19, 2005