Pros, cons of open source content management

As with all important areas of modern software, open source software (or Floss - Free/Libre Open Source Software - as it`s now known) has a play in the content management software (CMS) world. More correctly, it has numerous plays, some of which are at a level of sophistication equal to their commercial competitors.

So how does a business looking to undertake a content management project decide whether to pick one of these "free" products, or spend the money on commercial software, which often comes with some kind of use-based licence fee (whether periodic or volume-based)?

The Floss movement is trying its best to floss out what it perceives as the germs of commercial software. Richard Stallman, the high priest of free software, has famously said: "The enemy is proprietary software."

What Floss offers is software that is "free" in a more limited sense than "free beer", in its oft-used analogy. They mean "free from entanglements", usually of the legal kind. The "open source" bit is where a lot of the real action is. It means not only is the source code for the software available to anyone, but, usually, anyone is able to contribute to that code. Thus the software is built and maintained by a large community of contributors - not one isolated corporation.

The three key benefits of Floss are the lack of proprietary lock-in that a company holds over its clients and users, the distributed nature of code development and diminished (or in some cases no) licence fees.

The former gives the customer power to define its own destiny. The latter ensures rapid debugging of code, frequent releases and an almost risk-free development environment that never depends entirely on one or two great minds.

A lot of open source software is, in fact, free as in free beer. No cost. Download it, install it, use it, no charge.

Jarred Cinman is the product director of Cambrient

In the CMS world, a product like Mambo - one of the leading open source CMS products available - has several core development teams, and countless developers of modules and extensions. Since much of the development is rather informal, most open source products rely on public discussion forums for transfer of information. A glance at the Mambo forums shows just how active that community is.

And of course this discussion-driven, community-oriented model is part of the appeal. Hundreds - maybe even thousands - of available, interested participants able to answer queries, give advice and participate in the solving of problems faced by users, developers and administrators.

Which brings us to the issue of licensing. A lot of open source software is, in fact, free as in free beer. No cost. Download it, install it, use it, no charge. The most famous of all open source software, Linux, is free in exactly this way. Well, mostly.

CMS products like Mambo are essentially free. PHP Nuke is licensed for practically nothing - a few hundred dollars. And so the list goes on. You want free CMS you can get it, no problem.

The weaknesses of Floss are, of course, the strengths of commercial or proprietary software.

For starters, it`s all well and good to say the code is open source, but open is not the same as easily modifiable. Any piece of software by its nature is complex and displays the idiosyncrasies of the person or people who wrote it. If Microsoft decided to release the source code of Windows tomorrow, it would be a lot less useful than people think. Yes, virus writers would find ways to exploit it and eventually someone would be able to build on it - but at first, and for a very long time, it would be 40 million odd lines of useless text.

For this reason, a company running an open source CMS to run its Web site instead of using a commercial product is not necessarily able to forget about the risk of lock-in. Worse even than being locked into a corporation, it is locked into a collection of developers scattered here, there and everywhere. The company can`t go visit their CEO and insist on more attention. These developers work for no one, least of all some other business making money off their code.

Which brings us to the South African problem: installed base. We in this country have a pretty low adoption of open source software. At the level of operating systems and networking, Linux and related technologies have a good footprint here. One level up, at the application server level, software such as Apache Tomcat or JBoss also has reasonably good penetration. But at the actual application level, few companies have made the leap. Probably because of problem number one above.

This means that choosing an open source CMS leaves a business rather lonely. There may be many international users to connect with, but despite the ideal of a seamless global village, we all know it`s not that simple. Sometimes you just need to call in a physical, walking, talking, breathing expert.

Business is conservative and risk-averse, particularly in the IT area where it has burned a lot of money in the past. Despite all the gains in open source in recent years, it`s still a hard sell unless it`s plumbing software. When it comes to stuff that users see, use, configure and work with, commercial products still offer the kind of comfort, support and usability that businesses are looking for.

This will clearly not be the case forever. But it relies on two big changes: a strong installed base, among large, credible businesses; and, by extension, a strong developer community which can form the basis of contractible suppliers who are available to take responsibility for the rolled out solution.

For one obvious reason, proprietary software will not buckle under the weight of open source, especially not at the application level: people are willing to pay big money for applications because they want what that money can buy them. And that is called leverage. And that`s where open source still needs to figure out an angle.

* Disclosure: Jarred Cinman is the product director of Cambrient, which builds and sells a proprietary CMS utilising a number of open source components.