Oracle vs PHP - FUD and Ignorance in Journalism

LXer Feature: 10-Oct-2006 An article showed up in LXer's news queue that described how a business chose an Oracle solution over - get this - a PHP solution. Never mind that the article compares a database to a scripting language, but one can use Open Source PHP in conjunction with Oracle's non-free database system to write web applications. It just goes to show that we all need to know the difference between crap and shinola.

Line56.com wrote an interesting article about why vAudit chose an Oracle solution over a PHP solution. That's right, boys and girls! Oracle over PHP. There can only be two reasons why a case study-styled article would compare a non-libre database to a libre scripting language. The first reason is that the author is new to the technology industry and doesn't yet know his head from a hole in the ground. The other reason is that the author is intentionally misleading the audience.

Maybe I'm being mean or unfair. But ask yourself this one question. How does a technology news site get a story like this so wrong? Any technology manager worth their salt should at least know the difference between a database and a web scripting language. Not that all technology managers are worth their salt, and non-technology managers may not know the difference at all. Some executive reading the Line56 article may overrule the IT manager on the basis of just such an article. This is why I think trust is so important in journalism. Indeed, LXer's history is rooted in the issue of trust.

The Line56 article would have made sense if they had compared the Oracle solution to a comparable Linux, Apache, MySQL, and PHP (LAMP) solution. Instead, they simply said "PHP". Ironically, Oracle has a whole section of their web site devoted to using PHP with their database technology. Again, any half-awake IT manager will quickly recognize the gaffe. Unfortunately, the article seems geared more toward executives that may not catch on.

My guess is that the author got a tip from Oracle's PR folks, but didn't know what questions to ask. It could also be that the IT manager at vAudit did not give complete answers, in which case, the author should have pressed for more details or shot down the article, since it makes little sense as published. In any case, the article presents a very weak case for what's wrong with Open Source. Indeed, it's not what's wrong with Open Source, but what's wrong with the article that is the real issue in this case.

The Line56 article goes further to state that "service is non-existent or sporadic". Here is a far more likely scenario, without cross-examining the vAudit contact. The vAudit folks explored a few FOSS solutions, either canned or custom, by different vendors. Or they simply did not find a vendor that met their particular requirements. One wonders if the vAudit folks knew where to look for qualified help.

The simple fact is that literally hundreds of FOSS-oriented service companies exist in the United States alone, most of them focused on a variety of web-enabled database solutions. In fact, many of those companies have at least 8 years of experience, on average, offering precisely the kinds of services vAudit seems to need. So to say outright that service is "non-existent or sporadic" is either pure hype or pure ignorance, depending on the speaker's level of knowledge. Coming from a technology news organization, one is likely to think in terms of hype - not ignorance.

Let me be the first to say that I am not an expert in any particular field. However, before I write a story, I try to do enough research to understand what it is I am writing about. I have ditched stories because I did not feel comfortable with the few facts I had. Some writers call on "experts" to make exactly the kinds of statements that Line56 published. A journalist relying on such experts can be said to have relied on biased expertise, but a journalist making unsupported claims appears to have a poor comprehension of his subject, and risks credibility.

Look, I cannot fault vAudit if they honestly believe they have made the best business decision for their company. Maybe they did. However, the Line56 article makes broad sweeping statements with little or no basis in fact. The author's only point of reference, comparing Oracle to PHP, fails utterly. So, rather than provide substantive information, the Line56 article merely feeds readers a line of bull.

Executives should be wary of articles that are devoid of details. Be wary of overreaching statements. Be wary of comparing apples to oranges. Be wary of journalists who do not care enough to get the facts. Dan Rather's fall should teach all journalists an important lesson - it's all about trust. Rather failed to check the facts. So did this Line56 author. In saying that, I am implying that LXer editors and writers must continually strive to keep the trust of our readers. Trust is precious.