Response to John Naughon@The
Observer re: Bill Gates
jumbo-size problem

On Sunday May 7, 2006, John Naughon of The Observer wrote a piece titled A problem too jumbo-sized for Bill Gates to solve?. In general, this was quite positive towards open source software, however there were a few points where John didn't fully appreciate or understand the the key differences between the Microsoft approach towards developing software, and the open source approach. What follows is an email I sent to John clarifying a few of these differences.

To: John Naughon@The Observer

Subject: re: A problem too jumbo-sized for Bill Gates to solve?

Hi John.

A few suggestions for you to ponder:

1. You state:

Whatever other charges might be levelled against Microsoft, technical incompetence isn't one. If the folks at Redmond can't do it, maybe it just can't be done.

Do you have any real proof or basis for this concept? So many of Microsoft's design decisions have been shown to be woefully inadequate. For instance, ensuring that Outlook and Word automatically execute macros attached to documents. No other firm was foolish enough to embed such functionality in widely used products, as the potential for misuse and damage is immense, as we've seen with the thousands of Office and Outlook macro viruses and the havoc they've wreaked over a 12 year period.

Sure, Microsoft has talent. But that talent is focused on milking as much as it can from the monopoly rents it extracts from hundreds of millions of customers, not in building better tech. Microsoft is a living example of capitalism in action. Their aim isn't to build better software, but to make more money. They two are not synonymous, Not even close.

By this I mean that no matter how stupendous the new technology Microsoft creates is, it will almost certainly not generate the same profit returns as its current monopolies (Windows and Office) and whatever other spaces those monopolies earn them new monopolies in - for example using IE to crowbar as many users onto MSN Search rather than them selecting to use Google.

This shifts the advertising revenue from Google, which had won the search market fair and square through better tech, to Microsoft, which can win it by the fiat of owning 85% of the browser market, which in turn was delivered to it through its 90% ownership of the desktop OS market.

2. Complexity.

Microsoft, whether through design or stupidity, built into its product lines incredible levels of often unnecessary complexity. The open source space cannot match this level of complexity because it cannot (generally) direct thousands of programmers to power their way through complex molasses of code. Microsoft, through stock options and wages, can.

Therefore, the open source methodology has had to develop more efficient means of building viable applications and systems, which often translates into structuring the design so that it is inherently simpler and more modular. This reduces the 'barrier to entry' for other interested parties to join you in your efforts to construct viable software. Creating too high an edifice for your code means greatly reducing the likelihood that others have the time or patience to assist you.

Let me give you an analogy. Microsoft has the money to build monster trucks, with rocket packs. These vehicles can simply power through the problem of vehicular wind resistance at speed. Open source generally cannot afford this, so we end up designing sleeker vehicles which slice through the problem of wind resistance. The advantage of the latter approach is that as speeds increase, our approach offers the better chance of long term success.

Hope this helps in your understanding of the difference between the proprietary cathedral model that Microsoft uses, and the touted bazaar model that open source developers use.

Cheers,

Con Zymaris.