Open Source and software rental
Software pricing has always presented a conundrum both for vendors and for the customer. In the pre-internet days software pricing on servers was, roughly, proportional to the cost of the hardware. In the days of mainframes and minicomputers this seemed rational, but as the servers morphed into anything from a lonely Intel box to a high-end cluster it began to get complicated and it was quite possible that, with expensive software such as the big name database products, the structure of software pricing could limit or determine hardware choice.
This has now evolved to the point where, for example, Oracle's database technology is specifically designed to run on inexpensive servers (via Oracle's RAC technology), with the idea that if the hardware costs are low, the overall ticket price is lower. Such developments would be fine, if all vendors stayed completely in step in terms of pricing policy. But that doesn't happen. So the corporate technology buyers have a mess on their hands in trying to build rational and reasonably priced networks. The mess is made greater by the reality of desktop computing costs, which also bite out a huge chunk of budget for almost every company and complicate the issue still further. After all, what is the least expensive way to build a reliable network?
The truth is that once any vendor gets into a monopoly position, or even a position of significant leverage, they tend to maintain the price of the software. And by the way, I mean any vendor - Microsoft may be the vendor criticized most for this, but all vendors have a tendency to maximize the profits of success. In theory if their sales are climbing then their unit costs are trending down, but the consumer rarely sees the costs fall. This is one of the reasons that Open Source is important and why corporations really should have a policy of using it to some degree - and I'm not just talking Linux here, I mean Ingres and MySQL, I mean Apache, I mean Zope and maybe even a little desktop Linux mixed with Open Office. If it does nothing else, it may help to keep the vendors of proprietary software honest - and incidentally, a good deal of Open Source is quality software.
There is, however, a market force in play which may eventually alter the way that Open Source is perceived and deployed. This is the gradual move to software rental. When the ASP trend first occurred it was accompanied by a good deal of hype, and loudly proclaimed as "the wave of the future". Then after about a year it was declared by many to be a trend that failed - ignoring the fact that IT trends always occur twice, first as hype backed by immature product and the second time as a stable offering backed by a genuine value proposition. A trend is dead only if the second wave fails.
However, the second wave of the ASP trend has been reasonably successful, with the flagship vendors being Salesforce.com and Oracle. (Oracle Online is a significant success, in case you didn't know). This is moving the market gradually towards the idea of software rental as the natural way to price software. It is not a simple transformation because there are no simple and generally agreed metrics for measuring software usage, but it is a driving trend nevertheless.
An interesting aspect of software rental is that the price is just as notional as it is for any other kind of software license. However, if you are renting software that is hosted elsewhere, then you probably have no interest in what software makes up the technology stack - what you are buying is the application. This allows an ASP provider to circumvent some of the commercial difficulties of the Open Source license (the GPL), because the ASP charges for the service provided and not for the software. Of course this probably acts as an incentive for them to use Open Source, but it doesn't constrain their ability to make healthy margins - and the GPL does seem to do that in the traditional software licensing model.
So, as software moves towards a rental model, it is possible that the disruptive market impact of Open Source GPL will diminish. At the same time, Open Source products will take their place as viable products for specific areas of functionality. In other words the schism between proprietary software and Open Source software will be far less visible.
Copyright © 2004, IT-Analysis.com
Sponsored: Benefits from the lessons learned in HPC