Hammering at cost with open source

Irregular Bob Warfield riffs on Salesforce.com's decision to ditch Sun/Solaris in favor of Dell/Linux. In doing so, Bob makes the point that:

...for a SaaS company, the cost of service delivery is an absolutely critical factor.  Once you have software that runs well and scales horizontally on cheap commodity hardware, you’ve created a huge cost advantage for yourself.  As we speak, the cost to deliver service for the various public SaaS companies is all over the map, but Salesforce has always had one of the lowest if not the lowest cost on the map.

As I understand it, the benchmark developers like to keep an eye on is Google where the bare bones cost of delivery is reckoned to be in the 8-10% of revenue range. Bob then goes on to speculate whether Salesforce.com's  next move might be to ditch its Oracle database in favor of something like MySQL. I'll stick my neck out here and say it won't happen. Marc Benioff, Salesforce.com's CEO has Oracle blood running through his veins. If he is to have a shot at selling the company at some stage, then it would be suicidal to make that decision. Unless of course it was a sale to Google, something I can't see.

Vinnie Mirchandani raises a collateral point:

The question of course, is whether SaaS vendors benefit from at least the perception that Oracle is more "bullet proof" or do SaaS customers just want results (high uptime, performance etc) and don't really care what the underlying  technology is - especially if the economics are more attractive?

In our Irregular discussions, there was a view that brand matters and therefore the default selection would most likely be Oracle. But then I reflected upon a project in which I am currently engaged. We are eschewing any commercially licensed software as far as possible. Part of the reason is cost but equally, license discussions are a huge distraction, especially when you are juggling with components that come from a number of different sources. Even so, making the right choices, even in the open source world isn't as straightforward as you might imagine.

You have for example to think about the license implications of developing across software components where the rights of the licensee might impact what you do. Then there are the issues of how licenses are viewed across different geographies. It's enough to give any sane person sleepless nights and an ongoing migraine. And that's before the legal boys sink their teeth into clarifying the problems. It means the potential for a significant upfront investment to ensure that we don't trip up and find ourselves landed with an unwelcome licensing bill. I'm not sure which is worse.