There was a good debate last week at O'Reilly Radar about the implications of open source software (OSS) in the movement to Software as a Service (SaaS). Of course it included all the "we vs. them." "we invented Web 2.0," "OSS adoption is massive," "we invented SaaS," and "we don't get no respect" that these inward-looking OSS blogs love.
But it showed why the IT industry needs to nail down some definitions and concepts before discussions like these begin in earnest. Political blogs are easy to follow because their followers have standard definitions for right/left, republican/democrat, blue/red, and so forth (if a third party were to ever take off in the U.S. it would put the political blogs out of business). Auto buffs have good online and cruise meetings because they all agree on the auto industry model. But IT industry conversations, especially OSS conversations, are more like religous wars with everyone talking past each other.
The O'Reilly blog post and subsequent comments were centered around the meaning of conveyance in the OSS licenses (thankfully it means about the same as it means in any contract). But the string wandered around to--among other things--predict the demise of Google. The reason Google might self-destruct in this scenario is that--of course--it is not open, despite its major use of OSS. But Google does not deliver SaaS in a way that can be compared so some really imaginative thinking about the future of IT falls apart, and splinters further.
So when reading any of these blogs, move up a level and ask what the buzzwords mean.
First, for example, OSS, SaaS, Web 2.0 and utility computing platform are in fact buzzwords, here today and gone tomorrow. The concepts underlying the words have been around for years and someone just gave them new names (with the exception that utilitiy computing was the name MIT originally gave it even before Richard Stallman started hanging out in its labs). OSS is user-group software taken to the next level. SaaS is ADP payroll processing taken to the next level. Web 2.0 is--of course--Web 1.0 taken to the next level.
Second, the concepts are neither competitive nor causative. One does not have to choose between OSS and SaaS nor did OSS enable SaaS (although I believe OSS will be a big enabler of SaaS predominance before the end of the next decade). Web 2.0 was happening without OSS; I forget exactly what Microsoft called it in 1995 (remote scripting something--see this feature article on our website). SaaS was happening without OSS; we called it application service provision (and timesharing service bureaus before that).