Open source to shape cloud computing, but not dominate it

I don't understand the author's point. You can't compare a library/system like hadoop to an implementation of that system. Apache Hadoop isn't a business.

That article irks me as it makes no sense. First, there in fact is very little room for commercial software in the cloud: the clouds are built using either in-house developed tools, open source software, or combinations thereof.

Second, what people run on top of cloud platforms is almost always open source software: in fact, it would be hard to see open source distributed computing projects such as Hadoop and Voldemort gaining much traction without "the clouds".

Third, there's a whole business model for open source software: the software itself can be released free (as both in speech and in beer), while the customer would be charged for the service of running the software on vendor's premises (the vendor nets a profit, while the customer still as they don't have to build out the hardware infrastructure/perform systems administration/customize the software themselves).

Three misc thoughts:

1) To me there seems to be a long distance between "no place for open source" vs "open source will shape but not win..."

2) While it may be so that "the vendors of today are the same vendors of 10 years ago"; that does not hold true when sliding this 10yr window back in time. I certainly don't think its fair to say this will needs hold true in future either. Maybe the best that can be said is that "vendors appear to have time constant of around 10yrs"

3) What's with all this talk of "winning" anyway? Winning implies there is an end / finish line - tech is so transitory I don't know if its possible to say anyone has ever "won" anything, maybe "dominated" for some period of time, but its never really over, is it?

yeah, tell amazon to remove all their xen-based services.