Castles in the Clouds

Software-as-a-Service continues to gain momentum with more companies offering cloud computing solutions and even Development-as-a-Service.

Gartner seems to think SaaS will be adopted, apparently quite readily, if 15% of large companies will have SaaS initiatives by 2010. There is still COBOL and IDMS in this world. There are many factors that I feel are being glommed over. A couple of high level points:

Accountability--How anxious are companies to say, 'would you please hold onto all of our proprietary data for us?' For downtime they get to point a finger, but they still have to deal with the angry customers.

Security--Isn't this the ultimate all-your-eggs-in-one-basket situation? Again, companies are anxious to be beta testers?

Customization--Companies run differently... Look at how many canned applications you can find that do the same thing.

I can see smaller companies adopting cloud technologies more readily than large companies. They have budget and resource constraints. But I think it will take time before larger companies put stock into these offerings--hopefully not as long as grid computing.

I am just a DBA; I recognize that I am not business-oriented, as a misanthrope I do not possess the management mindset. Yet I see some very serious setbacks to SaaS adoption. Part of this is skepticism from seeing the industry and IT shops cycle through centralization and decentralization trends.

Let's hope that there is some substance to this hype.