After his presentation at SDForum's Third Ruby conference this afternoon, I talked with Anant Jhingran (pictured) about his role as CTO of IBM's Information Management division. Our conversation also touched upon the fact that he's using a silver Macbook Pro inside the enterprise that invented the black colour laptop; about his thoughts on enterprise search and Google; about MySQL competition; merging IBM's 2 databases (Informix and DB2) and finally on using Ruby in the enterprise.

First, let's start with what Jhingran had to say about being VP at IBM and using a Mac!

"Mac's are actually quite popular in IBM. But of course... Thinkpads dominate, but my estimate will be that we may have at least a few thousand Macs within IBM. And many of us would deal with the Web 2.0 community and others... In addition, of course, having a Mac doesn't hurt talking to the Ruby on Rails community and everything else", said Jhingran. Wow!

On enterprise search, Jhingran points out that it is not as simple as "external" search a-la Google. I also over heard at the event saying that even Google even has problems finding information inside its own walls. Although, it was a surprised to me, it was not to Jhingran.

"Search is going to become a capability that every piece of product and solution has. So where enterprise search is a horizontal technology, today nobody thinks about building an application without a database... People using the application do not think about the database. So tomorrow, the same thing will happen in search... Search is an ingredient technology of every application".

MySQL in the enterprise: an anecdote!

"A lot of the growth of MySQL is coming from projects that grew in the open. So while you get a lot of interesting anecdotes and factoids about MySQL in the enterprise, the bulk of the downloads and the revenue are not coming from the enterprise... Data and databases are one of the most precious commodity that an enterprise has. Our customers are very tough on us... For them, going to something that is entirely free is not something that they care about. Because they bet their business around the database. And for anything that you bet your business on, the cost to run the database is hardly relevant. FInally, in databases it turns out that the price of acquiring software is a very very small fraction of the cost of running it".

Software as a Service: applications like Salesforce.com, yes... but NEVER databases or anything else! And this is from an expert in database systems, with a Ph.D. in 1990 in that field from the University of California at Berkeley. But can this be true? Your comments please!

"The only thing that really works as a service is large classes of applications or to a degree outsourcing of business processes... the data has to move out of the enterprise out to the cloud. And it's not clear the benefit of this thing running on the cloud that enterprises want to take their most precious commodity and move it out... That said, it's entirely possible and we believe it's the case, that the kinds of capabilities of being able to flexibly grow your applications or your database are very much the kind of capabilities that our clients want inside their enterprise".

Merging Informix and DB2 is not what the clients want. What they want is to lower their cost of managing the data across these 2 databases.

"You'll see us talking about Data Studio and that's very much about a common tooling and administration across IBM and non-IBM databases".

Finally, Ruby and Ruby on Rails are enterprise-ready technologies. But enterprise Ruby applications are so far only deployed by Tech-savvy companies like IBM.

"A lot of our clients are not full off technologists. For them technology is a cost of doing business... We have seen a lot of small projects on Ruby on Rails, but I have not seen yet outside of IBM, it doesn't mean it does not exist, enterprise-class applications built on Ruby on Rails. What I was trying to say is that there are no reasons why we can't get there but they are many issues of skills, organisation, legacy data, etc".