We sent a vulture to IBM's new developer conference to find an answer to the burning question: Why Big Blue?

Regarding the disparate figures around cloud revenue...

word is software contracts have been changed to include "cloud" in them, so no matter where you are running yer WAS or db2 it's considered cloudy income.

IBM Droid» "Why IBM?" mused Duimovich. "We are driving a pretty aggressive optimization and ease of use story for cloud Java."

So, he drives stories. Hmmm.

Hmm, the Java VM has a bit of a problem with the Cloud.

JVM eats memory. basically, you only half compile a Java binary, leaving the VM for JITing.

That was OK when the JVM runs on a physical .

Its a huge problem on a VM - VM memory speed is shit.

Missing a disclaimer?

"In the cloud, he explained, that translates to workloads that cost half as much to run under memory-based pricing as a standard VM."

Shouldn't this include a disclaimer along the lines of "comparing a poorly configured Java application on an overpriced IBM cloud service to a Java application on Z where the memory pricing is a separate line item. This may not match any real world conditions'.

Unless IBM have examples that can't be right torn apart in seconds...

Ignoring the rhetorical nature of the headline, I always thought they were called Big Blue (aka It's Being Mended) because their kit was usually blue, grey and cream and it was a stonkingly large and powerful company.

You need to think in 1970 terms. No one got fired for buying IBM.

Why would you pick anyone who doesn't have a JVM team?

I would pick them because they don't have a JVM team. It means they're less likely to use Java as a solution.

YMMV

They've Got Stuff

They've got some technologies that might be valuable.

Their mainframes, while priced high, and aimed at legacy customers, are far more secure than Windows servers, and, in fact, I think that z/OS has an advantage over Linux and even BSD in this department. But if they made that tech available at competitive prices - would it sell enough to pay for its development? Or would it just throw away money by cannibalizing their existing customer base?

So I can understand why they're paralyzed... but the world does need genuinely secure servers, and IBM is uniquely positioned to meet that need.

But if they try to meet that need with what they have, and it fails, they could lose a big revenue stream for no return. I wish I knew what to recommend to them.

IBM's new slogan: "No one ever got fired for trashing IBM offerings while refusing to be named in the press."