I think they've actually got the long term covered. Everyone in The Enterprise (which is all Oracle cares about anyway) is already using Java. Those people want to buy official looking software with support contracts from someone.
They're going to be doing so for decades to come, either from Oracle itself, or someone who's paying Oracle truckloads of licensing fees to be allowed to publish their own Java implementation.
The only thing that could upset their subscription to money would be some inconvenient little company that decides to make a clean-room Java implementation that could target the same customers. This is what Google has done.
If they can win this in court or force a settlement that pretty much guarantees that anyone thinking of building that business model won't even bother to get off the ground, and Oracle can keep printing money.
It's funny that you should mention COBOL. Java turning into the next COBOL would be awesome for Oracle. IBM still makes billions selling systems to run COBOL[1], that's more than Sun ever made off Java.
Most companies buying mainframes, buy mainframes to run Linux and Java apps on it. The COBOL+mainframe market is not growing (very much). (source: talking with an IBM research director in Böblingen, Germany responsible for mainframe development)
Mainframes promise turnkey-reliability, where the IBM support guy comes in before you notice anything going wrong. The point is that "mainframe" is a strong brand and IBM has the monopoly on it. There are solutions magnitudes cheaper and just as reliable as a mainframe.
Comments
I think they've actually got the long term covered. Everyone in The Enterprise (which is all Oracle cares about anyway) is already using Java. Those people want to buy official looking software with support contracts from someone.
They're going to be doing so for decades to come, either from Oracle itself, or someone who's paying Oracle truckloads of licensing fees to be allowed to publish their own Java implementation.
The only thing that could upset their subscription to money would be some inconvenient little company that decides to make a clean-room Java implementation that could target the same customers. This is what Google has done.
If they can win this in court or force a settlement that pretty much guarantees that anyone thinking of building that business model won't even bother to get off the ground, and Oracle can keep printing money.
It's funny that you should mention COBOL. Java turning into the next COBOL would be awesome for Oracle. IBM still makes billions selling systems to run COBOL[1], that's more than Sun ever made off Java.
1. http://news.ycombinator.com/item?id=1567595
Most companies buying mainframes, buy mainframes to run Linux and Java apps on it. The COBOL+mainframe market is not growing (very much). (source: talking with an IBM research director in Böblingen, Germany responsible for mainframe development)
Mainframes promise turnkey-reliability, where the IBM support guy comes in before you notice anything going wrong. The point is that "mainframe" is a strong brand and IBM has the monopoly on it. There are solutions magnitudes cheaper and just as reliable as a mainframe.