Tuxedo is Oracle’s application environment for the non-Java languages. Like most “legacy” transaction servers, Tuxedo provides major large enterprise functionality to the programming languages prior to Java. Tuxedo had focused on C/C++ and COBOL until now. Among a couple of innovations, the most exciting news in the just-announced Oracle Tuxedo 11g release is the support for Ruby and Python. This pushes these newer languages immediately up the enterprise performance and reliability scale, making them comparable to COBOL, ABAP, and NATURAL.
The huge challenge for Oracle after this move will be to get access to the Ruby and Python developer communities. Most of them are looking more at open source runtime environments than at heavyweight enterprise transaction environments. However, this latest move by Oracle may resonate with these young open source natives, who’ve gone from university to their first job at banks, insurance companies, and other traditional mainframe shops. Ruby and Python on Tuxedo could be appropriate choices for those developers who want to move stuff off a mainframe but don’t want to get into COBOL on the new platform again.
The second announcement creates an even stronger impression that Oracle is attacking the mainframe vendor(s) in 2010. Oracle’s Re-hosting products — which is what Oracle calls its products that help you to move business logic off the mainframe — have been improved specifically to make the move away from CICS easier. Complemented with a batch application runtime, which provides IBM JES-like job management, the set of components that you require for migration out of the IBM mainframe world are becoming increasingly complete.
Despite all of the tools from Oracle and vendors like Micro Focus and Software AG that can help you migrate large, old business applications from mainframe hardware to distributed Unix environments, one major challenge can’t be neglected: A migration off the mainframe always involves manual steps and requires a deep understanding of the business logic. With this announcement, Oracle is entering into the fight for those talented workers who still understand mainframes (or can learn to do so). A pure product-based go-to-market strategy will not work.
I believe both Ruby and Python developers and mainframe experts should evaluate these new products. In particular, if you are working for a systems integrator, I am curious as to whether these scenarios will find a place in your service offering; please leave a comment on this blog and let me know.