"Our 'Live' efforts will have a natural side effect of increasing the use of genuine software," Ozzie said. What on earth is Ray talking about? He is now being processed in to the monster of benign attempts to change the software landscape in his current annointment to software CTO at Microsoft. I typically stay closer to the heart of enterprise software, but this quote struck me this morning as I was waking up. Are we to assume that web services are non-genuine? That, perhaps, businesses should not make serious investments in the cloud, for fear that:
1. It will be in-operable with current software infrastructure.
or
2. It will be merely a conduit toward more sophisticated rich-app implementations.
There are two points that may not be so obvious, but I would like to call attention to:
1. When did the advertising agencies of the Internet co-opt the term web services?
and
2. When are businesses going to actually implement an SOA infrastructure?
To me, it seems that we have been building out the intranet capabilities for 10 years, and dabbling in supply chain effectiveness through small attempts at an extranet infrastructure, but making little progress to an integrated architecture. We have had many launches, and claims of interoperability through XML, primarliy. But not much is going on with JAX and .Net as far as I can tell. My release of SunONE AS7 and JAX-RPC seems to be the only legitimate effort to marry the app server development experience with JAX technologies, not counting the dutiful support of the tools vendors. We need more, and it needs to come from Glassfish and JBoss if we are going to go anywhere with a JAX-based web services infrastructure for SOA.
In conjunction with the JAX support, there needs to be a burgeoning initiative around extranet development of integrated networks. I thought Loudcloud was trying to do this before selling out to BMC (i think). Or were they just trying to do N1 things. Either way, let's make some headway in moving away from LDAP-only architectures, and toward some federation among SSO platforms. Today, I am calling on Sun to open up the java.net homebase for a new release of an open source, identity-enabled Access manager in the form of Diamelle's proposed openIAM project. This will move us away from an intranet viewpoint, and toward the valued extranet services that the cloud calls for, and that vendors will ultimately need to support.
In short, Sun once again holds the cards, as JBoss scrambles to put together its Portal and ESB strategy, not to mention the release of JBoss AS5 with JEE5 support. Glassfish is leading the way to integration and open source excellence. That is a difficult pair to match, and will ultimately be valuable for Sun consultants and partners, which will in turn sell Sun hardware. By making the case for an extended enterprise, Sun will open up the market for Sun-enabled SOA infrastructures, just as I prescribed for Sun's software business to Mr. Bauhaus at the end of July '03. Right before being pushed out the door. Make my bet pay off, and embrace the Diamelle proposition of an expansive worldview of Identity and Access Management (IAM), turn us on java.net, and allow us to kick-start the developer initiatives around JAX, JBI, and IAM. Ultimately, in re-shaping the software landscape, allies are needed, even against out-of-touch mega-vendors, such as Micrososft.
I don't know how else to make my case than to just repeat it: now is a unique time in Sun's history, and truly in the history of software - - you have a Java product that could become the market leader. Grab hold of this opportunity, let me help you, and this could be the beginning of a truly great run...