Views on the Industry

GlassFish

It's Friday at 4, and I cannot come to terms with initiating a new project at this moment, and I am waiting to speak with my manager so there is no way to bolt out of here, so here we go with a topic that I have been waiting to talk about for about seven years, about the time that I learned what an application server was. Right after the acquisitions of WebLogic and NetDynamics and a year before joining the Sun-Netscape Alliance. The momentous occassion is described here:
http://news.com.com/Sun+tries+sharing+Java+again+still+not+open+source/2100-7344_3-5754765.html

And is hosted Here:
https://glassfish.dev.java.net/

And I am usually pro-Marc, but this time he has it very, very wrong. He cannot continue to perpetuate the myth that Sun does not have a viable app. server, sooner or later, customers will read that for fear. His Blog response to Glass Fish is revealing:
http://jboss.org/jbossBlog/blog/mfleury/2005/06/06/SUN_open_sources_XYZ_who_cares.txt

He has it right that IBM has declared WebSphere irrelevant and is now turning their hopes to Geronimo, I can feel him on that. WebSphere has been technically dead for - - well, for always. But they have not put it to rest from a product standpoint until today. Now for all customers investing in WebSphere solutions it is clear that the investment is lost, unless there is a miracle port to Geronimo. But Marc is not correct in his analysis of the java.net site, and he is not correct to discredit Sun's Glass Fish efforts. That is a perfectly sustainable move for the Solaris house.

For sure, it will take time to integrate the efforts of java.net developers in to productized format, who knows if Sun will ever take that plunge, but the time has come to recognize that the current measurements for app server leadership do not truly represent who is the market leader. JBoss is #1, Geronimo is #2, and Glass Fish is #3. That is not to mean that you have to be open source to be the leader, but that is how it is working out currently. I would give BEA the #2 position if they are acquired by Oracle, and then it would be a dog fight between Geronimo and Glass Fish for relevance. That is what the market wants to see, an actual hand-hand warfare between Sun and IBM on something. They talk about warfare, but they don't compete on any product line. Competing on the app server level would be the battle that everyone has been waiting for. It should come as no surprise that I give the edge to Sun, but only if they take the following steps, which mitigate the natural advantage IBM has with Linux:
1) Build Clustra capabilities in to N1
2) Integrate AS, EE w/ a new Portal
3) Integrate AS, EE in to Solaris

Until those steps are taken, IBM has too much of an advantage with WebSphere installed base, Global Services penetration, and DB2 and Eclipse. I like their odds, as soon as Jonathan answers my blog. I think he has done a more than adequate first step with Glass Fish, but now is the time to take on IBM, not waiting until Geronimo is technically built up to a JBoss level with Portals, and Mapping, and Eclipse integration, and all that comes with it. Buy Plumtree, release Clustra with nodes for the network, and integrate EE in with Solaris. Then you have a viable selling platform. And a competitive advantage over IBM. Currently there is no advantage while getting killed by Dell, and treading water at the high-end. Go after the margins of high-end apps., and take the plunge with EE. Nothing would make Sun stronger and more viable than to directly compete with the WebSphere/Global Services installed base soon to be running Geronimo. The time is now. By Java One 2006, everything should be well under way. Otherwise the buffer that is the market downturn will be gone, and Sun's dream of a productized Java will be someone else's...

posted by douglas dooley @ 1:01 PM