Tuesday, May 8, 2007
After JavaOne last year, I told myself I'd attend NetBeans day in 2007. A month or so ago I registered for JavaOne and signed up for NetBeans day (not a very smooth process, I actually had to call someone to register for NB days because I'd already registered for J1) anyway, just after that Sun started to promote CommunityOne Day. So here I am at the first CommunityOne day, the day before JavaOne starts. The reason I wanted to come to NetBeans day (now CommunityOne day) was because it is so small. James Gosling just walked by me and he wasn't mobbed :-) I attended a presentation earlier today where Jonathan Schwartz was speaking and he was standing right behind me before the session started. Where else could you talk to Jonathan Schwartz with only 100+ people in the room? The message today has been pretty clear, Sun is going to open source all of its SW and innovation is the key. Almost every talk I went to today emphasized innovation. The argument goes like this, if Sun isn't trying to control everything, then their engineers are free to think about and work on innovation, it's what will set Sun apart from the others. Jonathan described the market as a really big pie and that there is plenty of room for many players. He also described how Sun will make money with this strategy, pull-through business by selling support, services and hardware. I think for me personally, I'm really excited about what Sun is doing. By making everything open source and freely available, means I can use the SW at home and on my own personal projects and in turn I can use it at work. I can see CommunityOne day being really big in a year or three so if you get a chance, I'd suggest trying to attend next year, before it's as big as JavaOne itself. For me, being able to talk to Jonathan or stop James in the hall is reason enough to come to CommunityOne. The content is somewhat different from J1. Instead of only being about Java (you could argue J1 is also going this way) you get exposure to everything Sun is doing in the open source space.
Monday, May 7, 2007
I don't have anything planned till noon, so I thought I'd check out the "GlassFish: Getting Started and What's New in GlassFish V2" talk. Eduardo Pelegri-Llopart of Sun and the GlassFish Team (Karen) are giving an overview of GF V2. I got to see a preview of GF V2 last year, so I'm hoping to see some more cool things this time around. Karen started off, welcoming everyone to the first ever, "GlassFish Day". Since the are looking for user feedback, they've offered everyone who fills out a survey either a T-Shirt, memory stick (with GF v2 preloaded), or a sticker. Karen told us that instead of meeting Jonathan for lunch, he'll join us in this room for a Q&A for about 30 minutes, then it's lunch. Next up is Eduardo to talk about what's new in GF. In order to best answer everyone's questions, the leads for each technology in GF are here to answer questions, after the presentation. For more info on GF, check out http://blogs.sun.com/theaquarium, http://blogs.sun.com/stories/, https://glassfish.dev.java.net/. He is now showing a time-line of Project GlassFish, showing the launch of GF June '05, v1 final JavaOne '06, V2 beta 2 J1 '07. What is Gf? It's the Java EE 5 RI, it's enterprise quality, bases for Sun's AS 9.0 & 9.1. What does RI mean? When you create a JSR, the EG also needs to produce either a proof-of-concept or implementation to prove the technology can be created. Earlier RI of other JSR's were proof-of-concept, but in the case of GF, this RI is actually production ready, not a prototype. It has a dual license, either CDDL or GPL v2, you as the user of the framework get to decide the type of license you want. GF is also a community, like Apache, but not nearly as large - Eduardo's words, not mine :-) In addition, GF is also a collection of other projects, like jMaki, Phobos, BlogApps, JAX-RS, etc. An important note, all GF related discussions, etc. are kept outside of Sun, except for a few instances, GF is it's own project run outside of Sun, outside of the firewall. Next slide, "A Taste of Java EE 5", for those who have not seen it yet... first we are shown an overview of J2EE 1.4, its powerful, a standard, but too difficult to get started, ever simple apps need boring boilerplate, in Java EE 5 we eliminate the boiler plate, and make simple things simple. Java EE 5.0 = (J2EE 1.4).next, the theme of Java EE 5 is "Ease of Development". It's embraces a POJO based programming model. We are shown the standard, here is how it was and here is what it's like now slides for JAX-RPC, of course the Java EE 5 version is about 5 lines of code with no deployment descriptor (compared to the 30 lines of code and 50 lines of XML for the J2EE 1.4 version). Bill Shannon said, "Not your father's J2EE". That's it about Java EE 5, just a taste. Now on to the GF community, the principles of the community are:
- industry-leading technology, (no secret sauce)
- want adoption of code base (no secrete sauce for code base
- this is all of it (editors note: aka not like JBoss)
- community centered
- continuous improvements
- open standards - JCP
- OSI License
- clear copyright (SCA).
Very cool, Jonathan Schwartz was right behind my seat and I didn't even see him. Steve Olson just joined me, so we are going to listen to Rich and Jonathan together. Jonathan is talking about the the minimal foot print for GlassFish and how it can now run on a phone (GF v3). Jonathan and Rich are doing everything they can to remove all obstructions. There was an old saying at Sun, "Innovation happens elsewhere" and they want to foster innovation. They want to breakdown barriers and foster innovation. Investors keep asking Jonathan why Sun keeps giving away their ideas and an engineer @ Sun said they are not afraid of running out of ideas. Question about convergence, OSGi vs Java Modules, etc. Rich could not really answer the question but he did say, if we use late binding for the implementation, we can then decide on the right solution at deployment and not have to make a decision @ development. Editorial comment: I'm not sure how viable this suggestion really is. Companies that handle really large volumes of data cannot wait till later to chose the technology but I'm many applications will be able to leverage this type of architecture or development style. Jonathan talked about the choice to Open Source their software and how he sees the value Sun provides. If you are self-sufficient (as most are) you shouldn't have to pay for SW, however, if it is important for you to have immediate support when something goes down or you don't want to support it yourself, Sun can provide support and that is where the value is. The more open the more folks can innovate, so Sun is really moving towards the real revenue. Sun isn't Costco, they are not interested in bulk, they are interested in innovation. In a sense, they don't have a plan, that is to day you cannot plan innovation, it happens but you can plan for and create an environment that fosters innovation.