Recap of JavaOne 2008

Liferay enjoyed a great JavaOne this year, our third corporate appearance at the ground zero of the Java development world.  The emphasis this year was on openness, community, and the individual, reflecting the transformation our industry and culture have gone through over the last decade. In IT, that's been reflected in a shift from proprietary solutions controlled by the few to a landscape where unknown upstarts out-maneuver established vendors. It was also a nod to the buzz-idea of the season, which is the empowerment of the individual (reflected in Time's Person of the Year being "You" and the never-ending stream of business models chasing ways to "monetize" on mass participation). 

Visitors to our booth came looking for core developers by name ("Is Jorge here?" "I'm looking for Neil Griffin, please"). A by-product of the phenomenon of open source community is that many of us have never met each other in person. Conferences, especially the developer-focused JavaOne, are a chance to make our connections more personal—and of course, to talk about new ideas and features for Liferay. 

Several of our visitors came from Sun Microsystems, with whom we made an announcement on Wednesday morning of the conference. Their portal team has known about our relationship for some time, but now that it's "official," folks from many other Sun departments and product groups wanted to see what the buzz was all about. Many were of course excited to discover Liferay's capabilities and consider new possibilities for how to leverage Liferay and the related Sun products in their own work.

The announcement was taken positively by customers and community members as well. It was important to us for people to understand that this initiative was an organic growth of our extant open source community and not simply a marketing-driven, top-down initiative without real value. Analysts and reporters we spoke to about the announcement were surprised how exclusively tech-driven the relationship was, but I believe that's what makes it innovative. We are enabling real open source collaboration among several communities, bringing several voices to the table, and working together to build solutions that enterprises and individuals can use.  

I also had several engaging conversations with various business counterparts at companies like IceSoft, Terracotta, Intalio, and SpringSource. The common denominator from my conversations seems to be that open source companies genuinely care about delivering value and service to their community members, whether they are developers, end-users, corporations, or not-for-profits. We are all at a crossroads between proprietary and open solutions, "old school" and "new school" business models, and of course, profit vs. service motivations.  The trick is to build sustainable businesses without betraying the community. 

Finally, I'm very excited about Liferay's new developer network, in beta today on this site. We want to bring people out of the woodwork and recognize the contributions our community makes everyday to Liferay Portal. We also expect to see increased participation from this, as people see the benefits of raising their profile in the Liferay community. 

Blogs
Hi Bryan,

is the developer network code available in the latest release?

cheers,
Brian
Hi Brian,

Sure, take a look at http://lportal.svn.sourceforge.net/svnroot/lportal/plugins/trunk under the package wol-portlet.
Thanks Bryan,

Brilliant stuff...so the profile page and the friends page are all in the same portlet...but it seems like the blogs are still separate. The menu bar at the top I presume just has some friendly urls configured in the main web.xml file of the site.

Is there any proposed release date for all this cool new stuff?

cheers,
Brian
Hi Brian,

Right now everything is obviously geared toward the Liferay project, but the services and back end will be usable in other scenarios. I don't have a release date yet. How are you using Liferay, by the way? And have you created your profile yet on Liferay.com?
Hi Bryan,

Interesting stuff. I've been using liferay for a good few years now in two companies and it's really great the direction it seems to be going in at this stage. I've been the main technical lead on using liferay in a revenue assurance product in the telecoms area. Very successful product sold into all the major telecoms companies worldwide. Liferay was used as the display tier to aggregate data, charts, alarms etc.

I'm also building a business focused site where i want to use some of the new social network capabilities.

I love the direction you guys are going in...I've used jquery before and love it so it should make the frontend slicker and slicker.

My biggest single frustration has always been the tight coupling of the portlets with the portal however. It's come up on the forums before I know and is on the longer term list of things to do I think, but it would be great to actually be able to have a liferay bundle that is totally minimal with very few (if any!) portlets.

But broadly speaking...congratulations and as a long term user I think you guys are really going in the right direction.
Hi Brian,

In fact the discussion on loose coupling came up again this weekend (a few of us were in Vancouver to celebrate Alex Chow's wedding, and of course Liferay was one of the main topics of conversation). The pro side is that a modular architecture is more maintainable in the long run and you can strip the portal down to its bare essentials. The con is that there is slight overhead in putting everything through a common messaging system or what other solution you make to loosely couple things.

We didn't come to a conclusion but we were thinking that there are core services (including the CMS, collaboration, and social stuff) that should stay in the kernel and other portlets stripped out.

What are your thoughts?

And do you happen to work at the company that built the sites for the owner of the former "pimple of London"? emoticon
Hi Brian,

IMHO the pros outweigh the cons for the modular approach. I think one of the biggest difficulties for new users of liferay is if you have to dig into the code, it's a pretty huge codebase. Having a core kernel that's small and fast I think is the way to go...but absolutely, I see your problem about wanting to keep some services. CMS, Login and basic user stuff is probably a good idea to keep...but even still i'm not sure. What I'm sure of is that there are people out there who customize pretty much every aspect of the portal...I know we did...and making that as easy as possible is just going to keep the community growing. Definitely the ability to customize the user account is going to come up again and again. I love the idea of the social side of the account...but i don't know if all projects would need it, so is it a kernel feature.

It's a difficult problem i suppose, but long term, a simple clean kernel would be beneficial in loads of ways i think. I'd love to see a liferay deployment where I could even start it up with zero portlets in no time and then just drop each portlet I needed in. The database scripts, the whole lot could be modularized per portlet.

The sites we built are used internally within the telecoms operators to check how much revenue they are loosing due to systems going down and not customers not being billed properly, so I don't think the public ever sees them. O2 are a customer, but i reckon you're thinking of their public sites probably, which we had nothing to do with :-)
Brian,

Thanks for sharing your thoughts. The best I can do right now is encourage you to participate in the community. Questions like these come up frequently and it'd be great to have your voice be part of the conversation.

Also, if your company is interested in becoming a Liferay partner, we're going to use the Developer Network to feature key partners that have demonstrated expertise in Liferay through participation and projects. Let me know if you need more information about that.

By the way that presentation is available now at the JavaOne landing page.