RIFE/Laszlo is based on RIFE/Jumpstart and geared towards easily getting started with the creation of Rich Internet Applications using OpenLaszlo.

This release offers the following:

• added out-of-container tests of the REST backend
• added asynchronous zip-code lookup example
• upgraded RIFE to version 1.3.1
• upgraded OpenLaszlo to version 3.1.1
• upgraded Jetty to version 5.1.4

The demo of the basic RIFE/Laszlo application can be seen at http://rifers.org/laszlo, and more information about how to use it can be found in the readme.

RIFE/Laszlo can be downloaded from our download section.

Beware when you're creating URL instances by providing the entire spec without separating it into protocol, host, and file.

If your URL spec happens to be in the following format:

file://C:/somepath/file

JDK 1.4 will try to lookup the C host name. This will of course fail, but by default it takes 5 minutes to do so on Windows. Imagine what happens when you have several of these URLs ...

To work around this issue, make sure to instantiate your file protocol URLs like this:

String filepath = "C:/somepath/file";
URL url = new URL("file", null, filepath);

This will prevent the JDK from getting confused.

You can find proof of this behavior in this screenshot of a debugging session that Stéphane Meslin-Weber performed to track down slow startup times in RIFE:

This issue seems to have been fixed in JDK 1.5.

Update:

It seems that I jumped to conclusions too quickly and that the correct URL format should have an additional slash:

file:///C:/somepath/file

The confusing part for us was that JDK 1.5 doesn't seem to try to lookup C as a hostname, which made the issue very hard to track down since none of the core RIFE devs use JDK 1.4. We use retroweaver to generate a compatible version.

read more...

The slides of the Quickie that I did at JavaPolis about web continuations are available online as PDF:

  Static PDF document (1.01MB)

JavaPolis turned out to be a great event, it's rare to meet that many interesting high-profile people and have such a large amount of networking opportunities. The general atmosphere was extremely loose, I even had Neil Gafter crawling underneath some tables to help me connect my laptop to the internet. I only realized it was him afterwards.

I spent the entire first day talking about RIFE. After the 2 1/2h university session in the morning about RIFE's web engine, constraints and RIFE/Crud, I immediately followed up with a quickie that focused on web continuations. This generated a lot of interest and I spent most of my afternoon doing private demonstrations, explaining various aspects in more detail.

Since so many had already been said about RIFE during the day, I expected the evening BOF to have almost no attendants. I was amazed to see the room being almost fully packed. We had a nice interactive session and lots of people asked smart questions and tried to see how to fit RIFE into their current architecture or how it compared to their existing technology stack. I mostly gave some short explanations and then showed excerpts of a presentation or showed some code to make them see what exactly was going on. After the BOF itself, we still continued talking in the hallway for quite a while.

The rest of the week, I regularly ran into people that knew about our framework and that wanted some brief overview. I rehearsed the 'RIFE introduction paragraphs' pretty well now

All this culminated into Thursday evening when I recorded a technology video interview with Floyd Marinescu from TheServerSide fame. The results should be published fairly soon.

It great to see that many interest being built up. Exciting times are ahead as the Java community starts to realize the importance of The Long Tail and the need for small shops to be able to push out very functional solutions quickly in a maintainable fashion.