The Apache Lenya development community is very proud to announce the 1.2.3 release of Apache Lenya. Lenya is an open-source content-management system written in Java, with a WYSIWYG editor, workflow, internationalization capabilities, search, and more.

It's built on a series of Apache offerings, such as Cocoon and Lucene.

Changes in 1.2.3

The full changelog is available at the Apache Lenya website. Some highlights are:

• Replaced Sitetree with XMLHTTP implementation for greater performance
• Added synchronization to sitetree to improve stability under load
• Added caching to default publication
• Moved to Ant 1.6.3 and Jetty 4.2.23
• Update link attributes on publish
• Various small improvements to the asset management screens
• Improved windows installer to make Lenya version more explicit and launch Lenya start page at the end of the install

The project sounds fantastic. It just a shame that the link to the project takes you to a page that seems to contain JavaScript that eats ~95% of my CPU when I try and access it.

Any reason why you guys are not using the standard Maven build to manage your project?

yeah, that js has to go, it's only (stupid) purpose is to change the font size..

I just tried it... It really sounded interesting.

So I used the Win installer... Installation (in another folder than the default one) went fine, the server started up fine and I had the home page displayed.

That when it started going wrong... First... There is no explanation as to what to do once it's installed.

I went to the administrative page (using the default usr/pwd displayed on the login page), I try to add a user and it fails with an exception saying that it was unable to find some file (with the new username in it).

I clicked on the "Site" tab... Again an exception saying it was unable to find some other file.

After 2-3 minutes, I removed the thing.

So here are some small improvment that I immediately think of:
- If you write an installer, at least ask for a new username/password at installation time (default username/password are a bad thing),
- If you write an installer, also ask for the port number on which the server should listen,
- Basic functionality such as adding user should work!
- Have an homepage that clearly document the steps required to do something useful with the product.

My excepectation for a that kind of product, open-source or not are way higher than that (but not that high... I'm running Snipsnap as my personal information manager right now).

you were running into problems with jdk 1.5. see http://issues.apache.org/bugzilla/show_bug.cgi?id=33966 for help

also, 1.2.3 still has a bug where it cannot be installed in a path with spaces (since fixed). so that would explain your other error.. thanks for your input for the installer, though.

I played with this a while back and then gave up for the following reason:

1) It bundles its own customized tomcat . I couldn't get the war file to run in a different container

2) too much cocoon and XML -- all ISPs providing Tomcat/Resin hate cocoon xml

3) them I disciovere Mambo!! Learning PHP is less painful than Cocoon internals.

I played with this a while back and then gave up for the following reason:1) It bundles its own customized tomcat . I couldn't get the war file to run in a different container2) too much cocoon and XML -- all ISPs providing Tomcat/Resin hate cocoon xml3) them I disciovere Mambo!! Learning PHP is less painful than Cocoon internals.

Same here, but i discovered Plone!!

And learinig Python and Zope templates even better that PHP !! :))

I played with this a while back and then gave up for the following reason:1) It bundles its own customized tomcat .

jetty actually

I couldn't get the war file to run in a different container

what problems were you having?

3) them I disciovere Mambo!! Learning PHP is less painful than Cocoon internals.

mambo is nice, but serves a different purpose.

A little bit offtopic maybe...
But all these content management systems are, to me, making the same mistake. They try to be everything for everyone, trying to fit themselves into a narrow niche of relatively simple websites, which is already tightly occupied by e.g. Typo3, Mambo, Plone and tens of other great CMS not named here.

If Java CMSes want to position themselves differently, into a custom CMS development market, they have to be shipped as a set of loosely coupled components rather than a monolith. JSR-170 is a great step in this direction, and I'd expect more to come. I can think of detachable WYSIWYG, pluggable template engines, pluggable multi-level caching, pluggable renderers for mobile market, and all the rest, to help CMS developers delivering custom solutions per project concentrating on business logics, rather than try to saw another monolith in pieces.

What we need on Java CMS scene is sort of common CMS architecture, so that everyone could quickly assemble custom solutions from pieces.