Hi all, after digging into Apache Forrest, I introduced it to my project and immediatly it has been grown to our documentation system. We generated forrest-xml out of annotated Java Code and all were happy, except that some people didn't feel comfortable with handlng the complexity of forrest. The developers first were afraid of the tons of features forrest brought with it. Sometimes i had to implement own formats for very special documentation cases.
However at the end I think Forrest has very good ideas and concepts but two drawbacks: 1) its complex and one cannot start immediatly if you want a bit more than just writing content. Even implementing your own design can bring you some pain in the ass. 2) its java and i can only use it for projects which have an own server running, the rights to install it and so on. If I want to get the content from outside my company I have to open a port or rent a root server. This brought me to the conclusion that I need a lightweight Forrest. Means: less features but easier and an easy installation. I have stolen the great ideas of cocoon and forrest and wrote something similar in PHP. I call it PIWI and you can check it out here: http://code.google.com/p/piwi/ Its not complete yet and a very early state, but you can generate piwi xml from different sources, and transform piwi-xml (similar to document-v2.0) to HTML. Of course there is a site.xml. At the moment there is no session handling (needed for localization) and no form-handling, but we plan to implement this beginning in september. A cache would also be nice. Currently we develop at Google Code but I was curious about if the Forrest project or some developers here are interested in this implementation. I know there are not to much php folks at apache, but maybe PHP-Forrest would be a good start. If there is interest, I am willing to contribute that project and can imagine that the Forrest Project may be a good umbrella for that code. Best regards, Chris.
