If you haven't been to the Pentaho wiki lately (or you avoid it for fear that you will get lost forever in there) there have been some changes. Check out the new home page - http://wiki.pentaho.com which has a new style and even has useful information! Eventually, the new style or something similar will make it's way to the documentation pages.

But wait, there's more...

An even bigger and more important change is the Downloads section - http://wiki.pentaho.com/display/COM/...tion+Downloads. There is a brief explanation of what is available for download and a launch pad that helps you figure out what to downloads you need and where to find them:

* Need Help - I would like more information about the available downloads and builds
* Get Me Started - Show me the most likely downloads for a beginner grouped by task
* Stable Builds - I want to use or evaluate Pentaho, show me all the stable builds
* Release Candidates - I want a preview of what's next and to help with testing, show me all the RCs
* Milestones - I want to try new features in time to provide feedback to the developers, show me the milestones
* CI Builds - I want to help develop new features and test specific bug fixes
* Source Code - I want to write or debug code
* EE Evaluation - I would like to try the enterprise features and be assisted with my evaluation

Please give me feedback on anything that is still confusing or could be made clearer. The idea it to make finding the right download among the hundreds available, easy for new people unfamiliar with open source and also for developers that know what they are looking for. The next area I'd like to tackle is a guided evaluation of CE for both people new to BI and BI experts new to Pentaho. I'm looking for volunteers to help... hint... hint...

For the curious...

The pages with all the little download icons are generated using the BI Suite. The download data is scraped from SourceForge and pumped into a MySQL database using Kettle. The process is mostly automatic with 1 quick manual step to add some metadata about the new files since we don't have a consistent naming convention among projects. By scraping SF, we guarantee that all the links and file names are correct (1 version of the truth) and as files are added and hidden, the pages are correctly updated. I hope to have this be a fully automatic process soon. The BI Server is generating wiki code that is manually copied and pasted into the wiki. I hope to automate that step too and maybe even generate HTML using Pentaho Reporting and save the pages directly to the web site.

I plan to merge the scraping code into a Kettle SourceForge input step and do a tech tip or blog about the whole process. It could be generally useful for other open source projects and is an interesting use of Pentaho applications.