Together with Miguel the project leaders of the Jmol project.
You can visit (and possibly read) my blog at http://chem-bla-ics.blogspot.com/.
My Jmol Experience
Around the summer of 1996 I started working on a website on organic chemistry (Dutch Dictionary on Organic Chemistry) and we needed tools to add interactivity on the website. I quickly stumbled upon one of the Java applet examples showing a rotatable model applet that could read XYZ, which was part of one of the earlier Java releases. I adapted the source code so that it could read PDB files and used this applet on our website.
First, our website was using a file based backend containing the web content. The file format was a custom one, but when the first XML drafts were published in the net I converted this database into XML format. I actually found this new application very early because of the work of Henry Rzepa and Peter Murray-Rust on a chemical application of XML, or the draft actually at that time. When the final XML specification was released, there application transformed into Chemical Markup Language.
So, around 1999 (or so, can't remember the exact dates... have to work that out...) I came into contact with the Jmol and the JChemPaint projects of, respectively, Dan Gezelter and Christoph Steinbeck, and patched their work to include CML support IJC article, so that we could more easily offline maintain our database content, which did include 2D and 3D models of the molecules in our database. There was my first contact with the virtues of opensource software. Apperently, both liked my patches and where included in the main source tree.
Later I wrote more patches for both projects, among which an early draft of Rasmol script interpretation (far from functional compared to the version in Jmol 10) and interpolation for animations (still to be ported to Jmol 10).
In 2000, Dan, Christoph and I met at Notre Dame University, South Bend, USA, where Dan hosted a meeting on the future of Jmol and JChemPaint. The Chemistry Development Kit is a direct evolvement of both projects. In 2003 I worked on making Jmol based on the CDK, e.g. Jmol 9, but Miguel convincingly showed that the CDK core classes where to heavy for Jmol, so he completely rewrote Jmol, resulting in the current state-of-art Jmol 10 release.
But, this did not mean the end of Jmol and CDK integration (JmolCdkIntegration). But that's a whole other story.
Meanwhile, our dictionary website uses the latest Jmol version, and people seem to appreciate it very much, as we have over 1000 unique visitors each day.