One of the advantages of the Ext 4.0 documentation tool (JSDuck) is that it generates documentation that is crawlable/indexable by Google. I'm working on generating the documentation for my project and would like to reap the same benefits. I'm curious as to how Sencha is generating the HTML snapshots. It doesn't look like JSDuck emits the fragments as static content. Are you using a servlet filter with HTMLUnit like Google recommends or some other technology? I'll probalby go down that route unless I hear back that there is a better way.