Wow, thanks for documenting the idea, this could be a nice alternative to my tiddlywiki markdown setup. I could do all kinds of queries on my knowledge base. Still I could use grep as well for document counts and lists. It's neat that it's all in one compressed sqlite file - but I can tar ball my markdown directory as well. The only benefit I see is that I can use sql to query my documents and it seems like a step up from org-mode tables by using database tables.

It would be nice if there was cgi script that could serve a sorted and paged index with links to the html, md or gmi to a web or/and gemini browser.