libeze and blogz are now on code.zedzone.space.
I still haven't moved zedzone to use it but I probably will this week.
I've already prototyped all the code required to move blogz to using lmdb as an index. Conceptually it's very simple, just 3 tables and a few indexes. But there's a lot of boiler-platish code to map it to tables and enforce referential integrity across multiple secondary indices.
Although I haven't implemented it yet the tables support the required meta-data for versioning, renaming, tagging and so on. It will also support comments, wiki-like, and book-chapter like articles. So it will be a stepping stone toward that.
I'm thinking about another blob serialiser that uses a tagged format. This is to support simpler schema upgrades should they be required. On the other hand it may never be necessary and it start to fall into over-engineering so I need to see how it falls out once i've gotten there.