A draft of the first chapter of the ZODB book has been committed to our GitHub repository. It's just the introductory chapter, so it might not seem like much. However, the introduction is very important when starting a technical book, since it sets the tone and helps define the audience more clearly.
If you can spend a few minutes to read it and comment, that would be very helpful.
By the way, there is now a portlet at the top right of the blog that shows the latest commits to the repository. This may make it easier to keep track of the book's progress.
I would collapse familiarity and simplicity. Python programmers who come to the ZODB without RDBMS prejudice will be really impressed by both of these aspects.
ReplyDeleteTransparency and schemaless-ness. Also say that references in your python program works the same as ZODB.
- framework independent
- django, twisted, WSGI - works with any framework.
Regarding trnx support. I think it is vital to explain ZODB supports MVCC. Quickly people will find out about ConflictResolution and in the introduction I think TRNX/MVCC are two aspects that developers do not need to think about, ... usually ;)
I believe Undo and History should be collapsed into a single paragraph. Undo is rarely useful in production situation.
I would say Caching and talk about the two different levels of caching. In-memory and persistent disk cache.
Scalability, I would take out until you get more content. It is scalable in the sense of you can store terrabytes of data (esp blob) but it requires some design anticipation.
I would also draft potential hangups. i.e.
- zodb stores data in python native format.
It is better for someone to read introduction and get both sides (PRO and CON) and leave before wasting time.