We've built a document editor that might be suitable for reading (and writing) textbooks. It allows the text to refer to figures and details within those figures (e.g. "In Fig. 3, roller 22 is mounted on support 24.") It's originally built to prepare patent applications (see teampatent.com and ask for a beta account) but has wide applicability in other fields. It's collaborative like Google Docs, runs on Amazon Web Services so it can scale, and is built exclusively with open source (with a Python backend) so we have flexibility on how to deploy it. I can't reach ozanonay at her website so contact me at rocky at teampatent dot com.