-
-
Notifications
You must be signed in to change notification settings - Fork 252
Suggestions for further development
Johannes Lichtenberger edited this page Apr 5, 2018
·
3 revisions
For a more up-to-date list, have a look at the issues.
- horizontal scaling based on Akka(.io) or more likely with the help of Apache Ignite
- provide checkpointing and reapply changes after a power-failure... (full ACID compliance at least on a single computer/node)
- provide a B+-tree for large-index-structures (most probably based on the indirect page-tree), thus a faster COW variant but might be a bit slower for reading (because of leaf-page versioning and indirect pointer access). Version only leaf-pages.
- make index-structures dependent on self-tuning aspects (that is probably test-data sets from which to learn)
- include the index-structures and rewriting rules in the XQuery/XQuery Update Facility processor Brackit
- implement fine grained user-management
- implement n-writing transactions concurrently
- implement mapping from DeweyIDs to unique NodeKeys (for instance to facilitate fast ancestor lookup which is necessary for instance to implement the above item (concurrent write-transactions))
- implement nodes for XQuery which can be reconstructed in memory with the help of the path summary