Released OrientDB v1.4: new TinkerPop Blueprints API and many other features
From the post:
NuvolaBase is glad to announce the new release 1.4 of OrientDB: http://www.orientdb.org/download.htm!
What’s new with 1.4?
- Graph: total rewrite of TinkerPop Blueprints API that now are the default Java interface, support for light-weight edges (no document), labeled relationships using separate classes and vertex fields
- Storage: new Paged-Local compressed “plocal” engine (not yet transactional)
- SQL: INSERT and UPDATE supports JSON syntax, improved usage of indexes upon ORDER BY, supported timeout in query and global, new create function command, flatten() now is expand(), new OSQLMethod classes to handle methods even in chain, new encode() and decode() functions, support for new dictionary:<key> as target in SELECT and TRAVERSE
- new SCHEDULER component using CRON syntax
- new OTriggered class to use JS as hook
- MMap: auto flush of pages on regular basis
- Fetch-plan: support for skip field using “-2”
- Index: auto rebuild in background, usage of different data-segment
- Export: supported partial export like schema, few clusters, etc.
- Console: improved formatting of resultsets
- HTTP: new /batch command supporting transaction too, faster connection through /connect command, /document returns a JSON
- Studio: UML display of class
I did not mean to get so distracted yesterday writing about useless projects that I would neglect very useful ones, like the new release of OrientDB!
The “light-weight edges” sounds particularly interesting.
With no documentation you could not reliably merge the edges but as a design decision, that could be entirely appropriate.
You are always going to be faced with decisions of what subjects are represented at all and to what extent.
Nice to have a mechanism that doesn’t make that an entirely binary decision.