We've now released the latest version of dotNetRDF which is 0.1.3 Alpha (the official build number is 0.1.3.20723). This release is primarily an interim release to fix several parsing and serialization issues that were identified in the previous release but since we had already made a start on various new features and internal changes as we move towards the 0.2.x series of releases next year you get a few bonuses in this release.

Key features are:

• New Turtle and Notation 3 Parsers - slimmer, faster and eliminating the blank node handling bugs
• Many writers now completely thread safe, some parsers completely thread safe
• New SPARQL triple pattern ordering optimisation gives another 3 fold increase in query performance
• Experimental Joseki support - use with caution (mostly untested)
• Experimental Mono 2.6 build (source download only)

You can Download dotNetRDF to just get the library or you can Download dotNetRDF Source to get the full source code.

There are currently no Tags for this Content!

We unfortunately have had some major bugs with dotNetRDF reported this week. As these are sufficiently serious to break key features of the library we are slating a 0.1.3 release for next week - the reason this will be a new release and not a re-release of 0.1.2 is that we have already made progress with implementing some additional features in the library and altering the internal architecture of quite a number of classes. While these changes were all slated for the 0.2.x releases making another release allows us to fix these issues and release some of the small changes sooner - in particular it allows people to be aware of a number of classes now marked as obselete which we plan to remove in the 0.2.x builds.

The issues are as follows:

1. When using the Turtle/Notation 3 parsers there are flaws in Blank Node handling:
   1. If you use an anonymous Blank Node as the subject of multiple triples by use of predicate-object or object lists the first triple's subject gets given a different ID to the subsequent triples
   2. If you use a Blank Node ID which is the same as an auto-assigned one it gets remapped to a new ID but the first Triple's ID is not remapped correctly
2. When using any of the RDF/XML serializers on a Graph which has a namespace bound to the empty prefix the serializers fail.
3. When using the Fast RDF/XML writer in certain cases a Blank Node subject which has an rdf:type stated for it fails to be output

I apologise to anyone who's affected by these issues - I build this library because I want to use it myself so I appreciate how much of a pain it is when things are broken. Remember this is still Alpha quality software, while we plan to get to Beta quality around the 0.5.x releases we're still only on the 0.1.x releases so please be patient.

As a little sweetener one of the changes you'll get in the 0.1.3 release is a new query optimisation for our in-memory SPARQL engine which gives a three-fold performance increase and better scalability for complex queries.

There are currently no Tags for this Content!