Halyard 1.5 ("Lipno")

The biggest changes in Halyard 1.5 release are within its tooling.
HalyardBulkUpdate now support more complex configuration, it supports deletion of statements, and multi-stage SPARQL Update sequences.
New tools HalyardBulkExport and HalyardBulkDelete have been introduced.
HalyardElasticIndexer has been improved and more documentation about Halyard cooperation with ElasticSearch has been added.
Significant performance improvements for datasets with zillion of named graphs include: de-duplication of triples in default graph context, fixes in CardinalityCalculator, and improved HalyardStats generation.
Last but not least - underlying RDF4J has been updated to version 2.3.2

Halyard 1.4 ("Newark")

Halyard 1.4 release fixes several bugs, for example:
SPARQL queries containing optional sub-selects are now fully compliant with the specification .
Property paths in SPARQL queries should no more cause random performance problems.
And SAIL now response correctly on some no-standard situations.
Underlying RDF4J framework has been upgraded to version 2.2.4 and all the Halyard Strategy compliance tests are now simulating latency to better simulate and catch various asynchronous race conditions.
Default Halyard properties has been a bit adjusted and extended improve performance with respect to the latest measurements.
And at last but not least a new Halyard library module with RDF4J RIO extensions has been introduced to support various new formats.
Initially there are NDJSONLDParser as a new-line-delimited extension of JSONLDParser, and JSONParser providing an experimental synthetic mapping from general JSON file into RDF.

Halyard 1.3 ("Vienna")

Halyard 1.3 release contains several bug-fixes and new tools and feature. Halyard HBase SAIL newly cooperates with supplementary ElasticSearch for advanced text search capabilities over the stored literals. Halyard Bulk Update tool newly supports SPARQL Delete/Insert queries. Halyard Bulk load supports loading from locally provided Hadoop filesystems (for example Amazon S3 or Google GFS). New time-aware HBase SAIL has been introduced with basic support in several Halyard tools to open doors for time-aware graph operations.

You can find it in the documentation here: https://merck.github.io/Halyard/

Thank you,
Adam Sotona

Halyard 1.2 ("Under Attack")

Halyard 1.2 release contains several bug-fixes and improvements. Statistical tool is improved and it generates more advanced statistics. Performance of the SPARQL query evaluation is improved by re-ordering of joins based on the actual statistics. Bulk load performance has been improved for large number of source RDF files. New pre-split calculator creates optimal HBase table region splits for fast and smooth bulk load process of very large datasets.

You can find it in the documentation here: https://merck.github.io/Halyard/

Thank you,
Adam Sotona

Halyard 1.1 ("Spring")

Halyard 1.1 release contains several bug-fixes and improvements, new statistical tool, new support for cross-datasets queries using SPARQL SERVICE, and new documentation with more tutorials.

Check for documentation here: https://merck.github.io/Halyard/

Enjoy,
Adam Sotona

Halyard 1.0 ("Not In December")

This is an initial Halyard release so everything is new in the release notes.

Check for documentation here: https://merck.github.io/Halyard/

Feel free to add your feedback, comments and suggestions under Issues.

Thank you,
Adam