Håvard M. Ottestad

Marked this as stale to signal that the PR hasn't been active for a while and that we should consider closing it.

Using protected makes sense where a class is intended to both be used by users and be extended by developers. Public methods would be visible to users, while protected methods...

``` Benchmark Mode Cnt Score Error Units QueryBenchmark.complexQuery avgt 5 1.914 ± 0.009 ms/op QueryBenchmark.different_datasets_with_similar_distributions avgt 5 0.966 ± 0.002 ms/op QueryBenchmark.groupByQuery avgt 5 1.175 ± 0.028 ms/op QueryBenchmark.long_chain avgt...

# Updated benchmark results ``` Benchmark Mode Cnt Score Error Units QueryBenchmark.complexQuery avgt 5 1.976 ± 0.018 ms/op QueryBenchmark.different_datasets_with_similar_distributions avgt 5 0.932 ± 0.125 ms/op QueryBenchmark.groupByQuery avgt 5 1.361 ±...

Marked this as stale to signal that the PR hasn't been active for a while and that we should consider closing it.

I wonder if this could be related to https://github.com/eclipse/rdf4j/issues/3696 ? The [specification](https://www.w3.org/TR/sparql11-query/#func-not-in) for `NOT IN` says that it should be equivalent to `!=` so that `?a NOT IN (1,2,3)` should...

@eclipse/technology-rdf4j-committers We have started discussing this on the rdf4j dev mailing list and I've opened an issue here to track our decision. Comments are very welcome :)

Some more ideas: Async IO to produce Values that will be materialized in the future. Async Nio has a nifty way of handling closing of a resource where you can...

@JervenBolleman is this PR ready to merge?

@jeenbroekstra does this bug also apply to 4.0.0-M3? Is it something that can/should be fixed before 4.0.0?