Olaf Hartig

Yes, I can confirm that in the TPF work we have explicitly ignored blank nodes and, thus, the TPF interface does not support them. The reason has nothing to do...

Hi Bart, I think what you want is already possible as a result of the refactoring that I did over the last weeks. For instance, the refactored code base allows...

Hi @kjetilk Find an extended preprint of the full paper about brTPF here: http://arxiv.org/pdf/1608.08148

Yes we did. In the ISWC paper we (i) defined the notion of matching triples for a given triple pattern and (ii) formalized triple pattern selectors in terms of this...

@mielvds (null,p,null) is to be interpreted as (?x,p,?y) where ?x and ?y are two arbitrary variables such that ?x != ?y. @RubenVerborgh Regarding the encoding of variables that are specified...

I agree that blank nodes in TPF requests should be dealt with in the same way as (specific, named) variables are dealt with. Therefore, I would extend the `TriplePatternElement` interface...

@RubenVerborgh I think that automatically converting unspecified variables into named variables is not a good idea because it may cause problems if the named variable that you would assign to...

If it was only for pure TPFs, such an intelligent assignment would work (but it might have a slight impact on the server performance because it is extra work). However,...

The `TriplePatternElement` interface has been extended as discussed above. However, the actual bug that this ticket is about still exists.

Several of your points here seem to assume that the query templates we had for the BSBM-based version of LinGBM and the query templates we have now for the LUBM-based...