> I'm failing to see how this is workable - surly, you want the entire
service to not be there if it has some/all data missing?
Yes, I'd like the service to still be up and running even if some graphs are
missing.
What I'm trying to do is basically this:
- install fuseki on some remote server
- create a shared folder somewhere, where my friends and I can upload/remove
HDT files on the fly
So, it would be pretty useful and cool if, in the Fuseki config file, I could
specify a "HDT directory" property instead of a "HDT graph". Something like
this:
<#dataset> rdf:type ja:RDFDataset ;
rdfs:label "Shared Dataset" ;
ja:HDTdirectory "shared/folder/" .
so, instead of specifying all graphs manually, I just tell Fuseki where it can
find HDT files. Then, ideally, Fuseki should read a list of HDT files from the
directory, and for each file automatically create a new graph. For example, if
it finds a file called graph1.hdt, it will create a graph called <graph1> and
with source graph1.hdt.
This would be really rally cool, because it means we don't have to worry about
the config files, but we can just swap graphs in and out by copying files in
the shared folder.
However, since this seemed like a lot of work to add to Fuseki, I was trying to
think of a simpler solution... so I thought "I can write a huge config files
with a list of 100s graphs, and then Fuseki could just ignore whatever graph is
not available at the moment". This is the reason for my first email. However, I
still think that an additional property "ja:HDTdirectory" would be awesomely
useful if it can make it into Fuseki.
I hope my request is clear? Otherwise ask more questions :)
Sorry for the long email.
