[Mulgara-general] Setting SAIL with Mulgara?

Paul Gearon gearon at ieee.org
Wed Mar 12 16:28:56 UTC 2008


Hi David,

This is coincidental that you would ask this now.

Just last week we had general agreement to implement the SAIL API for  
Mulgara.  As you say, most of the interface is the same as what we see  
on the Session interface, so we don't think there will be a problem  
there at all.  Also, as you say, the only tricky part is the query  
language.

James Leigh (of Aduna) pointed me to the query algebra for Sesame, and  
it looks like the translation will be pretty easy (once I've finished  
the implementations for FILTER and OPTIONAL in our own algebra).

Other than the API and a consistent implementation of SPARQL, we would  
also get the SPARQL protocol adapter (which I haven't started on) as  
well as some of the other modules that operate on the SAIL API.

So it looks to be a desirable thing, AND not too hard to do.  The  
problem is priorities.  Andrae and I are swamped at the moment, and I  
know that Ronald is concentrating on those areas that directly affect  
Topaz.

If anyone (including you David!) wants to start on this, then the help  
would be greatly appreciated.  Andrae already had a few ideas about  
it, so anyone interested should write to ask him about it.

Regards,
Paul Gearon

On Mar 11, 2008, at 6:23 PM, David Legg wrote:

> I'd like a sanity check please!
>
> I would like to convert a project that currently uses Sesame to using
> Mulgara instead.  Unfortunately, the code is liberally sprinkled with
> references to Sesame objects which would make re-targeting it very  
> involved.
>
> Now that I've taken a cursory glance at the Sesame architecture it
> occurs to me that it has been designed to allow other triplestores  
> to be
> incorporated.  I think I need to create a Mulgara implementation of  
> the
> SAIL API (Storage And Inference Layer)... or maybe this already exists
> somewhere?
>
> I suppose one fly in the ointment is the query language.  Sesame uses
> SeRQL or SPARQL neither of which are supported in Mulgara (but that's
> changing real soon now - right Paul? ;-) ).  It's funny... it takes me
> back to all the dialects of SQL we had to worry about in ages past.
> Actually, it might not be that bad because since Sesame allows you to
> use one of two query languages it already has a mechanism to add a  
> third.
>
> Regards,
> David Legg
>
> _______________________________________________
> Mulgara-general mailing list
> Mulgara-general at mulgara.org
> http://mulgara.org/mailman/listinfo/mulgara-general




More information about the Mulgara-general mailing list