incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Leo Simons <>
Subject Re: [proposal] TripleSoup - a SPARQL endpoint for httpd
Date Mon, 29 Jan 2007 22:02:01 GMT
On Jan 29, 2007, at 9:36 PM, Alex Karasulu wrote:
> Noel J. Bergman wrote:
>> Will it do anything
>> with either Derby, ApacheDS or JackRabbit, in terms of potential
>> collaboration?
> Sounds like it's all C

Yep, that's the plan for now at least.

In "java land", there's two sizeable platforms for doing RDF stuff  
already -- jena [2] and sesame [3], so if I were working on one of  
the projects listed above (or any other java proejct), I'd probably  
try and integrate with those before I'd consider mucking about with C/ 
java interfaces.

> - an Apache Httpd module so it can't easily utilize these servers  
> meant for embedding.

Indeed, at least not easily...the way I'd see integration between  
these kind of pieces like these happening is all of them implementing  
the same (or compatible) interface on a HTTP level and then having a  
federated query system [1] across them.

That said, longer-term, there is a need for a really really really  
fast triple store for use from java, and if we get a lot faster than  
any java version is ever going to be, maybe all the JNI pain in the  
world would still not scare us away :-).

> Perhaps it could use Berkeley DB underneath for an optimized RDF  
> tuple store but there might be licensing issues with that tho.

I would guess so, at least for a bdb > 1.85. The current plan (and  
initial code donations) involves doing many many things "from  
scratch", even if RDFstore does have some bdb support [4].



[1] funny, Alberto just sent me this link today:

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message