Discussion:
Fuseki vs Marmotta
Laura Morales
2017-03-04 10:35:29 UTC
Permalink
What's the difference between Apache Fuseki and Apache Marmotta?
Jean-Marc Vanel
2017-03-04 10:48:10 UTC
Permalink
Apache Fuseki is a pure SPARQL server, with a native triple database.

Apache Marmotta is a complex beast, primarily an LDP server [1] , but it
mixes a lot of ingredients:
http://marmotta.apache.org/platform/index.html

Its persistence layer is only SQL databases.
It does offer a SPARQL service, but it seems loosely connected to the other
modules.
Especially, I'm not sure that after an LDP PUT or POST, the data will be
added to the underlying SPARQL database.

[1] LDP https://www.w3.org/TR/ldp/

2017-03-04 11:35 GMT+01:00 Laura Morales <***@mail.com>:

> What's the difference between Apache Fuseki and Apache Marmotta?
>



--
Jean-Marc Vanel
http://www.semantic-forms.cc:9111/display?displayuri=http://jmvanel.free.fr/jmv.rdf%23me
Déductions SARL - Consulting, services, training,
Rule-based programming, Semantic Web
+33 (0)6 89 16 29 52
Twitter: @jmvanel , @jmvanel_fr ; chat: irc://irc.freenode.net#eulergui
Laura Morales
2017-03-04 11:08:38 UTC
Permalink
What problem is a "Linked Data Platform" trying to solve that can't already be accomplished with a RDF server like Fuseki? If I'm correct also Fuseki has a REST front end available.



> Sent: Saturday, March 04, 2017 at 11:48 AM
> From: "Jean-Marc Vanel" <***@gmail.com>
> To: "Jena users" <***@jena.apache.org>
> Subject: Re: Fuseki vs Marmotta
>
> Apache Fuseki is a pure SPARQL server, with a native triple database.
>
> Apache Marmotta is a complex beast, primarily an LDP server [1] , but it
> mixes a lot of ingredients:
> http://marmotta.apache.org/platform/index.html
>
> Its persistence layer is only SQL databases.
> It does offer a SPARQL service, but it seems loosely connected to the other
> modules.
> Especially, I'm not sure that after an LDP PUT or POST, the data will be
> added to the underlying SPARQL database.
>
> [1] LDP https://www.w3.org/TR/ldp/
Jean-Marc Vanel
2017-03-04 11:27:48 UTC
Permalink
2017-03-04 12:08 GMT+01:00 Laura Morales <***@mail.com>:

> What problem is a "Linked Data Platform" trying to solve that can't
> already be accomplished with a RDF server like Fuseki?

Consider this use case:

- manage a team's public FOAF profiles with URL prefix http://xx.com/
- a team member X can upload her FOAF profile simply by an HTTP POST
request on http://xx.com/members/ at URL /myName
- then http://xx.com/myName is visible from Internet and contains
triples about <http://xx.com/myName> , which is the member's URI.


If I'm correct also Fuseki has a REST front end available.
>

Well the word REST used in
https://jena.apache.org/documentation/serving_data/ is simply meaning that
a SPARQL compliant HTTP server can be somehow considered a REST server.

But it's not what is generally meant by a REST server:
https://en.wikipedia.org/wiki/Representational_state_transfer
In this REST concept , one PUTs or POSTs data at some relative URL's, and
by HTP GET at the same URL retrieves the data.

This is what an LDP server basically does.
An LDP server also can be viewed as similar to an FTP server that typically
stores RDF data (but also any binary data).


>
>
> > Sent: Saturday, March 04, 2017 at 11:48 AM
> > From: "Jean-Marc Vanel" <***@gmail.com>
> > To: "Jena users" <***@jena.apache.org>
> > Subject: Re: Fuseki vs Marmotta
> >
> > Apache Fuseki is a pure SPARQL server, with a native triple database.
> >
> > Apache Marmotta is a complex beast, primarily an LDP server [1] , but it
> > mixes a lot of ingredients:
> > http://marmotta.apache.org/platform/index.html
> >
> > Its persistence layer is only SQL databases.
> > It does offer a SPARQL service, but it seems loosely connected to the
> other
> > modules.
> > Especially, I'm not sure that after an LDP PUT or POST, the data will be
> > added to the underlying SPARQL database.
> >
> > [1] LDP https://www.w3.org/TR/ldp/
>



--
Jean-Marc Vanel
http://www.semantic-forms.cc:9111/display?displayuri=http://jmvanel.free.fr/jmv.rdf%23me
Déductions SARL - Consulting, services, training,
Rule-based programming, Semantic Web
+33 (0)6 89 16 29 52
Twitter: @jmvanel , @jmvanel_fr ; chat: irc://irc.freenode.net#eulergui
Andy Seaborne
2017-03-04 13:19:59 UTC
Permalink
On 04/03/17 11:27, Jean-Marc Vanel wrote:
> 2017-03-04 12:08 GMT+01:00 Laura Morales <***@mail.com>:
>
>> What problem is a "Linked Data Platform" trying to solve that can't
>> already be accomplished with a RDF server like Fuseki?
>
> Consider this use case:
>
> - manage a team's public FOAF profiles with URL prefix http://xx.com/
> - a team member X can upload her FOAF profile simply by an HTTP POST
> request on http://xx.com/members/ at URL /myName
> - then http://xx.com/myName is visible from Internet and contains
> triples about <http://xx.com/myName> , which is the member's URI.
>
>
> If I'm correct also Fuseki has a REST front end available.

Yes - it supports the SPARQL Graph Store Protocol.

PUT/GET/POST/DELETE whole graphs.

... and it generalises it to the dataset+quads as well.

>>
>
> Well the word REST used in
> https://jena.apache.org/documentation/serving_data/ is simply meaning that
> a SPARQL compliant HTTP server can be somehow considered a REST server.
>
> But it's not what is generally meant by a REST server:
> https://en.wikipedia.org/wiki/Representational_state_transfer
> In this REST concept , one PUTs or POSTs data at some relative URL's, and
> by HTP GET at the same URL retrieves the data.
>
> This is what an LDP server basically does.
> An LDP server also can be viewed as similar to an FTP server that typically
> stores RDF data (but also any binary data).

The big thing that LDP adds is its container model.

Andy

>
>
>>
>>
>>> Sent: Saturday, March 04, 2017 at 11:48 AM
>>> From: "Jean-Marc Vanel" <***@gmail.com>
>>> To: "Jena users" <***@jena.apache.org>
>>> Subject: Re: Fuseki vs Marmotta
>>>
>>> Apache Fuseki is a pure SPARQL server, with a native triple database.
>>>
>>> Apache Marmotta is a complex beast, primarily an LDP server [1] , but it
>>> mixes a lot of ingredients:
>>> http://marmotta.apache.org/platform/index.html
>>>
>>> Its persistence layer is only SQL databases.
>>> It does offer a SPARQL service, but it seems loosely connected to the
>> other
>>> modules.
>>> Especially, I'm not sure that after an LDP PUT or POST, the data will be
>>> added to the underlying SPARQL database.
>>>
>>> [1] LDP https://www.w3.org/TR/ldp/
>>
>
>
>
A. Soroka
2017-03-04 13:26:34 UTC
Permalink
> The big thing that LDP adds is its container model.
>
> Andy

Yes. This is hugely useful, if it meets your use cases. It allows for a lot of automatic management for an important class of relationships.

https://www.w3.org/TR/ldp/#ldpc

---
A. Soroka
The University of Virginia Library

> On Mar 4, 2017, at 8:19 AM, Andy Seaborne <***@apache.org> wrote:
>
>
>
> On 04/03/17 11:27, Jean-Marc Vanel wrote:
>> 2017-03-04 12:08 GMT+01:00 Laura Morales <***@mail.com>:
>>
>>> What problem is a "Linked Data Platform" trying to solve that can't
>>> already be accomplished with a RDF server like Fuseki?
>>
>> Consider this use case:
>>
>> - manage a team's public FOAF profiles with URL prefix http://xx.com/
>> - a team member X can upload her FOAF profile simply by an HTTP POST
>> request on http://xx.com/members/ at URL /myName
>> - then http://xx.com/myName is visible from Internet and contains
>> triples about <http://xx.com/myName> , which is the member's URI.
>>
>>
>> If I'm correct also Fuseki has a REST front end available.
>
> Yes - it supports the SPARQL Graph Store Protocol.
>
> PUT/GET/POST/DELETE whole graphs.
>
> ... and it generalises it to the dataset+quads as well.
>
>>>
>>
>> Well the word REST used in
>> https://jena.apache.org/documentation/serving_data/ is simply meaning that
>> a SPARQL compliant HTTP server can be somehow considered a REST server.
>>
>> But it's not what is generally meant by a REST server:
>> https://en.wikipedia.org/wiki/Representational_state_transfer
>> In this REST concept , one PUTs or POSTs data at some relative URL's, and
>> by HTP GET at the same URL retrieves the data.
>>
>> This is what an LDP server basically does.
>> An LDP server also can be viewed as similar to an FTP server that typically
>> stores RDF data (but also any binary data).
>
> The big thing that LDP adds is its container model.
>
> Andy
>
>>
>>
>>>
>>>
>>>> Sent: Saturday, March 04, 2017 at 11:48 AM
>>>> From: "Jean-Marc Vanel" <***@gmail.com>
>>>> To: "Jena users" <***@jena.apache.org>
>>>> Subject: Re: Fuseki vs Marmotta
>>>>
>>>> Apache Fuseki is a pure SPARQL server, with a native triple database.
>>>>
>>>> Apache Marmotta is a complex beast, primarily an LDP server [1] , but it
>>>> mixes a lot of ingredients:
>>>> http://marmotta.apache.org/platform/index.html
>>>>
>>>> Its persistence layer is only SQL databases.
>>>> It does offer a SPARQL service, but it seems loosely connected to the
>>> other
>>>> modules.
>>>> Especially, I'm not sure that after an LDP PUT or POST, the data will be
>>>> added to the underlying SPARQL database.
>>>>
>>>> [1] LDP https://www.w3.org/TR/ldp/
>>>
>>
>>
>>
Loading...