Hi All, Reto, Based your comment in this other discussion I reason that your are working on updating Stanbol to the new Clerezza RDF api.
This is to inform you that (without further discussion) I am -1 for such an update. The main reasoning behind this was the decision of Clerezza for NOT using Apache Commons RDF (incubating) [1] as the RDF API for Clerezza. Because of that we are now in a situation where we have +1 competing Java RDF APIs (see http://xkcd.com/927/). Here is a Summary of the current situation as I see it: * Apache Commons RDF [1] is clearly not ready for production. * It only recently released it first version of the API. * We need to see how the adaption of this API is going. Especially by the big two (Apache Jena [2] and Sesame [3]). See if Clerezza also adapts Apache RDF commons somehow. * We need to evaluate its API against the needs of the Stanbol Enhancer (in-memory model, parsing, serializing) * We need to evaluate how to use this API in OSGI, JAX-RS, ... * Apache Clerezza introduces a new Clerezza Commons RDF API * The new Clerezza release (currently in voting) adapts this API * This API is similar to the Apache Commons RDF API but AFAIK incompatible * The new API is incompatible with the old. Therefore code changes are required in Stanbol after the update (esp. all EnhancementEngines need to be adapted) IMHO it is simple not possible to make any good decision between those right now. If Apache Commons RDF gets adapted by both Sesame and Jena it will probably be the better choice (esp. in the long turn). If not updating to the most current Clerezza version could make more sense. What I want to avoid are two RDF API changes within a short period of time. Such a change does not only require to adapt a lot of Apache Stanbol implementations. As classes of the RDF API are also used in central Stanbol Interfaces (most important the EnhancementEngine and ContentItem interface) this also affects users with custom Stanbol components. In conclusion: Sticking with the current Clerezza versions for some additional time (maybe a year) does not hurt. Doing so will allow for a much better informed decision for the future RDF API for Apache Stanbol. best Rupert p.s. In a questionary for the best RDF API choice for Apache Stanbol I would opt for the Sesame Model API [4]. As this in-memory model API is a really great fit for the use case of the Stanbol Enhancer. [1] http://commonsrdf.incubator.apache.org/ [2] https://jena.apache.org/ [3] http://rdf4j.org/ [4] http://rdf4j.org/sesame/2.8/apidocs/org/openrdf/model/package-summary.html On Mon, May 25, 2015 at 11:05 PM, Reto Gmür <m...@farewellutopia.com> wrote: > Ok, thanks for the prompt answers. For jsonld there currently is some > outdated support in stanbol but not directly in clerezza. I will port > rdf/json and try to have a release asap. > > Cheers, > Reto > On May 25, 2015 8:25 PM, "Ed - 0x1b, Inc." <e...@0x1b.com> wrote: > >> rdf/json. +1 >> On May 25, 2015 9:37 AM, "Reto Gmür" <r...@apache.org> wrote: >> >> > Hello, >> > >> > There has been a major update in clerezza and many modules have been >> > released as 1.0.0. >> > >> > Updating stanbol to the new clerezza artifacts I've noticed that several >> > modules include support for rdf/json. Howver the clerezzza bundle >> > supporting this format has not been updated to use the new RDF libraries. >> > As I though nobody would be using this format anymore I wasn't planning >> to >> > port it. >> > >> > >> > So, does anybody still needs or uses rdf/json (application/rdf+json)? >> > Otherwise I'l remove the support. >> > >> > Cheers, >> > Reto >> > >> -- | Rupert Westenthaler rupert.westentha...@gmail.com | Bodenlehenstraße 11 ++43-699-11108907 | A-5500 Bischofshofen | REDLINK.CO .......................................................................... | http://redlink.co/
