incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Reto Bachmann-Gmuer <reto.bachm...@trialox.org>
Subject Re: FOO proposal and Clerezza
Date Tue, 02 Nov 2010 07:17:01 GMT
On Mon, Nov 1, 2010 at 12:00 PM, Bertrand Delacretaz <bdelacretaz@apache.org
> wrote:

> Hi,
>
> On Mon, Nov 1, 2010 at 10:46 AM, Reto Bachmann-Gmuer <reto@apache.org>
> wrote:
> > ...The goal of the FOO proposal ("a modular software stack and reusable
> set of
> > components for semantic content management") seems quite similar to the
> one
> > of Clerezza ("an OSGi-based modular application and set of components
> > (bundles) for building RESTFul Semantic Web applications and
> services.")...
>
> I agree about the goal as described in the proposal - maybe we need to
> make it clearer there, as in reality I think the projects are further
> apart than that.
>
> My understanding is that Clerezza aims to be an out-of-the-box
> solution to build websites and web applications based on semantic
> storage of content. Please correct me if I'm wrong.
>
It shall be possible to use Clerezza as platform, but basically its a set of
components.


> The goals of FOO are broader IMO, and less "industrial" in many ways.
>
> While some components of FOO might be assembled to do the above, the
> main use of those components, is to provide *other* CMS platforms with
> semantic components, which might be more more experimental than the
> Clerezza production-ready stuff.
>
Not all of the Clerezza modules are part of the default platform
distribution, so we already have less production-ready module too.

Seeing that lots of things are done in parallel (templating system,
persistency abstraction, JAX-RS) I think FOO and CLEREZZA might as project
both have the highest benefit with a joined community. This is not to say
they couldn't exist as twin project and benefit from each other, but at the
current state I'm not sure if such a high level of parallelism is positive
for the young communities.



>
> So, although there's a sizable interesection between the FOO and
> Clerezza components, I think the end goal is quite different,
> sufficiently so to warrant two distinct projects.
>

I'm not sure, on one hand FOO provides what could be seen as a distribution
and front-end adaptor for Clerezza components. On the other hand (from what
I understood talking to some of the community members) FOO also wants to
exploit more possibilities of the backend and support inferencing and rule
based reasoning. So my impression is, that on one hand it aims to provide
some simple APIs for integration within existing CMS systems (as requested
by the industrial members of the IKS consortium) on the other hand it also
wants to use more of the semantic web possibilities. I think both directions
of development could fit well within Clerezza, but as separate projects I
think the problem is they address two interwoven and overlapping set of
layers which leads not only to duplicate efforts but also to a complicated
architecture with cyclic dependencies.


>
> >... Since several people are involved in both project, I'm curios on how
> the two
> > projects goals or approaches differ, FOO is for Semantic Content
> Management,
> > Clerezza for Semantic Web Application. Is there really a difference
> (taking
> > into account how Celrezza isn't just focuses on the web presentation
> layers
> > but on the whole stack and that FOO is based on OSGi too)?...
>
> IMO the best is to view Clerezza and FOO as sister projects, with lots
> of exchanges between their communities and a sizable chunk of common
> software components, but different end goals and audiences. This will
> enable both projects to follow their own goals while finding
> synergies.
>

With the recent development of Clerezza towards adding support for Social
Web features alongside the semantic stack the target audience is clearly
broader than traditional CMS vendors. Still, I don't think the target
audiences are too distinct and looking at the current FOO code I think
there's quite some duplication and that a lot could be simplified by using
the pattern supported by Clerezza.


>
> Maybe this needs to be better spelled out in the proposal and project
> charter, I'm open to suggestions on that.
>

I have nothing against having multiple semantic web projects at Apache with
a mixture of collaboration and competition. However I think the differences
either in what is implemented or in how it is implemented should be clear.
Currently I see as differences (FOO <-> Clerezza):

- jersey with proprietary jax-rs extensions <-> triaxrs (clerezza jax-rs
impl)
- resource bound to path <-> resource bound to path or rdf type
- freemarker <-> plugable templating engines with scala server pages as
default
- no systematic conneg (production of html by resource classes) <-> resource
classes normally produce RDF, conneg by rendering system
- uses rdf stores partially provided via clerezza but also uses SQL <-> data
Access over an RDF abstraction layer


Reto


>
> -Bertrand
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message