I started generating /doap/{tlp-id}/pmc.rdf
Please tell me what to add in these files
I don't understand what to write into /doap/foundation/tlps.rdf
And for /doap/{tlp-id}/{project-id}.rdf, the content can't be generated: there
is a lot of handwritten content we can't guess automatically
What we can do is copying projects hadcrafted content to the structured
location: but even extracting the {project-id} is not easy
any idea?
Regards,
Hervé
Le mardi 12 mai 2015 18:23:12 Sergio Fernández a écrit :
> Hi,
>
> On Tue, May 12, 2015 at 8:47 AM, Hervé BOUTEMY <[email protected]>
>
> wrote:
> > I think we could generate an authoritative DOAP url for TLPs from
> > committee-
> > info.txt
> > then give instructions to projects to update their software DOAP files to
> > point
> > to these reference TLPs DOAP files
>
> Exactly. I think a conclusion we could arrive with the years using DOAP
> files in ASF is simple: handcrafted files are
>
> So I think the goal for projects-new.apache.org would be to automatically
> generate the DOAP files, including: TLPs (PMCs), projects and releases. We
> have that data somewhere (LDAP?), I'm pretty sure, we just need to get it
> and process it.
>
> > I can generate tonight http://projects-new.apache.org/doap/tlp/ as a POC
> > for
> >
> > https://svn.apache.org/repos/asf/infrastructure/site-tools/trunk/projects/
> > data_files replacement
>
> OK, bootstrap it, and then I'll jump it to do the proper RDF coding there.
>
> I'll propose to publish at least:
>
> 1) /doap/foundation/tlps.rdf
> 2) /doap/{tlp-id}/pmc.rdf
> 3) /doap/{tlp-id}/{project-id}.rdf
>
> 1 will provide the basic information, linking to 2 containing the pmc
> information, as we already do in json. Then 3 would also link to 2, where
> in most of the cases {tlp-id} and {project-id} will be the same, but we do
> support subprojects (e.g., lucene) and components (e.g., commons).
>
> One we have that infrastructure in place, we can tell all TLPs to directly
> link to those files, avoiding the manual duplication of that data.
>
> I think with that we are coming closer to a satisfactory solution ;-)
