There is a problem of new ontologies popping up for things that have already been formalized. So much energy is lost by people reinventing the "wheel" (in this case, usually the "tag" or "person" classes).
But this wasted energy is missing at the other end - existing ontologies have bugs and problems and need continuing care. FOAF is a good example, every year an updated version is published.

Now Ivan Herman from W3C hinted me to look at this mailinglist at google groups:

http://groups.google.com/group/rdf-schema-dev

Description: A place for authors of RDF schemas to gather and talk shop. The focus is on smaller schemas that don't already have an established community so discussion of improvements to FOAF, SIOC, DOAP and the like are off topic, although reference to those schemas and their designs is not.

But it seems this list is not very active.

My idea would have been to create a "ontology making discussion group". I discussed it with Ivan Herman and others, this is a mixup of his thoughts and mine.

W3C would set up a public mailing list (public-ontology-announce@w3.org), anybody can sign up, archives and RSS feed are publicly. If somebody wants to do an ontology, he/she sends a mail to the list saying 'this is what I want to do and this and this is where the discussions happen' and people may react if such work is already available or somebody is already busy on this. Then, when the work is done, it is announced on that list. Discussions on the work is not supposed to happen on that list, only announcements.

Basically, it sums up to these rules:

• announce your work beforehand
• announce your work when its done
• allow feedback to your work and critique (using the discussion forum of your choice)

Then your ontology is listed on some W3C website, giving it an official blessing, or on a special section in schemaweb.info.

Although W3C says "we don't do ontologies", the GEO ontology having the W3C namespace, or SKOS, have higher weight than others, just because of having the ASCII characters "W3C" in their namespace.

Creating a discussion forum and listing ontologies that have been created according to this procedure would give non-W3C ontologies a similar weight.

This is to ensure that we do not duplicate work and that we begin to establish some formal process of doing ontologies, its meant as a crystallization point around which we can further improve the whole deal of making ontologies.

Any feedback? Does this exist already?