[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Semantics of objects in UDD : seeking more standardization, interoperability ?

On 12/02/09 at 18:00 +0100, Olivier Berger wrote:
> Hi.
> (message about UDD, CC-ed to qa list, although I'm not so sure it
> belongs here)
> I've just already responded (Message-Id:
> <1234456666.7959.46.camel@hortense>) to Enrico Zini on d-d@l.d.o about
> the lack of semantics (in the sense of Semantic Web standards used to
> convey ontological references, like RDF schema, etc.) in the outputs of
> DDE, so this message's subject is quite similar.
> I guess the same kind of criticism indeed transitively applies to UDD
> (DDE exports UDD data, IIUC).
> The meaning of the data in the database comes from the choices done at
> database schema design time, and in the code that injects data to UDD
> (sometimes not in obvious ways, as it seems there are inconsistencies
> preventing links between tables, for good reasons).
> I'm thinking about trying to help improve that situation, maybe by
> providing some docs first (UML diagrams and more ?). The goal is to
> enable the use of "commonly agreed" semantics for representation of some
> data, at extraction time, for instance.
> For instance, that could imply mapping bug attributes to EvoOnt BOM (or
> baetle) ontology (http://www.ifi.uzh.ch/ddis/evo/).
> As a (rencent) maintainer of bts-link, and in the frame of project
> (https://picoforge.int-evry.fr/cgi-bin/twiki/view/Helios_wp3/Web/) I'm
> thinking about mapping Debian "facts" to such standard ontologies. That
> would help for instance making semantic links between bugs in different
> bugtrackers in a more standard way, which would supposedly enable easier
> implementation if all bugtrackers understood the same RDF dialects, for
> instance.
> We're trying to evaluate the mapping of different bugtracker's data to
> EvoOnt at http://code.google.com/p/baetle/wiki/EvoOntBomMappings for
> instance (debbugs is on our list).
> I have written a short piece on the subject
> (http://www-public.it-sudparis.eu/~berger_o/weblog/2009/02/10/udd-swim-flossmetrics-facts-databases-about-libre-software-distributions-going-semantic/) after the FOSDEM, and I'd like to get your comments on these ideas.
> Of course, more semantics may not be necessary from the strict immediate
> needs of Debian-specific tools. But I imagine that using standard
> representation formats (for the same semantics of similar objects), QA
> tools could be "ported" from distributions to other distributions, and
> eventually manage inter-distributions data.
> An immedate example would be sharing the "database" of links between
> related bugs (forwarded-to like) allover the net between all
> distributions bugtrackers, enabling maintainers to explore the graph of
> interrelated bugs to ease identification of similar work done by other
> maintainers.

I obviously don't object to UDD being used for that. How would you like
to proceed? Would it be enough to add VIEWs to the DB to map Debian data
to standard ontologies?

I've just made dumps of the DB available (see http://udd.debian.org/),
so you could use that if you want to work on an offline copy.
| Lucas Nussbaum
| lucas@lucas-nussbaum.net   http://www.lucas-nussbaum.net/ |
| jabber: lucas@nussbaum.fr             GPG: 1024D/023B3F4F |

Reply to: