Xesam meta-meta-data spec needs attention.

[Mikkel Kamstrup Erlandsen]

2007/5/12, Evgeny Egorochkin <phreedom.stdin at gmail.com>:
>
> On Saturday 12 May 2007 00:14:55 Mikkel Kamstrup Erlandsen wrote:
> > Basically we have four desktop ontologies that I know of:
> >
> > Strigi:
> >
> http://websvn.kde.org/trunk/kdesupport/strigi/src/streamanalyzer/fieldprope
> >rties/ Tracker :
> http://svn.gnome.org/viewcvs/tracker/trunk/data/services/
> > Spotlight:
> >
> http://developer.apple.com/documentation/Carbon/Reference/MetadataAttribute
> >sRef/index.html
> >Nepomuk: ? I couldn't find one... Sebastian, Jos, Phreedom?
>
> Parts of Nepomuk are still being drafted. Naturally it's RDF-based. I can
> ask
> Nepomuk if they are ok with releasing some parts of the draft, though you
> can
> consider Strigi onto a reasonable approximation of Nepomuk onto with some
> workarounds due to:
> * absense of file types(as of yet)
> * ban on multiple-node descriptions of files.
> * some extensions in situations where Strigi had to have something before
> Nepomuk
> * some files like strigi_font or strigi_trash are way to specific and will
> be
> moved out of the onto and into appropriate analyzers sometime soon.
>
> Otherwise the design ideas/criterias for Strigi onto are the same as
> Nepomuk
> and this leads to a quite predictable result.
>
> > I've looked a bit into these and I think the Strigi one is the one
> closest
> > to what I had in mind. It has received a lot of thought and seems
> generally
> > coherent.
>
> There are several minor corrections I'd like to make, but reluctant to do
> so
> until I have a critical mass of fixes and,preferably, xesam draft since
> this
> also involves analyzer code changes.
>
> Of course, I'll let you know what I'd like to have fixed and why.
>
> Also, I'm prepared to explain any and all design decisions behind
> Strigi/Nepomuk onto.
>
> > It would be ideal to store the ontology drafts in a VCS, but we
> > have none yet - I don't know if the wiki would be disastrous or not...
> Any
> > ideas on a temporary storage idea until we get a subversion module at
> fdo?
>
> Wiki is ok. The only way I see this working is to authorize one person to
> make
> changes to the onto. It's too small and has too many internal dependencies
> to
> allow ad-hoc editing. Changes should be proposed, discussed and only then
> committed.
>
> You still can use diff on wiki article contents and send patches for
> discussion/approval.




Right. The key maintainer would still have to discuss things on the list I
take it?

Anyway, I would like to propose that Evgeny is our ontology-man (if he has
the time (I haven't really talked to him about it)).

To all you other Xesames: I can tell that Evgeny really digs this whole
ontology deal. Atleast better than me :-) He is the one responsible for the
Strigi ontology.  By talking to him on IRC it is clear to me that it is
founded on a lot of thought and solid reasoning. On top of that it seems
that he has insight into the Nepomuk world and I really think we need to be
compatible with them.

If you are at all in doubt of his technical merits I bet you can ping him on
#xesam at FreeNode...

Cheers,
Mikkel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freedesktop.org/archives/xdg/attachments/20070512/cf4f74a7/attachment.htm