I like Chris' suggestion, too. Seems coherent and easy to remember and apply.
+1 Rubén Pérez TELTEK Video Research www.teltek.es 2012/9/25 Greg Logan <[email protected]> > +1 to unifying, and to be honest I like Chris' suggestion most. I don't > particularly care whether we go with an experimental or a registered > one, but I think at style of URN will make it very clear that the > namespaces do not resolve to anything. > > G > > On 12-09-25 11:14 AM, Christopher Brooks wrote: > > On Tue, 25 Sep 2012 19:07:17 +0200 > > Tobias Wunden <[email protected]> wrote: > > > >> Looking through the code base one can see that we now have XML > >> namespaces where needed, which is good. However, it turns out that we > >> have not been too consistent with regards to how we create these > >> namespaces. Some of them are prefixed with http://, others aren't. > >> > >> Therefore, the first part of the proposal is to unify the two > >> approaches and I hope to get a +1 from everyone on this. > > > > +1 > > > >> The second part is to agree on the namespace naming scheme we use. On > >> the table is the choice between going with a prefix of http:// vs. > >> going without it. Technically speaking, the namespace should be a > >> URI, which both URN (name) and URL (location) comply with and there > >> is no requirement on the prefix, so > >> > >> engage.opencastproject.org > >> > >> is as good as > >> > >> http://engage.opencastproject.org > >> > >> with the slight difference that using the http:// somehow implies > >> being able to visit that location and get something meaningful, which > >> is not the case with the Matterhorn namespaces. I am looking for > >> reasons to go with one or the other, but to make voting easier, I > >> will propose to go *without* http://. > > > > I support without, but would like it to not just be unique but a URN > > instead. The syntax for a urn is: > > > > urn:<NID>:<NSS> > > > > The namespace id (NID) must be registered with IANA, but I believe this > > is free: > > > > http://tools.ietf.org/html/rfc3406#appendix-A > > > > So a completed unr would look like: > > > > urn:opencast:engage > > > > We can use, without registration, the experimental URNs. No guarantee > > of being collision free, but it's unlikely to cause a collision. That > > would look like: > > > > urn:X-opencast:engage > > > > I am fine with either of these. > > > > If we are using http then it should resolve to something, at least a > > wiki page. > > > > Chris > > > > > > _______________________________________________ > Matterhorn mailing list > [email protected] > http://lists.opencastproject.org/mailman/listinfo/matterhorn > > > To unsubscribe please email > [email protected] > _______________________________________________ >
_______________________________________________ Matterhorn mailing list [email protected] http://lists.opencastproject.org/mailman/listinfo/matterhorn To unsubscribe please email [email protected] _______________________________________________
