Sounds like you should just do what you need with a private extension, then. On Oct 4, 2012 6:21 AM, "Andreas Kuckartz" <[email protected]> wrote:
> The "why?" is a good question. > > And I have a simple answer: It would be the most simple way to implement > my use case, because I then know that the annotate.js library can be > used (https://github.com/szabyg/annotate.js). > > annotate.js demos are available here, the second one is updated ever 24 > hours: > http://szabyg.github.com/annotate.js/ > http://dev.iks-project.eu:8081/enhancervie > > I will have a look at the implementation changes which would be implied > by using RDF/XML instead of RDFa. But I fear that it might complicate > that work to much. > > Cheers, > Andreas > --- > > Ralph Meijer: > > While I'm sure we could alter XEP-0071 to add support for RDFa, I have > > to wonder why that is desirable. > > > > As I see it, the main purpose of having XEP-0071 was to standardize > > existing efforts to have light *presentational* mark-up for instant > > messages. In practice, a client would mostly use a chat UI element with > > a few helper widgets for adding styles (like bold) and URLs. One > > wouldn't typically write HTML directly. > > > > As an aside, I have personally gone as far as patching Gajim to further > > restrict the allowed elements and styles (mainly because of iChat and > > Adium). > > > > RDFa, on the other hand, would likely be for marking up a message > > *semantically*. Standard practice in XMPP is to just add a new child > > element to the <message/> stanza for that. I.e. as a sibling of <body/> > > and (in this case) <html/>. You could simply add RDF/XML constructs, > > while keeping our restrictions on the use of XML namespace prefixes in > > mind. > > > > For non-IM purposes, I have used embedded Atom Entry documents (as > > Publish-Subscribe payloads), using link elements for denoting triples. > > > > I also have to note that we have traditionally shied away from using > > all-encompassing vocabularies in favor of application-specific ones. > > E.g. XEP-0080 defines its own way to record addresses and geo-location > > information, instead of using an existing gazillion page RFC. :-) > > > > In any case, I'd welcome alternative points of view, of course. > >
