[otr-users] Pidgin plugin sends and parses HTML

[Ian Goldberg]

On Tue, May 13, 2008 at 01:01:55AM +1000, Scott Ellis wrote:
> >
> > Uhm. I can only find one place where it mentiones HTML at all. And while
> > it
> > mentions that it may contain markup, it still doesn't qualify as allowing
> > to
> > put HTML into a place where only text/plain is allowed. Of course the text
> > to encrypt may contain HTML, if an HTML message is about to be sent. Just
> > as
> > it may contain rtf, M$ .doc or any other markup if that is what is to be
> > sent. But the data type of the data to be encrypted can only be determined
> > by the underlying protocol, otherwise an extensive chapter on integration
> > would HAVE to be part of the spec. It isn't.
> 
> 
> The actual text transferred over the underlying protocol is made up of
> plaintext chars - and as such none of the rules of the underlying protocol
> are being broken. Even jabber XEPs cannot lay claim to the *meaning* of
> plaintext within messages - just as you and a friend are not prevented from
> using some code language you make up yourselves over jabber. Under this
> interpretation the unencrpyted messages of OTR conversations have nothing to
> do with the transport protocol. The phrase that was used by the developers
> in my earlier conversations on this topic was 'higher level protocol'. It's
> ugly and inconvenient to most of us, but it does make sense from a certain
> point of view.
> 
> It claims that using libOTR is
> > as simple as replacing the plain text with the output of the function.
> 
> 
> You're very right there - in most cases it doesn't perform 'as advertised'.
> But it does work that way for a lot of clients - almost anything Qt or Java
> based, for example.
> 
> Can I suggest this discussion continue on the dev mailing list though?

Agreed.  I'll start a thread over there.

   - Ian