"you can't break what's already broken."
Microsoft, Qualcomm and Lotus, vendors of such widespread email/groupware packages as Outlook, Eudora Pro or Notes, have co-authored a proposal to improve threading of email messages using HTML and XML (Extensible Markup Language, one abstraction level above HTML).
How does it work?
In an email exchange following the proposal's ideas every piece of text is associated with the message it originally appeared in. This is done with the message ID every email already must have. This way the text can always be attributed to a certain sender and a certain time, allowing to sort the individual quoted parts within one message with respect to their time or to reply to the author of an arbitrary part of the message.
Additionally to the message ID the text parts are attributed with a style sheet. This allows for laying out the text on a per-author basis. For example, everything in a message written by William would be in huge, red, bold and italic letters.
Additionally (yes!), the message (a message with quoted text really consists of multiple messages) and its authors are described in XML blocks. The message properties include such essential things as the author's email address, her name, the subject,... The properties describing the author can provide a URL associated with her, her phone number, a style sheet class name etc.
Take this dull conversation:
Quotant Jimbo: > Quoting Jamba: > > > Quotant Jimbo: > > > > > Quoting Jumbo: > > > > > > > Jimbo said: > > > > > > > > > Hi! > > > > > > > > Hi. > > > > > > I love you! > > > > Who? > > Me. You. Huh?
It could look like that:
Quotant Jimbo:Quoting Jamba:Me. You.Quotant Jimbo:Who?Quoting Jumbo:I love you!Jimbo said:Hi.
But this is just one possible way of displaying this thread and certainly not a particularly good one.
No more >?
Apparently, the authors of "HTML Threading" thought something had to be fixed. It is also apparent that they wanted to make sure that HTMail will become a standard. They have explained their motivation and their dislike of > in the proposal itself (this is probably the most interesting part of it). They say:
"It [the >] conveys only hierarchical information."
Very true and in some rare cases even a problem.
"It may be hard for the user to infer, and impossible for a messaging UA to determine, the author of that line of text."
Yes indeed. That's why an alternative way of quoting that is entirely based on plain text has been thought out. It addresses this problem quite well, but is not 'standard' and thus of limited use.
"And when the text is quoted multiple times [...] lines which are hard-wrapped can overflow onto the next line, resulting in awkward, hard to read messages."
This is not a problem of the > quoting approach but of the appalling incompetence of email clients when it comes to being an editor. Maybe their programmers should take a look at Par to see how paragraph reformatting can be done or, better yet, allow filtering text through it.
The document also claims that it should be easy to read a HTML-threadified email with a 'traditional' mail reader. I am sure this was achieved, but I have no idea how.