[standards-jig] message formatting (XHTML IM)

Julian Missig julian at jabber.org
Thu Jun 19 18:26:24 UTC 2003

On Thursday, Jun 19, 2003, at 11:15 US/Eastern, David 'TheRaven' 
Chisnall wrote:

> I have a few thoughts, some of which may be sensible, some of which 
> may result from lack of sleep:
> 1)  I really don't like the idea of parsing /italic/, because this is 
> often used in paths and URLs that people paste into chat dialogs and 
> having a client randomly italicise parts of a URLwould be irritating. 
> These forms of mark-up are often used in the geek community, but not 
> so common outside it.  For the people who use it the client parsing it 
> would be unrequired, for those not using it it would be confusing. 
> However, this is not a protocol issue.

Easy way around that: only parse it if it's /word/, no /'s in the 
middle, nothing on the sides of the slashes. Also, Mozilla Mail parses 
_word_, *word*, and /word/... and it leaves the _'s, *'s, and /'s in. 
That's another solution.

> 2) I'm not sure what the advantage of being able to enter your own 
> markup is to most people.  MSN etc. don't let you do this, they all 
> give you a wordpad style interface.  There's nothing to stop client 
> having keyboard shortcuts (ctrl-i for toggle italic, for example), for 
> users who prefer to use a keyboard, but the majority of IM users have 
> grown up with MS Word and derivatives and expect that kind of UI for 
> entering rich text.

I know I don't consider "being able to enter your own markup" an 
advantage in this discussion. No matter which markup we use, clients 
will have the keyboard shortcuts and buttons.

> 3) The client I currently use (JAJC) supports HTML IM very well, and 
> has a clear UI for doing so.  With the exception of the MSN transport 
> not correctly parsing it, it seems to work fine, and does everything I 
> would expect of an IM system.

That's nice. Is it correctly using the currently XHTML JEP?

> 4) MathML.  Hmm.  While it would be very useful to be able to enter 
> mathematical formulae in an IM (I usually fall back to entering LaTeX 
> syntax and relying on the other party to parse it in their head, which 
> is far vrom ideal) my view on MathML is that is a demon-spawned 
> hell-child.  It is not human-readable (the spec even states something 
> to this effect) and if you aren't going to be human-readable then this 
> somewhat defeats the point of using XML.  I have yet to find a good 
> program for creating it, so I suspect that it would require a major 
> investment of effort on the part of client developers to implement.

Well, MathML will be a different beast in a different JEP. I imagine 
the only client developers who are going to implement it are either a) 
insane, or b) have a library/interface from something else that they 
can use.


More information about the Standards mailing list