ietf-822
[Top] [All Lists]

Re: SGML, RichText, etc.

1993-01-21 08:03:13
approach, especially when one considers that real font names are not 
constrained
to 7-bit ascii content.  Instead I considered useing a 

       <X-MacFONT> 12345 "Sonata" </X-MacFONT>... 
...
I would hope that any reworked version of RT would establish more definate
semantics, and possibly even punt <Smaller><Bigger> for explicit sizes (ie,
<Size-999> ... </Size-999>).

I'm seeing a real difference in the desires for what "richtext" (or
whatever we are going to call it) is expected to do for us here.

One contingent, in which I would count myself, wants this thing to be
quite lightweight and *very* device and facilities independent in the
markup (if not going all the way to the generic approach of SGML).  I
don't want to see anything more specific than <bold> and <smaller>, and
I want the display semantics associated with those to be explicitly left
up to the UA implementation, with no standardized guidance and as little
folklore guidance as possible.

This doesn't mean that things like the <smaller><smaller> abc </smaller>
xxx </smaller> issues don't need to be clarified: clearly they do.

The other contingent seems to be looking for something in which it is
possible to express most, if not all, of the specifics of a moderate to 
high-quality formatting markup language of the TeX, MS-RTF, IBM-RFT,
troff, etc., varieties, presumably with the assumption that display
engines will interpret the font and size specifications into "something
close".

I'm opposed to that.  I think those sorts of things are extremely
good--and appropriate--candidates for applications types.  But Richtext
is going to fail in [what I've thought of as] its primary goal of
permitting slightly more fancy email on an interoperable basis unless
almost everyone can support what almost everyone else is putting out.

If there is disagreement on that point, can we talk about it, rather
than getting bogged down in syntax?
   
   john

<Prev in Thread] Current Thread [Next in Thread>