Subject: Re: Alternative *ML syntaxes [was: Re: StudlyCaps ]
From: Erik Naggum <erik@naggum.no>
Date: 18 Nov 2002 18:08:50 +0000
Newsgroups: comp.lang.lisp
Message-ID: <3246631730734601@naggum.no>

* Rob Warnock
| Perhaps in the case of HTML there are no conflicts of this type (I
| haven't checked thoroughly), but if one tried to generate output into
| somebody's arbitrary XML form, say, where they hadn't defined the DTD
| with that conflict in mind...

  I would strongly urge you to please find an actual conflict before even
  considering the problem.  Where there is a conflict, solve it locally.
  This does not need a global solution or even any problemizing because it
  happens extremely rarely and when it does, you can deal with it in the
  mapping.  This is, after all, an editing tool.  A little human effort to
  take care of problem that never happens is better than a lot of human
  effort to ensure a problem that never happens /could/ never happen.

| (Or does HTML and/or XML already require that the tag & attribute names
| be disjoint?)

   Attribute names are local to an element (please note the terminology),
   but element names are global.

| That's why I somewhat prefer \foo{...} to {foo ...}, since I tend to
| use curlies in writing plain text (e.g., sample C code), and almost
| never use "\". (MS users MMV.)

  The whole point here is to make the syntax more navigable with Emacs, not
  less so than *ML.

| TeX has shown that \entity and \func{...} can coexist, has it not?

  Yes, by attaching magic meaning to whitespace.  I want unescaped { and }
  to be markup, unconditionally.

  I believe we have different goals with the syntax.  Besides, I want to
  clean up the fantastically ugly mess that is TeX, too, not mimic it.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.