Robin Fairbairns <[log in to unmask]> on Sat, 17 Oct 1998 at 11:49:11 +0100 wrote (responding primarily to Carlisle (and Aberg)): > latex's syntax has a little bit of regularity, a little bit of block > structure, ..., but all told it's so uneven that it's silly to imagine > `formally' specifying it. > added to which, the extreme difficult of faulting tex primitives, > etc., that fall outside the scope of the syntax makes the utility of > such a specification doubtful. Comment: I do imagine that LaTeX can be modeled as something like a "categorical limit" of SGML DTD's. If I want to be able to have robust translations of the documents that I author to other formats, then I would choose one of those DTD's. (Well, I might add a few other goodies to it that flow to LaTeX.) But my original question was motivated by several lines from "ltx3info.tex" and my wondering whether these lines were durable for the LaTeX-3 future. Here are clips of a "newenvironment" and a "newcommand": %%%%% clip \newenvironment{citations}{% \list{}{% \renewcommand{\makelabel}[1]{\normalfont\itshape ##1}% }% }{% \endlist } %%%%% clip \newcommand{\AmSLaTeX}{$\mathcal A$\lower.4ex\hbox{$\!\mathcal M\!$}$\mathcal S$-\LaTeX} %%%%% endclips (I do not see why "citations" could not have been done entirely in Lamport LaTeX. (I am unclear what "list" and "endlist" are.)) In the "AmSLaTeX" macro I believe that "lower" and "hbox" are ordinary TeX. Will this usage survive? Thanks. -- Bill Hammond