Re: [office] Critical problem

["Florian Reuter" <freuter@novell.com>]

Actually Latex is a little bit different.

The \begin{enumerate} \end{enumerate} and \section are just abbreviations.

What happens behind is that counters are defined using the
\newcounter{sectioncounter}
\setcounter{sectioncounter}{1}

and then the counter can be displayed using e.g. \arabic{sectionscounter} or \roman{sectioncounter}.

So the \section command is implemented like
\newcommand{section}{1}  {
\arabic{#1}
\addcounter{sectioncounter}{1}
}

So this is really a very different concept.

~Florian



>>> Bruce D'Arcus <bruce.darcus@OpenDocument.us> 03/13/07 2:36 PM >>>

I really don't have the time to get drawn into this, and I'm not a list 
person, but ...

On Mar 13, 2007, at 6:42 AM, David Faure wrote:

> Personally I don't have enough experience with <list> to be able to 
> answer that.
> You do, so you probably know the answer: if you know for sure that 
> there is no way
> to model this with <list>, then please say so, and I will simply 
> conclude that:
> * OK, we have two different ways to model lists in OpenDocument
> * the numbered-paragraph way is more flexible - and this is why we 
> need it.
> (And personally I could do with that way only, i.e. without <list>).

... it seems to me this makes sense, and that if there's a problem 
mapping between the two:

a) perhaps then text:list is the problem
b) in any case, am not sure we have some formal (or even practical?) 
requirement that we must be able to map between the two?

Superficially, I prefer the list-id approach because it involves 
identifying a conceptual object of sorts, rather than just using some 
essentially presentational attributes (continue-numbering).

BTW, re: LaTeX, the thing there is that lists can span across other 
structures (sections and paragraphs).

\begin{enumerate}
\item one
\item two

\section{Foo}

\end{enumerate}

Bruce