LATEX-L@LISTSERV.UNI-HEIDELBERG.DE

 Options: Use Forum View Use Monospaced Font Show Text Part by Default Condense Mail Headers Message: [<< First] [< Prev] [Next >] [Last >>] Topic: [<< First] [< Prev] [Next >] [Last >>] Author: [<< First] [< Prev] [Next >] [Last >>]

```Michael J. Downes wrote:
>>   But another common module separator would be :, but that is already
>> occupied. One could change the argument spec separator to ., so that names
>> soul look like \foo:bar.n or something.
>
>Then TeX dimens of the form 0.1pt in the source code would not scan
>correctly (the "." has to have catcode 12).
>
>These issues were already hashed out at great length by the LaTeX team
>a long time ago. Don't forget that TeX is not like other programming
>languages---for example, TeX has no notion of quoted strings, so the
>right quote character ' could also be a candidate character for use in
>control sequence names, unlike C or Pascal identifiers (... the octal
>number notation is used so seldom in practice that a function
>\octal{321} would easily serve).
>
>In my opinion, there has been so far too much discussion and not
>enough substantial work with the published expl3 proposal. I do not
>say it is flawless, but I do say it is a waste of time to rehash it
>endlessly without any real work being done.  When someone can post to
>CTAN a working package of nontrivial size using the current proposed
>conventions, then I will be ready to listen to their opinions with
>respect.

So, if I was allowed to interfere, I would suggest that the names should
be named \tex/foo_bar:<argspec>, with "_" a word separator and "/" a module
separator. This way the L3PL team can forget about higher module
abstractions for now, but on the same time ensuring that the L3PL does not
block that the modules concept being introduced fully at some later point.

If "/" does not prove to be suitable, then it can easily be changed to
something else at some later point.

Hans Aberg