I am using chemnum package and since last updates of l3packages, l3kernel and l3experimental when I compile my code there are 14 errors. The first one is:

LaTeX error: "kernel/command-not-defined"

Control sequence \chemnum_aux_file undefined

Despite this error the code is working and it gives me a good output, but this error should mean something:

\documentclass[a4paper, 10pt]{report} %environnement du document
\usepackage[T1]{fontenc} %prise en charge des caractères spéciaux
\usepackage[frenchb]{babel} %langue
\usepackage{graphicx} %prise en charge des graphiques eps
\geometry{margin=55pt} %marges
\usepackage{indentfirst} %indentation après le titre de section
\usepackage[bottom]{footmisc} %bas de page en toute fin de page
\usepackage{csquotes} %pour la biblio
\usepackage[backend=biber,style=chem-acs,subentry,maxnames=20]{biblatex} %style acs
\usepackage{chemnum,chemstyle,chemscheme}%pour les schémas de chimie

\usepackage[version=3]{mhchem}  %prise en charge de l'écriture chimique



\begin{scheme}[H] %le H permet de fixer l'image ici
        \label{13} %permet de faire reference dans le texte


Is there a way to fix it?

  • A number of things appear to have been broken by recent updates to l3kernel and l3packages. For example, I (and others) are finding that unicode-math no longer works as it should. – Evan Cooch Jul 25 '12 at 15:10

I already fixed that and sent an update to CTAN. Version 0.5c is already available there.

| improve this answer | |
  • Ok, I will update teh package – Thomas C Jul 25 '12 at 15:27
  • @cgnider Could I check exactly what broke: we did try to avoid any issues, so I wonder which change in the code caused this. – Joseph Wright Jul 25 '12 at 16:15
  • @JosephWright chemnum writes the compound numbers to an aux file \chemnum_aux_file: and chemnum was missing a \iow_new:N \chemnum_aux_file: so this was rather a bug in chemnum, I guess. – cgnieder Jul 25 '12 at 16:21
  • @JosephWright in another of my packages I had troubles, though, as it was relying on xparse's (now missing) \NoValue to fool some commands to be fed with no value instead of an empty one. I fixed that, too, and probably should have designed the function using it differently in the first place, so ... – cgnieder Jul 25 '12 at 16:27
  • @cgnieder On the first point: OK, that was one of the 'breaking' changes we knew about. On the second point: Part of the reason for the change was that having \NoValue 'accessible' was asking for trouble, as there was the danger of people doing \foo[\NoValue] which was never the intention. So again this is 'to be expected'. I assume you are on LaTeX-L if only to pick up my announcements? – Joseph Wright Jul 25 '12 at 16:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.