Jean-Marc Lasgouttes wrote: > > Your diagnosis and suggested fix are right. IIRC we decided with JMarc > > it was not worth going recursively into insets for toc purpose in > > order to avoid potential visual problems (like big math insets or > > branch, etc). So maybe we should add something between 'no inset' and > > all insets'. I remember an old commit from JMarc that distinguish > > between general insets and formating inset (like spaces, etc), maybe > > that can be leveraged here. > > I think that each inset should decide what it does depending on the > flags that are passed to it. I do not think that a 'no inset' policy > ever makes sense.
Sounds sensible. Anyway, I'll leave this to you. Jürgen
