https://bugs.documentfoundation.org/show_bug.cgi?id=32363
--- Comment #47 from László Németh <[email protected]> --- (In reply to RGB from comment #45) > Built-in workarounds are still workarounds, specially when they are "second > order built-in workarounds" (this one uses the previously introduced "inline > headings" workaround)... Anyway, I give up. Yes, I agree, that we have still workarounds. WYSIWYG applications don't follow the SGML logic, that is why is very hard to map a structured description (https://en.wikipedia.org/wiki/LaTeX#Typesetting_system) to a WYSIWYG application. allotropia started to implement loext: style separators (still incomplete), while I've added a workaround to solve the ToC/PDF ToC issues with ODF-compliant inline frames (also I've added the first PDF ToC unit tests to LibreOffice), which seemed to me broken by design: inline frames are div-like objects, i.e. they cannot be multi-line without breaking the in-line heading – but I realized, this is not true for the DIV implementations of the browser, that is why the (X)HTML export of the inline frame-based workaround for the style separators is OK now... We simply can't solve everything at once, especially if we don't know all the problems, and especially not in all depth. (See my previous comment about the different needs for the PDF ToC, not too mention the interoperability requirements, which are still the main driving force for our developments.) For me, this bug report was a great help in understanding the problem better. The next step would be to thoroughly test the proposed solutions to find and file new bugs. It's OK to ask for more LaTeX-like solutions, too. (I thought of mapping the short titles to RDF metadata of the headings, which is the ODF-compliant way of metadata annotation. Later we could define canonical RDF IDs, like for field prefix and suffix, and extend the document layout/fields to handle them. The only problem with this “clean” solution, that LibreOffice is a WYSIWYG editor, which allows much more, than LaTeX, because we always know the layout. Using *only* recent LaTeX solution for the short titles, you never know, that shortening will really work or not. You need a LaTeX processor, too, and fix your LaTeX again and again based on the actual layout, which work flow is not ideal sometimes.) -- You are receiving this mail because: You are the assignee for the bug.
