As you probably know, if users can afford Trados, it's usually because they translate for companies, which in turn pay them to keep their corporate confidentiality: we cannot afford (literally and figuratively) to go against these corporate confidentiality matters.

What Trados troubleshooting team should do as a start is to reintroduce the Word format that previously seemed to be among the only workable (though last-resort) solutions against that problem.

Then, to try and repeat the problem (in order to, as you say, “build some robustness around this where needed”), I can only give you general pointers as to what may cause issues in the file I have to translate right now.

Features of the file are as follows:
• DOCX file of about 3 Mb;
• 48-page document with table of contents;
• Includes about 36 figures and tables, including 31 images, plus a large image as a page cover;
• Highlighted copy in several colours throughout, including in the page footers (because the file was first processed in LogiTerm – for end-client purposes – before processing in Trados Studio 2017);
• After creating target file from Trados, I try to open it in Word 2017 – when I keep the table of contents, the Trados target file is unreadable; when I remove the table of content before processing it in Trados, the target file is fine, but I get this message before first opening it (in French, since that is the language of my system and applications), and I just choose the negative option: “Ce document contient des champs qui peuvent faire référence à d'autres fichiers. Voulez-vous mettre à jour les champs de ce document ?”; English equivalent would apparently be: “This document contains links that may refer to other files. Do you want to update this document with the data from the linked files?”

I think it summarizes the main file elements... but the key here is: TABLE OF CONTENT.
Once removed, everything goes fine.

Hi Louis La Bonté 

If only it was as simple as this.  In fact if it was we wouldn't be having this conversation!  Perhaps if anyone has this problem with a file they are willing to share then it would be helpful to have the file.

Here's one I spent a little time trying to use to reproduce the problem you painted:

Louis La Bonte.zip

I have no problems with this at all using the new filetype.

The reality is that there are so many things that could be causing this problem it is not going to be as straightforward as you have suggested.  So I'll just leave this problem open for the time being until we get a file in here, or in support, that allows us to find out what is causing this issue and then fix it.  It's not as widespread as you might think or we would certainly have a lot more complaints than this.

Regards

Paul

Unknown said:

but I get this message before first opening it (in French, since that is the language of my system and applications), and I just choose the negative option: “Ce document contient des champs qui peuvent faire référence à d'autres fichiers. Voulez-vous mettre à jour les champs de ce document ?”; English equivalent would apparently be: “This document contains links that may refer to other files. Do you want to update this document with the data from the linked files?”

Just on this Louis La Bonté this is not a problem.  This is to be expected.  A TOC is generated from the contents of the file, so if you translate it then you are going to have to update the links and regenerate the table.  You don't actually translate the table of contents itself.