Encoding bug when compiling

Scrivener Scrivener for Windows
,
1

Context: Until recently, I had exclusively used the old native version of Linux on Linux Mint 22. I started using a Mac Mini and I have been testing its compatibility with the Windows version. Maybe because of the legacy metadata or some other version conflict, my experience when writing between a Mac device and another computer with Windows 11 (trial period before buying it) has been a disaster.

In this thread I highlight a work-breaking bug on the latest version of Scrivener of Windows: when I compile my manuscript to PDF, some chapters (independent documents) appear with incorrect coding (strange characters). This seems to happen randomly between the documents. If it is re-exported, other chapters may appear right or wrong.

I do have observed a pattern, it seems that only text contents may be corrupted; the headings never corrupt. The text is set “as-is” in the editor. The section type assigned to the document seems to be irrelevant.

2

Might be me, but I can’t tell what software you are running. Specifically, what versions of all software (Linux, Windows, Mac, Scrivener, …) are running to produce this error?

3

On the latest version of Scrivener for Windows (3.1.5.1). After migrating my project from the old format, I was creating build profiles on Mac/Windows and this bug occurs only on Windows.

4

Does this happen with another output format, for instance .rtf?

Or this could be a cross-platform font issue. L&L has an article on this topic in their Knowledgebase. Ensure that the Mac font(s) in use have been installed on Windows. See if changing the font(s) on both platforms makes a difference, particularly if the font(s) you are currently using are out of the ordinary. For example, try something vanilla like TNR on both Mac & Windows.

2 Likes