Hi @endrebsb,
I just wanted to let you know we’ve fixed the issue internally, and the application now works properly with non-low-ascii paths.
The fix is out now in the latest insider build of Logic 2: Logic 2.4.49 Insider 2
I still need to audit the automation server code, but the bug we just fixed affects both the GUI software and the automation server. Specifically, it’s a bug we introduced a few months ago. The crash was occurring when we tried to handle a UTF-8 string as if it was the current code page. For low-ascii characters, there were no problems. for multi-byte UTF-8 characters, depending on the system code page, there was a good chance that there would not be a crash, but the paths would get corrupted. The path in question is ignored if it doesn’t exist. Some UTF-8 characters are actually invalid characters in the system code page, and these specific characters caused the crash. Interestingly enough, this is pretty rare if you’re using the western European code page. that code base is single byte, and almost every byte represents a valid character, so most UTF-8 would translate to garbage, but valid strings. other code pages - especially multi-byte code pages like the Japanese CP932, were much more likely to interpret the UTF-8 sequence as an invalid character and crash.
Also the crash required a Logic MSO connected.