i30817
commented
11 months ago Thanks for notifying me. I don't have a computer at this time (this is a android tablet), so a fix will take a while, possibly too long since I'm broke. You can avoid the problem with renaming the affected files and rescanning the playlist, if it was made with the manual scanner... the reason is that by changing the filename, you disconnected the playlist entry from the actual file, so it won't start if you try to load it from RetroArch because it can't find the file. You can also only change the playlist game label, but that breaks whenever you rescan instead.
It should be unusual, most games .dats will use utf-8 I think, and so most game files. Since libretro doesn't have thumbnails for demos for those two you probably won't lose anything. The default server the program uses for thumbnails are the Png files at https://thumbnails.libretro.com although you can change that in the options.... as long as it uses the same subdirectory name structure and png thumbnails.
I actually find it strange libretro is even putting utf 16 on playlists, probably something they do to be able to reference a windows file with utf16 filename. They probably should only allow it in the filename not the label (replace the characters with no equivalent by the utf8 missing character symbol) since it will display wrong for them too, most likely. The program only uses the label to find the pngs, because it's more flexible (some ways to build playlists with the manual scanner or other utilities don't have names taken from the filename to have better names, for instance using mame .dat files).
Feathered-Serpent
Hey there,
found that little program and wanted to give it a try. Though after testing some, it doesn't seem to be able to handle UTF-16 characters in a playlist. After trying it a few times, I finally found a message, that helped me finding the problem:
The file is named "Chojn╟ⁿw '94 Compo Music Pack #1 (1994)(Altus).d64" I don't know if the symbols are an error or not. Maybe the program should catch the UnicodeDecodeError and skip that one file instead of stopping.
After I renamed the files (just replacing ⁿ with a _; where a few dozen) I ran the program again, only to find another file name with a character it doesn't like. This time the filename was "PETSCÏÏhead (2020-01-13)(Seven).prg" I renamed the Ï to I and seemingly after that it started downloading thumbnails.
Windows 11 Pro; Python 3.12.0 64bit.