A setting to overwrite files/bulk optimization? about optivorbis

[murlakatamenka]

@abbuw you can leverage other utilities to make your CPU cores busy with optimizing a folder, such as fd:

fd . --type file --extension ogg "/path/to/folder" -x optivorbis {} {.}_optimized.ogg

After that, you can verify that optimized files are valid and remove the original ones, then strip _optimized from filenames. Quick and dirty, take it or leave it 😄

[murlakatamenka]

Yes, GNU parallel is another alternative.

@abbuw didn't specify the platform he uses, so I've written about cross-platform fd only. parallel is via MSYS2 or WSL on Windows as I understand.

[AlexTMjugador]

Hi! 👋

In-place optimization is a bit tricky for OptiVorbis. Its two-pass optimization techniques require the input file to be available while the output file is generated, so a temporary output file must be used and atomically moved to the original path after optimizations. Getting this done right across platforms requires attention to detail, so it was not in scope for the first public release.

On the dragging and dropping part, while I understand that it can be a significant UX improvement, at a technical level that gesture is just another way of passing command line parameters, and I'm unsure whether I want to change the CLI interface to accommodate that use case. For example, if two files are dragged and dropped, two paths are passed to the CLI, so what should OptiVorbis do: optimize the file referenced by the first path to the second path (the current behavior), or optimize both files in place? OptiVorbis does not know if the user dragged and dropped files or used a command-line prompt, so it's impossible to guess what the user wants from the paths alone.

Anyway, your suggestions are intriguing, and even if they do not make it to the application, I welcome user-contributed ways to make these operations easier! Thank you for your feedback 😄

[AlexTMjugador]

Thanks for the tip, @murlakatamenka! I've used tools like parallel several times myself to execute commands in parallel across file sets, but I didn't think of it when I wrote that reply 😄