What would you like to see in a CLI booru downloader?

I am developing a CLI program (which I eventually plan to communicate with a chrome/firefox extension) but the gist is it can bulk download based on tags and download single post.

It will also have custom file/directory structuring based on a config file.

What features would make you use it / would be cool to have?

Curious but how would it differ from mikf/gallery-dl which has multi-site (including other boorus) support?

WRS said:

Curious but how would it differ from mikf/gallery-dl which has multi-site (including other boorus) support?

It would be very similar, but I already have a working prototype written in Zig with multithreading, focusing on speed and efficiency from the start. I also plan to integrate it with an extension so you can easily use it from the browser with the same library location.

There was also an idea of having a background process version that can automatically download all new images for a given query, though that would not be my current focus.

FYI going multi-threading might have worked a decade ago, but nowadays everyone (including us) will think you're an AI scraper and ban your IP the moment you do that.

nonamethanks said:

FYI going multi-threading might have worked a decade ago, but nowadays everyone (including us) will think you're an AI scraper and ban your IP the moment you do that.

even with proper rate limiting blocks (thread sleeps)?

SageEvan said:

even with proper rate limiting blocks (thread sleeps)?

At that point there's no benefit to using multithreading. It might be useful to have in your architecture for sites that do benefit from it though.