As an option, deduping should ideally look and the root domain or root URL less any variables after the ? in an URL and treating http the same as https as well as ignoring leading www subdomain.
Duplicates finder algorithm is improved!
Now all garbage from URL like different protocol, WWW, trailing slashes, useless query parameters* and hash strings are ignored!
* like referral id or advert source
@Rustem Mussabekov: Please don’t remove the ability to bookmark different fragments of the same page (the bit after the pound sign #)
Being able to bookmark different sections of the same page can be very useful indeed - e.g. you can directly jump to a particular section of the page from Raindrop or add a section-specific description)