The /s/<id> urls are nice but they aren't direct links to file. It would be really useful if it was possible to reference access single files in a similar manner. Maybe something like /f/<short id>/[optional file name]. The current direct links are very long and use UUIDv4 which is not the most friendly thing to type on a keyboard.
🎤 Pitch
I often need to transfer single files between virtual machines and often these virtual machines don't have an easy way to copy and paste a long url. They also often lack a user interface or only have access to very old browsers. Having a short, self-hosted, direct download url that auto expires would be supper useful.
Some tools (like wget) will by default ignore the Content-Disposition header meaning that the name of the downloaded file will just be whatever was after the last /. Having the ability for a user to put anything there would make it trivial to download the file with the desired file name without needing an additional rename step. Some applications also make assumptions about the file type based on the file extension they see in the url rather than the Content-Type header.
The file name should of course be optional to allow for nice looking, easy to type short urls.
W1BTR
commented
2 months ago
🔖 Feature description
The
/s/<id>urls are nice but they aren't direct links to file. It would be really useful if it was possible to reference access single files in a similar manner. Maybe something like/f/<short id>/[optional file name]. The current direct links are very long and use UUIDv4 which is not the most friendly thing to type on a keyboard.🎤 Pitch
I often need to transfer single files between virtual machines and often these virtual machines don't have an easy way to copy and paste a long url. They also often lack a user interface or only have access to very old browsers. Having a short, self-hosted, direct download url that auto expires would be supper useful.
Some tools (like wget) will by default ignore the
Content-Dispositionheader meaning that the name of the downloaded file will just be whatever was after the last/. Having the ability for a user to put anything there would make it trivial to download the file with the desired file name without needing an additional rename step. Some applications also make assumptions about the file type based on the file extension they see in the url rather than theContent-Typeheader.The file name should of course be optional to allow for nice looking, easy to type short urls.