I was testing input boundaries in my client (deedum), and I realized that I wasn't counting the full length of the uri. I believe the text of the link on 3.2.1.gmi could be improved in this regard.
The relevant spec is:
URL is a UTF-8 encoded absolute URL, including a scheme, of maximum length 1024 bytes. The request MUST NOT begin with a U+FEFF byte order mark.
As I fixed that, and was testing the boundary conditions, I'm pretty sure that requests with a url of length 1023 and 1024 are incorrectly rejected with a 59 error code. I double checked and I am not sending the byte order mark, so I'm guessing your server may be counting the CRLF that follows the URL.
I was testing input boundaries in my client (deedum), and I realized that I wasn't counting the full length of the uri. I believe the text of the link on 3.2.1.gmi could be improved in this regard.
The relevant spec is:
As I fixed that, and was testing the boundary conditions, I'm pretty sure that requests with a url of length 1023 and 1024 are incorrectly rejected with a
59error code. I double checked and I am not sending the byte order mark, so I'm guessing your server may be counting the CRLF that follows the URL.