Setting up a file server can be intimidating to the uninitiated. There are many servers to choose from, and then you need to decide how to install it — Docker? Kubernates? Well, what’s …
I’ve literally never felt like I needed a file server to be easier and support more protocols, and this seems like it’s trying to do way too much at once. HTTP is beautiful and convenient but as a bespoke javascript-heavy API it’s really not a particularly great way to manage files, FTP sucks and if you still need it for something you need to re-evaluate your life choices, TFTP is useful in extremely niche applications that I wouldn’t want to hook my entire file server up to anyway and certainly don’t want running along side these other options, WebDav is fine but again really only necessary in niche applications which you don’t need or want your entire file system hooked up to (or if you don’t know how to VPN) and this doesn’t seem to support SFTP/SSHFS which is what I would consider a modern standard for a secure file transfer protocol.
Just use Samba and/or ssh, every Linux distribution comes with packages for them. Both are widely used and battle-hardened, and between the two they are compatible with almost everything. You don’t really need all that other stuff.
In some ways it’s the inverse of the UNIX Philosophy: instead of doing one thing perfectly, this program is doing everything [9001] could think of, and doing it “good enough”.
As a believer in the UNIX philosophy (armed with an understanding of why it is fundamentally useful) it horrifies me that either developers or users think this is a good thing.
I’m not sure if this would fall under the definition of “skill issue”, but I’ve never had any luck getting Samba to work on even a local network.
This may be the same argument at “immutable” distros, having an environment where the user cannot fuck it up will be the best way to spur adoption rates, and eventually ambitious users will use more granular methods like you described.
I’ve literally never felt like I needed a file server to be easier and support more protocols, and this seems like it’s trying to do way too much at once. HTTP is beautiful and convenient but as a bespoke javascript-heavy API it’s really not a particularly great way to manage files, FTP sucks and if you still need it for something you need to re-evaluate your life choices, TFTP is useful in extremely niche applications that I wouldn’t want to hook my entire file server up to anyway and certainly don’t want running along side these other options, WebDav is fine but again really only necessary in niche applications which you don’t need or want your entire file system hooked up to (or if you don’t know how to VPN) and this doesn’t seem to support SFTP/SSHFS which is what I would consider a modern standard for a secure file transfer protocol.
Just use Samba and/or ssh, every Linux distribution comes with packages for them. Both are widely used and battle-hardened, and between the two they are compatible with almost everything. You don’t really need all that other stuff.
As a believer in the UNIX philosophy (armed with an understanding of why it is fundamentally useful) it horrifies me that either developers or users think this is a good thing.
I’m not sure if this would fall under the definition of “skill issue”, but I’ve never had any luck getting Samba to work on even a local network.
This may be the same argument at “immutable” distros, having an environment where the user cannot fuck it up will be the best way to spur adoption rates, and eventually ambitious users will use more granular methods like you described.
I probably will use this.