tanteDoes anyone know if a Mastodon server can block client applications based on their name?
Janik Besendorf@tante of course it can or are you asking if this is already implemented in the code?
@besendorf I wonder if that is already available in the configuration UI
Stefan
elsa sloth-ringen 🔜🚫GPN :(@tante for your own users yes (can be easily done with a web server rule), for someone else's instance's users no
Daniël Franke 
@tante Asking for a certain seldo creation?
@ainmosni yeah. Since it runs in people's browser's IP blocking or headers don't work, but it registers as a specific Application name with a mastodon server which could be blocked.
@tante the problem there is that from another instance you could potentially only block posts that were made by that instance, afaik, you can't see if an app is registered in an account at all.
Mx. Eddie R@tante
Doesn't currently exist, and might not be possible -- it would require the bad actor's instance to cooperate as well as your own. People like Seldo (this week's LLM bro) just wouldn't play nice.
Doesn't currently exist, and might not be possible -- it would require the bad actor's instance to cooperate as well as your own. People like Seldo (this week's LLM bro) just wouldn't play nice.
@silvermoon82 tha code uses "Zetgeist" as application name which one could filter. If they start messing around with that thats a whole different story, then he can no longer claim to be a decent citizen
Preslav Rachev
Uriel FanelligTs can block based on user-agent. Otherwise, you need some nginx in front, to process headers.
Thomas Cloer@tante ChatGPT sez:
"Blocking purely by name is inherently weak because:
* Client names and User-Agents are self-declared and easily spoofed
* A malicious client can just rename itself"
(taken from a much longer answer)
Lars Marowsky-Brée 😷
Tekniquelly correct@tante Sure, you can block them at the webserver layer like: https://www.xmodulo.com/block-specific-user-agents-nginx-web-server.html
However, if their scraper lies about the user-agent, there's not much you can do.
@tante There's only so much you can see there, though. And a malicious client could change their agent to be something like "Mastodon/4.5.7 (http.rb/5.3.1 +https://mastodon.soc/)" and they'd look just like another Mastodon server talking to you.
C.Suthorn 
@[email protected]
I don't know, but it wouldn't make much sense? Only on servers with local-only posts—for example, Mastodon Glitch servers, but not regular Mastodon servers. And on servers that aren't federated. On all other servers, a post could simply be shared via federation to another server and then read on that other server using any client.