Post A9t5PHfJ6OfESgT2um by feonixrift@hackers.town
 (DIR) More posts by feonixrift@hackers.town
 (DIR) Post #A9t5PHfJ6OfESgT2um by feonixrift@hackers.town
       2021-08-01T16:56:31Z
       
       1 likes, 0 repeats
       
       IMHO step zero of shifting focus back from product to protocol is to stop surfacing which software people use in the API.It's done little to encourage/boost alternate software (human communication is better for that) and rapidly becomes a filter for dunking on folks who make different choices without regard for their reasons.
       
 (DIR) Post #A9t5PIBv98kg5pj5kW by feonixrift@hackers.town
       2021-08-01T16:59:59Z
       
       0 likes, 0 repeats
       
       You want to know my capability string? Fine. Ask that. That's relevant. But what client, server, or p2p software I run is nobody's business but mine.
       
 (DIR) Post #A9t5PIeHRhR9VmzjxA by penguin42@mastodon.org.uk
       2021-08-01T17:01:21Z
       
       0 likes, 0 repeats
       
       @feonixrift Providing a place in the protocol for that isn't unreasonable though - as long as it's not required;  it's pretty useful to be able to spot you've got a bug when speaking to $client or $server.
       
 (DIR) Post #A9t5dw0dttFRdnQuES by feonixrift@hackers.town
       2021-08-01T17:03:58Z
       
       0 likes, 0 repeats
       
       @penguin42 Yeah there are uses, definitely. That's a good one. I'm being slightly hyperbolic but I honestly believe the downsides may outweigh the advantages, when considering something that is already a human to human communication protocol, ie. where one-off cases you can actually ask the human on the other end what they run.