#social 2019-06-22
2019-06-22 UTC
xmpp-social, ajordan and vitalyster joined the channel; vitalyster left the channel
# 
up201705417 Would it be okay if we added a 'preferredWebfingerEndpoint' property to the Actor activity in GNU social?
# up201705417 I was thinking about it as it would allow instances to attempt the preferredWebfingerEndpoint in case the Actor URI was offline...
# fr33domlover up201705417, you can always add extension properties, and whoever doesn't know them just won't use them. No harm in that. However if the same web app / same server serves the actor and the webfinger, it's not very likely one will be online and the other offline. Is that sort of thing likely for GNU Social? Also side note, Actor isn't an activity :)
# fr33domlover (But is definitely a JSON-LD object)
# up201705417 fr33domlover: By default that will be the case (if one is offline the other will be), this would be a functionality
# up201705417 if the user defines a custom webfinger endpoint
# up201705417 it would allow things like setting a preferred Webfinger endpoint as in diogo.site pointing to loadaverage
# up201705417 then, if loadaverage goes offline, I could change my webfinger in diogo.site to point to another instance/account
# up201705417 it would be something to prevent missing any notices while main instance is offline
# fr33domlover up201705417, I'm slightly confused, however I wonder if delivery retry already cover this. If GNU Social tries to deliver a message and the recipient is unreachable, does it give up or will it keep trying to deliver once in a while?
# up201705417 well, it was just a thought, but yeah, it doesn't look like a good idea
# up201705417 >does it give up or will it keep trying to deliver once in a while?
# up201705417 right now it is just doing best efforts, we still have to implement that