@blaine A slight problem with federated micro-Twitters versus fully decentralized IndieWeb setups in a way.
Mastodon instances amplify their users in a different way to actual independent web pages / the IndieWeb.
But still, yeah, you’re right. I’m just sad that eg the WebFinger setup ended up a lot less static site friendly than the original and is much more catered to micro-Twitters than to actually independent sites.
@voxpelli ❤️
For me, the webfinger stuff was/is just an idea - how can we give people an online name that they control?
(There's some new thinking in this direction that is very exciting, and I think better than any of our previous approaches!)
@blaine Yeah, I remember WebFinger itself never really being intended to be a spec in itself, rather just a pattern for how to apply Host Meta rules to usernames
Which new thinking are you referring to? IndieWeb ones? Crypto ones? Browser based ones?
Please also note that webfinger is not part of any ActivityPub spec.
https://www.w3.org/TR/activitypub/#conformance
It is just a masto thing.
But the fediverse is so many wonderful softwares.
@sl007 @voxpelli the http spec also doesn't have anything to say about html, nor does webfinger say anything about ActivityPub, as it should be. Webfinger isn't "just a masto" thing, either - any social software will need to use the pattern if adoption and sharing is a concern.
It's perhaps best to think of webfinger as "DNS for people"
“any social software will need to use the pattern if adoption and sharing is a concern.”
I mean, the fediverse consists of many softwares.
Just see our Conferences
https://sebastianlasse.de/#home
And those who do only implement ActivityPub and no webfinger is superfine too.
I do not see any advantage.
Yes. Maybe.
Just hoped there would be arguments to use it.
Quite interesting when former twitter people tell the fediverse what it needs.
I did not say that webfinger is bad.
What I am saying is:
New implementors look at the official ActivityPub spec.
An implementor can be any human being in this world, even marginalized or disabled or even an anticapitalist working remote (crazy, I know).
This why I want that every kid can implement it !!!
They do not know that they need webfinger for implementation x but butter cream cookies for implementation y …
They just want to use it and interoperate. That worked all the time. Also without webfinger.
The problem is that webfinger is not part of the spec.
That's it.
If you believe it should be, this is what the Fediverse Enhacement Proposal Process (FEP) was made for or if it should be an official extension, the group can vote.
Just want to understand.
“I've been having this debate with technologists” - I am journalist and artist, sorry.
@sl007 I appreciate that, thanks.
For context: I wasn't just a Twitter employee, I created much of it but have been written out of the history because I am an anti-capitalist. I both tried to make twitter decentralized, creating the first fedi-instance with @ralphm and, when I was forced out of Twitter, was very involved in designing the protocols that became ActivityPub a decade later (thanks to the work of *many* people who came after).
@blaine @sl007 +1, the web consists of URL:s and hyperlinks. Hyperlinks links together two URL:s and can specify a relation from one to another.
None of URL, Hypertext Markup Languages or Hypertext Transfer Protocols needs to know the semantics of a relation and that relation can be specified independently of them and by a user be composed into a new creation.
Relevant spec: https://www.rfc-editor.org/rfc/rfc8288.html
We eg. did that at Flattr. We used rel-payment.