acct-uri
parent
2b417c88a2
commit
e3af1df5af
|
@ -36,6 +36,77 @@ acct:reiver@changelog.ca
|
||||||
|
|
||||||
</pre>
|
</pre>
|
||||||
</section>
|
</section>
|
||||||
|
<section>
|
||||||
|
<h2>Why acct URIs</h1>
|
||||||
|
<p>
|
||||||
|
The TL;DR of <em>why acct URIs‽</em> is —
|
||||||
|
</p>
|
||||||
|
<ul>
|
||||||
|
<li>
|
||||||
|
<p>
|
||||||
|
… because people seem to be comfortable with something that looks more like an <strong>e-mail address</strong> or as <strong>username</strong> as an <strong>identifier</strong> but the technology only understands URLs and URIs;
|
||||||
|
so you still have to turn things that look like <strong>e-mail address</strong> and <strong>usernames</strong> into URLs and URIs;
|
||||||
|
which is what <strong>acct URI</strong> is;
|
||||||
|
and “no”, mailto URI won't work;
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
… because a new URI scheme makes it easy to identify that it should be used with <ziba-link transform="lowercase">WebFinger</ziba-link> (rather than being able to do something else with it).
|
||||||
|
</p>
|
||||||
|
</li>
|
||||||
|
</ul>
|
||||||
|
<p>
|
||||||
|
There are many conceptual places where some kind of an <strong>identifier</strong> is a core part of it.
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
For example, on any social protocol I am aware of, there is some type of notion of an <strong>identifier</strong>.
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
BBS echo-mail has them.
|
||||||
|
BBS net-mail has them.
|
||||||
|
Internet e-mail has them.
|
||||||
|
Internet finger-protocol has them.
|
||||||
|
Internet gemini-protocol has them.
|
||||||
|
(Multi-User) Linux & Unix operating systems have them.
|
||||||
|
Twitter has them.
|
||||||
|
Etc.
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
The <strong>identifier</strong> enables you to send messages, to share photos & images, to control data, etc.
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
With a single centralized system, dealing with <strong>identifiers</strong> is comparatively more straight-forward.
|
||||||
|
But when dealing with a distributed, decentralized, or federated system, things can get more complex — for example, how do you point to and interact with someone or something from a different node on the network?
|
||||||
|
Etc.
|
||||||
|
</p>
|
||||||
|
<p>
|
||||||
|
Experience (with OpenID and other systems) seems to suggest that (at least currently) most people are more comfortable with using something that looks more like an <strong>e-mail address</strong> or as <strong>username</strong> as an <strong>identifier</strong> rather than a URL or a URI.
|
||||||
|
For example —
|
||||||
|
</p>
|
||||||
|
<ul>
|
||||||
|
<li>joeblow@example.com</li>
|
||||||
|
<li>jandoe@something.example</li>
|
||||||
|
<li>actor@host</li>
|
||||||
|
</ul>
|
||||||
|
<p>
|
||||||
|
<ziba-link name="id">Fediverse ID</ziba-link>s look very similar to e-mail addresses.
|
||||||
|
They just have a U+0040 at-sign ("@") at the beginning of them:
|
||||||
|
</p>
|
||||||
|
<ul>
|
||||||
|
<li>@joeblow@example.com</li>
|
||||||
|
<li>@jandoe@something.example</li>
|
||||||
|
<li>@actor@host</li>
|
||||||
|
</ul>
|
||||||
|
<p>
|
||||||
|
But <ziba-link transform="lowercase">WebFinger</ziba-link> only understand URLs and URIs.
|
||||||
|
So those have to be turned into URLs or URIs.
|
||||||
|
Which, usually, is very straight-forward:
|
||||||
|
</p>
|
||||||
|
<ul>
|
||||||
|
<li>acct:joeblow@example.com</li>
|
||||||
|
<li>acct:jandoe@something.example</li>
|
||||||
|
<li>acct:actor@host</li>
|
||||||
|
</ul>
|
||||||
|
</section>
|
||||||
<section>
|
<section>
|
||||||
<h2>Fediverse Users</h2>
|
<h2>Fediverse Users</h2>
|
||||||
<p>
|
<p>
|
||||||
|
|
Loading…
Reference in New Issue