W3C home > Mailing lists > Public > www-tag@w3.org > June 2012

Re: Registration of acct: as a URI scheme has been requested

From: Harry Halpin <hhalpin@ibiblio.org>
Date: Fri, 22 Jun 2012 19:02:21 +0200
Message-ID: <CAE1ny+6ws_ue7H8EC_sudx+zNvKsdZH4jUX7JgXRdvfT0x2YjQ@mail.gmail.com>
To: Michiel de Jong <michiel@unhosted.org>
Cc: Melvin Carvalho <melvincarvalho@gmail.com>, www-tag@w3.org
On Fri, Jun 22, 2012 at 2:01 PM, Michiel de Jong <michiel@unhosted.org>wrote:

> Hi Melvin,
> On Fri, Jun 22, 2012 at 12:22 PM, Melvin Carvalho
> <melvincarvalho@gmail.com> wrote:
> > Yes, I believe the current draft supports any URI.
> true. that's why the acct: scheme is needed: to make sure user
> accounts identified as user@host can be written down as something that
> falls under "any URI". if you just write 'user@host' as a string, then
> that's not a URI.
> > Therefore, webfinger in theory does not have a dependency on acct:
> that is an incorrect conclusion. in fact, the opposite is true.
> In Webfinger, we talk of user accounts (user@host) as such, without
> implying any reference to email or jabber. This gave us two options:
> - change webfinger to say it accepts "either any URI, or bare
> 'user@host' strings"
> - keep the "any URI" sentence in the spec, and make 'acct:user@host' a
> URI scheme.
> I think the first one would have been a simple solution, but it would
> also have constituted a deviation from web architecture, using
> "custom" strings instead of URIs. In my personal opinion, both options
> had their merits. But that's not the point. The point is, after weeks
> of discussion, the second one was chosen by the authors, and that's
> why you are now seeing this thread at TAG.
> Nobody is forcing you to use webfinger if you don't want to, but we
> are already using webfinger, including the acct: scheme, in
> production.
> Therefore, IMHO anybody who wants to forbid what we do would need a
> very good reason. It would damage the momentum of not only the
> federated social web (of which webfinger is one of the cornerstones),
> but also of "unhosted" web apps (we allow users to connect their
> cross-origin remoteStorage via webfinger).

It seems like it would be better to use just mailto: (i.e., the original
WebFinger design was around *email addresses*, not accounts per se) or a
http: URI rather than mint a whole new scheme. The concern I think would
naturally be that 1) minting URI schemes for every new app is a bad idea
and that 2) registering new URI schemes usually takes a while, and thus
would officially subtract from the momentum.

However, if somehow the momentum of the WG depends on getting this URI
scheme registered and the group feels that registering a new URI scheme
makes sense and somehow increases their momentum, power to the people.

Regardless, I'll have to join the mailing list to check it out.


> Please just allow us to use webfinger and be happy. :)
> Cheers,
> Michiel
Received on Friday, 22 June 2012 17:02:54 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:56:45 UTC