loggedInEmail vs loggedInAs

[Ben Adida]

Hi folks,

As we are implementing navigator.id natively in Firefox, we have the
opportunity to see the point of view of a brand new implementation of
the spec.

One issue that's come up is that loggedInEmail, the parameter to
navigator.id.watch(), is the one remaining place where we explicitly say
"email". This may be suboptimal if we ever want to enable identities
that are not strictly emails in the SMTP sense.

How would people feel about loggedInAs, or loggedInUser instead? Of
course loggedInEmail would keep working for a good while.

Btw, the latest spec is now on github, in case you missed that email a
while back:

https://github.com/mozilla/id-specs/

Yes, we take pull requests on the dev branch :)

-Ben

[Jeff Schnitzer]

On Tue, Jun 19, 2012 at 2:58 PM, Ben Adida <b...@adida.net> wrote:
>
> How would people feel about loggedInAs, or loggedInUser instead? Of course
> loggedInEmail would keep working for a good while.

Either are fine by me.

Jeff

[Dan Callahan]

On 6/19/12 4:58 PM, Ben Adida wrote:
> How would people feel about loggedInAs, or loggedInUser instead? Of
> course loggedInEmail would keep working for a good while.

+1 for loggedInUser.

While we're at it, could I suggest something slightly further afield?
It's hard to talk about "loggedInEmail", because you have to dance
around the idea of a user being logged into your site, but that being
different from the user that's logged into Persona. And all you have are
the words "logged in."

What if we said "expectedUser" instead? I think it'd convey the function
much more clearly: "Set 'expectedUser' to the email address of the user
your page is being prepared for. If the Persona user is different from
what you expected, either the onlogin or the onlogout callback will be
invoked to signal the change."

-Callahad

[Ben Adida]

On 6/19/12 3:30 PM, Dan Callahan wrote:
> While we're at it, could I suggest something slightly further afield?
> It's hard to talk about "loggedInEmail", because you have to dance
> around the idea of a user being logged into your site, but that being
> different from the user that's logged into Persona.

Hmm, are we sure about that? This is the bias that we, Persona team,
have, because we're always hacking on Persona. But I get the sense that
web sites simply think of the users logged into their sites, and that's it.

Folks building web sites, what do you think?

-Ben

[Jeff Schnitzer]

On Tue, Jun 19, 2012 at 3:40 PM, Ben Adida <b...@adida.net> wrote:
>
> Hmm, are we sure about that? This is the bias that we, Persona team, have,
> because we're always hacking on Persona. But I get the sense that web sites
> simply think of the users logged into their sites, and that's it.
>
> Folks building web sites, what do you think?

I certainly don't have any conceptual notion of whether the user is
logged into Persona or not, and never think of it that way. That
doesn't mean that expectedUser is bad though.

loggedInUser/loggedInAs seems more congruous with the callbacks named
onlogin and onlogout, and feels more "obvious".

Jeff

[Lloyd Hilaiel]

I'm also a +1 for loggedInUser. I prefer it to loggedInAs because it's more
explicit *who* we're talking about (not the RP's site itself, not an api key,
but the user).

I'm also a bit resistant to 'expectedUser', because from the perspective of the
RP this user is "expected" to sign in, they already have.

api design === splitting hairs :)

lloyd

> _______________________________________________
> dev-identity mailing list
> dev-id...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-identity

[Lloyd Hilaiel]

https://github.com/mozilla/browserid/pull/1806