Break the Login Status API out of the FedCM spec #642
Conversation
|
Uhm is this a request from someone? Also I'm not sure but I imagine there needs to be some procedure to add a new spec to the WG (even if it is all extracted from the current spec). Finally, in terms of editor I would suggest to name the one who actually wrote this which I think is @cbiesinger |
As far as I know the primary requirement is that the spec is in the charter, which it is (https://www.w3.org/2024/03/wg-fedid-charter.html#deliverables) I see this was also added to the agenda for the next call |
Yeah, we have always intended to break the Login Status spec out of the FedCM spec, but never got to it: a browser vendor should be able to implement the Login Status API without having to implement the FedCM API.
privacycg/is-logged-in#55 (comment) and more recently:
I'm already working with @hlflanagan and @wseltzer to figure out how to go about this. We would likely want to also move the spec to its own repo and the login-status repo from the CG to the WG: https://github.com/fedidcg/login-status We also want to coordinate with @johnwilander if he would rather have us use the FedID CG repo or the Privacy CG repo as the basis for the FedID WG repo (e.g. keeping the history and the issues). Either works for me, but we can have that discussion separately.
Yep, chatted with @cbiesinger who agreed to edit here too. Done. We also want to coordinate with @johnwilander and @bvandersloot-mozilla if they still feel comfortable sharing co-editorship here (TPAC 2023 we had the three of us as co-editors). |
samuelgoto
changed the title
samuelgoto
changed the title
spec/login-status.bs
Outdated
| 1. If |client| is null, return. | ||
| 1. If |origin| is not [=same origin=] with the [=/request=]'s | ||
| [=request/origin=], return. | ||
| 1. If |client| is not [=same-origin with its ancestors=], return. |
There was a problem hiding this comment.
You seem to have copy pasted an outdated version. We updated this to same-site...
There was a problem hiding this comment.
Are you sure? This seems to be up to date with HEAD?
https://github.com/w3c-fedid/FedCM/blob/main/spec/index.bs#L400
There was a problem hiding this comment.
Ah, yes, interesting, for some reason I missed this PR in my branch:
I sync and merged and I think this is now up to date. Sanity check?
There was a problem hiding this comment.
Now it seems fine. I guess the plan is to create a new repo for this, right? So anyways the change in this repo will only be to remove the existing spec text.
There was a problem hiding this comment.
. I guess the plan is to create a new repo for this, right?
Yes
So anyways the change in this repo will only be to remove the existing spec text.
Yes, that's correct. I wanted to move into a single PR so that it is clear that the break was clean, rather than in two separate PRs into different repos.
spec/login-status.bs
Outdated
| 1. If |client| is null, return. | ||
| 1. If |origin| is not [=same origin=] with the [=/request=]'s | ||
| [=request/origin=], return. | ||
| 1. If |client| is not [=same-origin with its ancestors=], return. |
There was a problem hiding this comment.
Are you sure? This seems to be up to date with HEAD?
https://github.com/w3c-fedid/FedCM/blob/main/spec/index.bs#L400
SHA: 57d30a6 Reason: push, by samuelgoto Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Move the Login Status API out of the FedCM spec * Remove the login status spec from the FedCM spec * Set Christian as the editor * Use the right user id for Christian * Address feedback * Merge with PR 548e7b2 * Add link references to the login status spec * Remove the reference to same-origin with its ancestors --------- Co-authored-by: Sam Goto <[email protected]>
Resolves #641.