Closed Bug 744457 Opened 13 years ago Closed 13 years ago

signing into Marketplace does not sign you into Dashboard/AitC (and vice-versa)

Categories

(Firefox for Android Graveyard :: Web Apps (PWAs), defect)

15 Branch
ARM
Android
defect
Not set
normal

Tracking

(blocking-kilimanjaro:-)

RESOLVED INCOMPLETE
blocking-kilimanjaro -

People

(Reporter: Harald, Unassigned)

References

Details

Current solution: Marketplace and Dashboard will both prompt a Persona login on first launch. Having 2 separate logins, this will cause issues when the user is logged in with different accounts in the Marketplace and the Dashboard. Installing an app for a Marketplace account will be synced to the other AitC account. "Sign in to Marketplace/AitC" should sign in both services, until "Sign in to Browser" is defined.
Blocks: 738546
Whiteboard: [marketplace-beta?]
I don't understand this bug fully, aren't we simply going to reuse the same ID that the user logged into the marketplace with for AitC?
Yes, that's the behavior we need to document in this bug. If I didn't miss a ticket, so far we don't have anything tracked and would have 2 separate logins.
Whiteboard: [marketplace-beta?]
Version: Firefox 14 → Firefox 15
Blocks: 744273
Whiteboard: [Phase1]
Nominating for k9o, as this relates to AITC integration into fennec native for the android web runtime for apps.
blocking-kilimanjaro: --- → ?
No longer blocks: 738546
Assuming by Dashboard you mean persona.org, it is indeed the case that signing into Dashboard does not sign you into Marketplace, which is marketplace.mozilla.org, and vice-versa, in my testing. It isn't clear what issues this will cause, though. Harald: can you elaborate? The more specifics you can give, the better we'll be able to understand how important it is to fix this bug, by when we need to do so, and the best way to do it. It also isn't clear that this is a Fennec bug. It might be, but it could also be a Marketplace bug, or a Dashboard/AITC one. cc:ing bwalker for help triaging and jfong for input from the persona.org side of things.
Summary: Sign in to Marketplace & Dashboard/AitC → signing into Marketplace does not sign you into Dashboard/AitC (and vice-versa)
(In reply to Myk Melez [:myk] [@mykmelez] from comment #4) > Assuming by Dashboard you mean persona.org, it is indeed the case that > signing into Dashboard does not sign you into Marketplace, which is > marketplace.mozilla.org, and vice-versa, in my testing. With the AITC client, which hasn't landed yet, signing in to the Marketplace will automatically sign you in to the dashboard (for Firefox users only, of course). We haven't made any plans for this working the other way around (dashboard -> marketplace). Is there a use-case for this?
We need some info from Ragavan or Jenn on whether or not this really blocks. Going to minus until we get that.
blocking-kilimanjaro: ? → -
I think this isn't as much of an issue as it might have been before. This is the sign in conflict that I'd worry about: 1. The last user of the machine is, say, mywife@gmail.com, and she signed into persona.org 2. I go to the Marketplace, log in as ianbicking@gmail.com, and buy an app 3. I don't want that app to be associated with mywife@gmail.com's aitc account, it should be part of my account. I don't believe with the current architecture that this will be a problem, because at step 2 I have to sign out of persona.org and sign back in. Here's another possible issue: 1. I have two persona email addresses, ianbicking@gmail.com and ianb@mozilla.com. 2. I sign into the Marketplace as ianbicking@gmail.com 3. I sign into persona.org as ianb@mozilla.com. Do I see the app from 2? I don't think this can even happen currently, as I don't sign in as ianb@mozilla.com, I sign in as the person who has both ianbicking@gmail.com and ianb@mozilla.com addresses. If these two scenarios aren't problems with the current architecture, then it seems like we're good. If they are problems, then we need to consider how to resolve them?
(In reply to Ian Bicking (:ianb) from comment #7) > I think this isn't as much of an issue as it might have been before. > > This is the sign in conflict that I'd worry about: > > 1. The last user of the machine is, say, mywife@gmail.com, and she signed > into persona.org > 2. I go to the Marketplace, log in as ianbicking@gmail.com, and buy an app > 3. I don't want that app to be associated with mywife@gmail.com's aitc > account, it should be part of my account. > > I don't believe with the current architecture that this will be a problem, > because at step 2 I have to sign out of persona.org and sign back in. > > Here's another possible issue: > > 1. I have two persona email addresses, ianbicking@gmail.com and > ianb@mozilla.com. > 2. I sign into the Marketplace as ianbicking@gmail.com > 3. I sign into persona.org as ianb@mozilla.com. Do I see the app from 2? > > I don't think this can even happen currently, as I don't sign in as > ianb@mozilla.com, I sign in as the person who has both ianbicking@gmail.com > and ianb@mozilla.com addresses. > > If these two scenarios aren't problems with the current architecture, then > it seems like we're good. If they are problems, then we need to consider > how to resolve them? Right. The one problem here that's different than desktop is the initial proposal for AITC integration for desktop targets myapps.mozillalabs.com for integration, which is fine for desktop, but that won't work for mobile (myapps isn't designed for phone use). Something entirely different needs to be thought about here, but generally, I am confused what this bug now is, cause what dashboard are we even using for AITC on android? Is persona.org intending to target mobile? Is there something even entirely different we are doing? Flagging productwanted - Need Jen to look at this and determine how AITC integration for Android should work.
Keywords: productwanted
Whiteboard: [Phase1]
Status: NEW → RESOLVED
Closed: 13 years ago
Keywords: productwanted
Resolution: --- → INCOMPLETE
Product: Firefox for Android → Firefox for Android Graveyard
You need to log in before you can comment on or make changes to this bug.