Closed Bug 1020156 Opened 10 years ago Closed 10 years ago

Need a way to have about:accounts open with a default email address.

Categories

(Firefox :: Sync, defect)

defect
Not set
normal
Points:
3

Tracking

()

RESOLVED FIXED
Firefox 36
Iteration:
36.2

People

(Reporter: markh, Assigned: adw)

References

Details

(Whiteboard: [diamond])

Attachments

(1 file, 2 obsolete files)

For the sync migration, UX calls for us to open about:accounts with the email address set to a specific value.  ie, we will find the email address specified in the current legacy-sync setup, then open about:accounts such that it defaults to creating an account with this address.  The user will still be able to change the value - it's just a suggestion.

cc zaach as I suspect this will require a change to about:accounts *and* to the back-end.
Flags: firefox-backlog+
Adding myself to cc list, since I suspect that jbonacci and I will be splitting the testing duty for this feature.
Flagging as diamond. Zaach, feel like mentoring this if somebody wants to pick it up?
Flags: needinfo?(zack.carter)
Whiteboard: p=3 [qa+] → p=3 [qa+][diamond]
I've opened a content-server issue here: https://github.com/mozilla/fxa-content-server/issues/1258

I assume this will work similar to force auth, where the browser appends an "email=" query parameter to the URL.
Flags: needinfo?(zack.carter)
Assignee: nobody → stomlinson
I am taking the content server portion of this.
The server portion of this is ready. You can open the sign up page with an email query parameter, i.e. `/signup?email=<email-to-prefill>`.

I assume we'd use this for all legacy-sync users who would visit about:accounts. Are there more UX details for someone interested in writing the about:accounts patch?
Thanks Zach/Shane. The client side of this patch currently depends on other bugs (likely bug 1019985), but once we tackle those this should be more straightforward.
Assignee: stomlinson → nobody
Points: --- → 3
Flags: qe-verify+
Whiteboard: p=3 [qa+][diamond] → [diamond]
Blocks: 999910
No longer blocks: migratesync
Assignee: nobody → adw
Status: NEW → ASSIGNED
Iteration: --- → 36.2
This cleans up URL param handling and in the process fixes this bug, if I understand it right.  All URL params except for "action" from the about: URI are forwarded on to the remote page, including "email".  So you can open about:account?email=foo@foo.com and it'll fill in the email address.

I don't know why wrapper.init try-caught constructing the URL and setting the iframe src.  Bug 913199 seems to be where this was introduced, but I didn't see a reason.  I wouldn't expect that to throw, so I removed it.

If this looks OK, I'll move on to tests next.
Attachment #8514581 - Flags: feedback?(mhammond)
Comment on attachment 8514581 [details] [diff] [review]
Forward URL params from about: URI to remote URL

Review of attachment 8514581 [details] [diff] [review]:
-----------------------------------------------------------------

LGTM
Attachment #8514581 - Flags: feedback?(mhammond) → feedback+
This is the same as the previous but has two new tests.  They seem kind of redundant since the existing tests hit these paths...  And an end-to-end test that makes sure the email param is copied to the text field in the page would be nice, but I don't think that's possible.  Or should our automated tests use the test servers from bug 1014411?
Attachment #8514581 - Attachment is obsolete: true
Attachment #8515291 - Flags: review?(mhammond)
Comment on attachment 8515291 [details] [diff] [review]
Forward URL params from about: URI to remote URL + tests

Review of attachment 8515291 [details] [diff] [review]:
-----------------------------------------------------------------

While end-to-end tests would be ideal, I think the test coverage is OK as it stands.

::: browser/base/content/aboutaccounts/aboutaccounts.js
@@ +334,3 @@
>        });
> +      break;
> +    default:

I think this switch statement is going to cause you merge issues due to bug 1079835, but I'll let you use your judgement how to resolve that.
Attachment #8515291 - Flags: review?(mhammond) → review+
Indeed, here's an unbitrotted patch that still uses the switch statement and accommodates bug 1079835.  Carrying forward the r+.
Attachment #8515291 - Attachment is obsolete: true
Attachment #8517110 - Flags: review+
https://hg.mozilla.org/mozilla-central/rev/ee0f03650a55
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Flags: in-testsuite+
Resolution: --- → FIXED
Target Milestone: --- → Firefox 36
Setting for Tracy. We'll pick it up if we have the time.
QA Contact: twalker
I see this bug has automated tests, is there need for manual testing? If so can you please offer some guidance? Thanks
Flags: needinfo?(adw)
Thanks Bogdan, I think this is well covered by the automated test and my own and others' manual testing in the past.
Flags: qe-verify-
Flags: qe-verify+
Flags: needinfo?(adw)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: