1013423 - Handle registration failing due to the server being unreachable

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: FindMyDevice, defect)

Description

[Guilherme Gonçalves [:ggp]]

We need to define the client-side behavior and UX for when registration fails due to the server being unreachable (which can happen because of a network error, for example).

Please see bug 966033 for my early exploration on this, in both client behavior and appearance. I would expect the UX solution for this to be similar to the one we take on 998489, but this would be place to decide it.

Comment 1

[Erin Lancaster [:elan]]

To john for consult and jsavory can collaborate.

Comment 2

[John Gruen [:jgruen]]

It sounds like you could use the generic error message we're using on Firefox Accounts on lines 47-48. If it's a matter of the user not having their network enabled, lines 35-36. Both of these are displayed in error sheets.


https://github.com/mozilla-b2g/gaia/blob/master/apps/system/fxa/locales/fxa.en-US.properties

Comment 3

[Erin Lancaster [:elan]]

Tagging late l10n so it's understood that the at least one set of strings mentioned above will be used for the gaia client.

Comment 4

[Francesco Lodolo [:flod]]

late-l10n keyword is not useful in these cases, better to cc someone from l10n-drivers to get advice if you have doubts.

So, you plan to reuse 2/4 strings from Firefox Accounts for Find My Device. As long as the context is the same (font-size, role of the text, etc.) it should be fine.

Probably worth having some generic error messages to reuse across apps in 2.1?

Comment 5

[Erin Lancaster [:elan]]

can qa check this and see if bug 1021428 is enough?

Comment 6

[Kate Glazko]

Bug 1021428 handles the case of being offline before registration is initiated.

We have yet to test what happens if registration fails due to server being unreachable, for example, if one has wifi and then loses wifi connection in the middle of attempting to register, or if the server goes down when one is registering.

This isn't trivial to test, and not certain whether Bug 1021428 is enough to handle this situation.

Comment 7

[Jared Hirsch [:jhirsch] (he/him) (Needinfo please)]

Actually, maybe this bug doesn't need to land for 2.0.

There is a similar bug in FxA, and it was decided that, if we detect network offline at the start of the flow, catching network outages during the flow wouldn't block 2.0. See bug 998464, especially https://bugzilla.mozilla.org/show_bug.cgi?id=998464#c8, as well as the linked bug 1026603

Comment 8

[Erin Lancaster [:elan]]

+1 I'm good with aligning with an established FxA convention. Let's get this feature to market and see what users think, too.

Comment 9

[Jason Smith [:jsmith]]

Clearing qawanted since this isn't trivial to analyze & non-blocking issue.

Comment 10

[Michael:mtreese]

[Blocking Requested - why for this release]:

Comment 11

[vkrishnamoorthy]

this is not a release blocker also ni jsavory.

:jsavory is this covered in the wireframes?