Firefox Accounts on the web...

Chris Karlof ckarlof at mozilla.com
Tue Nov 5 15:53:01 PST 2013


On Nov 5, 2013, at 4:08 AM, Lloyd Hilaiel <lhilaiel at mozilla.com> wrote:

> On Nov 4, 2013, at 8:59 PM, Chris Karlof <ckarlof at mozilla.com> wrote:
> 
>> 
>> On Nov 4, 2013, at 7:04 AM, Lloyd Hilaiel <lhilaiel at mozilla.com> wrote:
>> 
>>> In discussions in madrid and over the past couple weeks, it’s becoming increasingly clear that we will have to host web based resources for various parts of the sign-up flows we’re implementing.
>>> 
>>> This is some squishy work that will cut across multiple efforts.  I’m curious if anyone has started trying to aggregate all of the different screens that will need to be web hosted across sync and fxa on fxos to figure out the best way to host all this jelly.  From software layout to url strategy…
>>> 
>> 
>> It's something we've been (somewhat) ignoring in favor of native efforts, but we're on it. Warner, John, and Ryan have been queued up. So far, I've been making general assumptions about our reliers' requirements, but should make those more explicit.
> 
> I think this prioritization is exactly right, because the web is instantly updatable.  And stuff.  So we have considerably more time.
> 
> That said,  I mention this because in more and more conversations, it’s becoming obvious that we will need both jelly to support native, AND a web-only FxA sign-up / sign-in flow.  Because this is a rather large hole in our current story, and because we have a number of talented freeish agents on our team, I wonder if it behooves us to fully parallelize this and have some folks dedicated to it?
> 
> As a starting exploration, we have a fantastic group in Brighton UK right now (including ozten, john, dan, shane, francois).  Would it be useful (and not distracting) for ongoing FirefoxOS & Sync.next work to focus them on this problem to lay down a strategy for what singing into the web looks like?  John & ryan have already been thinking about this problem.
> 
> In order for them to be successful, I think a set of requirements would be useful to generate.  Like quickly running through and saying for each UA, what jelly is required.  Then defining a minimal scope for the landing page.  (we’ve already gotten requests from QA folks around the org to have a set of acct management features accessible.
> 
> At a high level, a proposal and some pixels for how FxA on the web works would help people understand how we’re building something that works everywhere.
> 

I agree this is a good place to start. John and Ryan are on it.

We also have a "product" problem: How does FxA work across Mozilla products/services?

IMO, FxA Web should a single sign on system for all consumer facing Mozilla relying web services. Sign in to FxA, you get signed in to all the relying Mozilla web properties. Sign out of FxA, you get signed out of all relying Mozilla web properties. IMO, this is crucial to a coherent, smooth UX. It's proven at Google properties, and it's simple. It's not at all clear to me that we got buyin on this strategy from our relying services, e.g., Marketplace, Where's My Fox, etc. This is step 1, IMO.

If there are alternative sane cross-product strategies, I'd like to hear them.

-chris

> lloyd 
> 
>> -chris
>> 
>> 
>> 
>>> lloyd
>>> _______________________________________________
>>> Sync-dev mailing list
>>> Sync-dev at mozilla.org
>>> https://mail.mozilla.org/listinfo/sync-dev
>> 
> 



More information about the Sync-dev mailing list