Thanks, Chris. Fwiw, there have been discussions on this topic and that
group of decision makers you noted is meeting early on Monday to try to
forge consensus. But this is very helpful!
As I read over your plans, these are the questions that immediately jump
to mind.
1. I think part of the hold up in agreeing to a "let 'em be" route has
been a question of the "costs" of maintaining current and new sync.
Some of those costs are what you identify as open questions around UX
that need to be answered. I don't know if there are other costs,
technical or otherwise; I suspect most of the costs are UX related in
trying to minimize confusion. Getting clear on these costs would be a
huge help.
2. You express a lot of concern around upselling current sync users to
new sync. We will come back to the question of upselling in a moment,
but in this plan, what is the experience of an current sync user who
decides to create a FxA? We could think about ways to get really
targeted with our messaging and limit the amount of noise we make to
current sync users about new sync, but we can't completely shield the
current sync user from learning about accounts. The current assumption
is still that we're going to make a lot of noise around the launch of 29
(for reasons Austrails and to get as many users to create an account as
possible). We can of course revisit that assumption, but I'm unclear
about the implications for current sync users in your plan. What's the
risk of a current sync user creating an account?
3. This may be related to #2, but I'm unclear on what exactly won't work
with Sync on Fx29 in your idea below:
The default "Create Account" screen for Sync will change to use FxA. We
include warnings that this won't work with Sync on Fx29- or Sync with
pairing. For users looking for "Existing Sync", we also include a link
that will direct them to the Existing Sync "Setup Sync" and "Pair A
Device" screens.
Thanks!
CW
On 1/3/14 3:59 PM, Chris Karlof wrote:
Hi all,
I see today listed as the deadline for a decision on a transition
plan. I haven't seen significant discussion on this over the last day
or two.
What I've seen is:
- https://wiki.mozilla.org/User_Services/Sync/Migration
- https://services.etherpad.mozilla.org/sync-migration
-
https://www.lucidchart.com/documents/edit/4678-1408-52b1b652-9823-7a810a00c462 (Sync
Migration tab)
- Brendan asking "why given how sync works now, we can't keep the
option for those users who want the same secrecy property they have now"
The most concrete transition proposal I've seen is by Richard and Ryan
F., largely captured in the Lucid chart linked to above. It's largely
sensible, but it includes elements that I don't consider minimal
(e.g., auto transitioning users and upselling).
Here's my proposal for a bare minimum strategy for an *initial*
transition from Existing Sync to FxA Sync.
tl;dr *Just let Existing Sync users be. Support FxA Sync and Existing
Sync side by side in Fx29 with no attempt to upsell or transition from
Existing Sync -> FxA Sync.*
*What this means:*
1) For users that are already connected to Existing Sync, Existing
Sync will continue to work when they upgrade to Fx29. There will be no
upsell to FxA Sync, and no attempt to transition them to FxA Sync in
Fx29. Just let 'em be. In Fx29, these users will see no evidence of
FxA Sync without disconnecting from Existing Sync first.
2) The default "Create Account" screen for Sync will change to use
FxA. We include warnings that this won't work with Sync on Fx29- or
Sync with pairing. For users looking for "Existing Sync", we also
include a link that will direct them to the Existing Sync "Setup Sync"
and "Pair A Device" screens.
3) The default "Login" screen for Sync will change to use FxA. We
include warnings that this won't work with Sync on Fx29- or Sync with
pairing. For users looking for "Existing Sync", we include a link
that will direct them to the Existing Sync "Setup Sync" and "Pair A
Device" screens.
*What this means for the future:*
1) We don't necessarily have to commit to a sunset date or strategy
for Existing Sync *now*. New users will (most likely) get FxA Sync and
old users have the capability to continue to use Existing Sync, if
they desire.
2) In Fx29, we lay what technical groundwork we can for a future
sunset date, e,g, deprecation messaging channels.
3) In Fx29, we lay what technical groundwork we can for future
transition strategies and upsells, e.g., starting to record the
Firefox version of connected sync clients.
*Why I think this strategy is a good idea:*
1) It's the simplest thing I can think of that both gets FxA Sync out
the door and minimizes disruption for existing Sync users. It's simple
enough that I feel confident that it "won't be terrible" for Existing
Sync users. If an Existing Sync user doesn't attempt to the add a new
device, her experience will be identical to what it is now.
2) It buys us time to design, build, and test transition and upsell
strategies from Existing Sync -> FxA Sync that we can land in
Fx30+. IMO, upsells and auto-transitions are non-trivial, and add both
technical and schedule risk. We have 18 days left!
3) We mitigate negative press (for now). We aren't sunsetting pairing
based Sync yet. For users that want to continue to use that, they
can (for now). This strategy also addresses Brendan's question of "why
given how sync works now, we can't keep the option for those users who
want the same secrecy property they have now". If we're lucky and
clever, we potentially add pairing as advanced security option for FxA
Sync in future releases, in which case won't ever have to sunset it
completely.
*Proposed next steps:*
1) Make a decision (Chad, Karen, Jonath, Mayo, ?). Deadline: Dec 6.
2) Flesh out the details and UX more, particularly what some of the
"Create Account" and "Login" screens would look like multiplexing FxA
Sync and Existing Sync. (Note: even if we don't go with this proposal,
we need to do this work anyway, unless we plan on completely
eliminating the option of connecting to Existing Sync accounts in
Fx29.) Deadline: Dec 7.
-chris
_______________________________________________
Sync-dev mailing list
Sync-dev@mozilla.org
https://mail.mozilla.org/listinfo/sync-dev
