feat(marketing): Enable Marketing Email prefs for more users.

[shane-tomlinson]

issue mozilla/fxa-content-server#2676

@rfk - Whether this merges is dependent on the response to https://bugzilla.mozilla.org/show_bug.cgi?id=1179873#c3. If Basket handles all of this correctly from fxa-register, hooray, we can merge. If we have to pass along the user's language to the /subscribe basket endpoint, we'll have to hold off until that work is complete.

[rfk]

I'm r+ on the change here, but we'll hold off merge until mozilla/fxa-content-server#2676 gets resolved.

[rfk]

The supporting code for this is now merged and will go out with train-47.

[vladikoff]

once trian 47 is live will be merged

[rfk]

once trian 47 is live will be merged

Oh hey, train-47 is live! This is good to go, but I don't want to merge it last thing on a Friday evening and have the change go straight to production - let's deal with it on Monday when prepping the cut for train-48.

[shane-tomlinson]

Oh hey, train-47 is live! This is good to go, but I don't want to merge it last thing on a Friday evening and have the change go straight to production - let's deal with it on Monday when prepping the cut for train-48.

We should be good to merge, the merge goes to dev. Whenever new experiments are merged, a new branch is created for that train, which is then used until a new experiment is added.

As an aside, @jrgm, @vladikoff, @rfk, @dannycoates - I find it really difficult to keep track of which repos use which process for marking a sha as "used by train XX". Some repos use branches, some tags, others, like this repo, cut a new branch only whenever the experiments change. Can we settle on one process, to use everywhere? Branch or tag on every train, even if there are no updates. A unified, well-understood process may have prevented the fxa-content-server-l10n mixup last week, and would prevent confusion like such as whether we should wait to merge this PR now or on Monday.

[rfk]

Whenever new experiments are merged, a new branch is created for that train,
which is then used until a new experiment is added.

OK, cool, thanks @shane-tomlinson. I do remember this discussion, but I don't recall the details of the outcomes and it's not obvious to me where I'd go to discover the process. I filed #29 for us to remember to do something about that.

Merging.