This addresses part of the ‘is Notify dependable’ group of needs.
The ones it specifically and partially addresses are:
- is it reliable
- how is it supported
There’s more to come in this section, we’re doing this bit now because
it’s a nice conclusion to the page.
Also reword some of the `<p>`s so that they flow better from the
headings.
This was a bit me, and a bit sitting down and hammering this stuff out
with Stephen and Sheryll.
This is the first step towards a fully-fledged ‘product’ page.
The needs for the top, blue section of the page are:
- what is Notify?
- can I use it?
- can I test it out?
- how do I create a log in
The needs for the next section of this page (which is the only one added
by this commit) are:
- how will Notify help me work?
- will it work with my service?
This commit shows 4 features of Notify on the home page which address
those needs. They are illustrated because:
1. We want to catch people’s attention – users are reluctant to scroll
on this page because they just want to click the ‘create account
button’. But we hypothesize that they will get on better
with Notify if they look at some of this stuff first.
2. The concepts that they’re talking about are hard to explain with just
words because they’re quite abstract. The illustrations help us be
more specific.
3. Feedback we got from user research was that the product page didn’t
give users any sense of what it was like to actually use Notify.
This copies the style that Tim and Stephen have been developing for all
product pages.
It also pulls out the CSS for this into its own file, so that it could
potentially be reused.
This is part of the new header style that Tim, Stephen et al have been
working on.
This means that we lose the feedback link, so I’m trying out having it
in the top right.
We don't want large numbers in Production to start overlapping other columns
in the tables when they have less space available, and putting these messages
at the top of the page under the h1 means that we don't need an extra column
on the page yet.
This link looked odd floating above the left column, and although we may want
to have admin navigation on the left we aren't sure what that would include
yet, so move this link to the header alongside the Platform admin link.
When we make the numbers on this page more filterable the date range will be
one of the options to change, so it makes sense to move it to the side now
instead of leaving it above the big numbers.
This page currently includes all notifications for all services, including
those sent using a test key. Stats on all other pages exclude test key usage,
though, which can lead to confusion for admins comparing numbers between
pages. Adding the option to toggle between including and excluding test key
usage on the platform admin page gives context for this difference, and allows
admins to see live usage of the platform as well as load associated with test
key usage.
- This allows us to set a custom header for admin calls only (not needed in client calls)
- Adds request-id from Middleware to the API call to ensure the API logs against the same request ID.
> Make the header bar red
>
> Red for admin is a good reckon.
– 286fc308d9 (part of https://github.com/alphagov/notifications-admin/pull/130)
Starting to think it’s not such a good reckon. Users could take a guess
at what it meant, but they often guessed wrong.
However, changing the colour of the header bar _is_ useful for us
internally to see which environment we’re in. So this commit makes three
changes:
1. On live, the header bar is always standard GOV.UK blue
2. On other environments, the header bar is some other colour (local is
very different, staging and preview are related colours)
3. If an enviroment has a different header colour, it has it even when
you’re not logged in.
* all current invocations of get_services now call get_active_services
EXCEPT for platform admin page (where we want to see inactive services
* cleaned up parameter names and unpacking (since *params is unhelpful)
* fixed incorrect kwarg name in conftest
