Captions benefit blind screen reader users who navigate through tables,
particularly in this case because we potentially have several tables
with headings (one table per group), so when navigating through tables
it might be hard to know which group the headings belong to.
In this case they also benefit sighted users. Usability tests have shown
the "Groups and headings" section is a bit confusing, so adding a
caption like "Headings in Districts" helps clarifying Districts is a
group and the table refers to headings in that group.
The very same table is rendered in the "headings" step of the budget
creation wizard. However, in that case the information of the caption is
redundant because the page is specific for headings belonging to a
certain group. We're making the element invisible but still keeping it
for screen reader users in order to ease their navigation through
tables.
In the past it would have been confusing to add a way to directly
enable/disable a phase in the phases table because it was in the middle
of the form. So we would have had next to each other controls that don't
do anything until the form is sent and controls which modify the
database immediately. That's why we couldn't add the checkboxes we used
when using the wizard.
Now the phases aren't on the same page as the budget form, so we can
edit them independently. We're using a switch, so it's consistent with
the way we enable/disable features. We could have used checkboxes, but
with checkboxes, users expect they aren't changing anything until they
click on a button to send the form, so we'd have to add a button, and it
might be missed since we're going to add "buttons" for headings and
groups to this page which won't send a form but will be links.
Since we're changing the element with JavaScript after an AJAX call, we
need a way to find the button we're changing. The easiest way is adding
an ID attribute to all admin actions buttons/links.
It was broken since we fixed issues with other image fields in commit
394a94cbf, because there we added a fieldset for the image fields, and
so they inherited the styles in budget phases fields (these styles were
added before it was possible to attach an image to a phase).
Since we were using an icon font with no text, screen readers were
announcing things like "Enabled, L", trying to read the icon generated
with CSS.
Using text and replacing it with CSS with an icon solves the problem.
We could also use aria-label, but I prefer using "Yes/No" so the text
can be shown/hidden with CSS. Also useful when using
`save_and_open_page` during tests, since the displayed page will not
have any CSS rules applied.
Out of several existing techniques to hide text [1], we're setting the
font size to 1px in combination with moving the content off-screen
because that way we can override it in the `::before` element.
Just moving the content off-screen has the inconvenient of the content
still being taken into account when calculating the text indentation.
And just using a 1px font would make a 1px-sized square appear when
selecting text, which could confuse users.
[1] https://webaim.org/techniques/css/invisiblecontent/