This page says what has been checked and what has not. It is written from the code, and every line below it is asserted by a test that runs against the app itself.
Each choice, tile and slot is an ordinary button, so Tab and Enter reach all of them. There is no dragging anywhere in the app — you tap a tile, then tap where it goes. Every page begins with a link that jumps past the navigation.
The line the app speaks and the line it shows are the same string, at every step. When it will not speak something, it does not show it either.
Clear, Rounded, Standard, Dyslexia and Playful — the Dyslexia setting is the OpenDyslexic typeface. They are chosen per reader in that reader’s settings, and they now load on the lesson and preview pages as well as on the picker.
One size setting per reader, from 12 to 24 point, applied by all six activity renderers rather than by some of them. Tiles and slots grow with the letters they hold, so a large setting does not push a board off a phone screen.
If the device asks for reduced motion, the honeycomb stops pulsing and the honey stops falling. There is also a per-reader control and a calm mode, and any one of the three is enough to still the screen.
Every piece of text the app draws is checked against the surface behind it, in a real browser, against the WCAG AA threshold. The lowest reading on the surfaces we check is above that threshold rather than at it.
When honey arrives, a screen reader is told how much — warmly for a young reader, plainly for an adult. A lesson that earns none says nothing at all, which is the same thing the screen does.
The claims on this page are assertions in a test file. It loads the real pages, reads the styles a browser computed, and fails if any of the above stops being true. Before it existed, nothing would have noticed.
No independent accessibility audit has been carried out, and no disabled person has tested this app with the assistive technology they use. Everything above was measured by us, on our own machines, against the criteria we chose to measure.
The app has not been checked at double the browser’s zoom, or in a window as narrow as a small phone held sideways. Those are real WCAG requirements about text reflowing rather than being cut off, and we have not put a number on either.
Contrast is measured for text. It is not measured for the outlines, borders and focus rings around controls, except for the one search field where we replaced a missing focus ring and checked its colour.
In the picture-matching activities the options are pictures, so a screen reader announces the picture symbol itself. Somebody working entirely by ear cannot currently tell those options apart. This is the clearest known gap in the app, and it is a consequence of a design that carries meaning by picture and sound so that reading is never required first.
The High Contrast setting darkens text that has no colour of its own. Text that carries a specific colour keeps it, so the setting does less than its name suggests.
Contrast is swept on the pages the test can reach when it runs. A page it did not open is a page nobody measured, and the test prints the list of what it opened so that gap is visible rather than assumed.
Saying an app conforms to a standard means somebody tested it against that standard and it passed. What we have is narrower and more specific: a list of properties we check on every run, and a list of the ones we do not. We would rather hand you both lists than a word that stands in for them.
If something here does not work for you, the contact page is the way to tell us. A gap you can name is worth more to this list than anything we would have found on our own.
A free account opens every reading pack in the library. There is no payment path on this site today.
Already have an account? Sign in.