What problem are you trying to solve?
Browsers provide spell checking by comparing text against built‑in dictionaries (local or server‑side). This works well for general language, but breaks down on pages that rely heavily on domain‑specific terminology—product names, proper nouns, fictional universes, technical jargon, and other vocabulary that is valid in context but absent from standard dictionaries. If your site allows authoring of any of that sort of content, authors are faced with lots of words marked, incorrectly, as errors. So many false positives is actually counter productive as it makes it unlikely that authors will trust the system and find actual spelling errors.
There is currently no way for a page to supply its own domain-specific vocabulary programmatically.
What solutions exist today?
Users can "learn" words individually, but in many scenarios this also has the non-web-like drawback of potentially leaking something about them to other applications that might consult the dictionary. Authors can try to integrate some custom solutions for spell checking that bypass native features entirely.
How would you solve it?
We have a general proposal laid out after many discussions in an explainer.
Anything else?
We also have prototypes in Chromium and WebKit, have had an early TAG review and have open standards positions with some discussion on Firefox and WebKit.
What problem are you trying to solve?
Browsers provide spell checking by comparing text against built‑in dictionaries (local or server‑side). This works well for general language, but breaks down on pages that rely heavily on domain‑specific terminology—product names, proper nouns, fictional universes, technical jargon, and other vocabulary that is valid in context but absent from standard dictionaries. If your site allows authoring of any of that sort of content, authors are faced with lots of words marked, incorrectly, as errors. So many false positives is actually counter productive as it makes it unlikely that authors will trust the system and find actual spelling errors.
There is currently no way for a page to supply its own domain-specific vocabulary programmatically.
What solutions exist today?
Users can "learn" words individually, but in many scenarios this also has the non-web-like drawback of potentially leaking something about them to other applications that might consult the dictionary. Authors can try to integrate some custom solutions for spell checking that bypass native features entirely.
How would you solve it?
We have a general proposal laid out after many discussions in an explainer.
Anything else?
We also have prototypes in Chromium and WebKit, have had an early TAG review and have open standards positions with some discussion on Firefox and WebKit.