The spec requires doing nothing here; we throw NS_ERROR_NOT_IMPLEMENTED. (richtext2 tests this, but tests for the wrong result, so it will still fail even if we fix this.)
Created attachment 625485 [details] [diff] [review]
This returns false, like we already do for formatBlock and fontSize. I changed createLink too while I was at it. I just updated the spec (and spec tests) to require this behavior -- previously it required returning true here.
The updates to test_runtest.html.json are partly because of the test changes and partly because of the code changes -- there are new expected fails because we don't support the same set of commands for formatBlock as the spec requires, for example. It's not because the patch actually causes any regressions (I think!).