| 00:38 | <Michael Ficarra> | I've updated the spell-checking PR to do what we previously discussed: https://github.com/tc39/ecma262/pull/3335 |
| 00:40 | <Michael Ficarra> | it now uses a GitHub action to add annotations to lines that use words other than dictionary words and words that already exist in the document (and are not composed of such words) |
| 00:41 | <Michael Ficarra> | I'd appreciate a review, though I don't think the action will run until it's been merged |
| 00:43 | <Michael Ficarra> | also it wouldn't output any warnings anyway because the PR doesn't introduce any misspellings, but we could try that out with a follow-up PR |
| 00:52 | <Michael Ficarra> | I'd appreciate a review, though I don't think the action will run until it's been merged |
| 01:21 | <bakkot> | commented |
| 01:21 | <bakkot> | you could introduce a misspelling just so we could see it work, and then revert? |
| 01:59 | <Michael Ficarra> | aha! it works! |
| 02:03 | <Michael Ficarra> | and the view from the action itself |
| 03:38 | <jmdyck> | You remember how in PR 3361, the second commit changed "a |Foo| Contained within code" to "any |Foo| x such that code Contains x is true" (https://github.com/tc39/ecma262/pull/3361/commits/f2b2d524)? I just discovered that's invalid, because Contains takes a symbol, not a Parse Node. |
| 03:40 | <jmdyck> | (I didn't notice it before, because I can't analyze the algorithm fragments in Annex B. I noticed it now because I merged main into my monkey_patch PR, putting those fragments into 'mainline' algos.) |
| 03:42 | <jmdyck> | (Of course, the prior wording was presumably invalid too, but more obscurely so.) |