In the latest release of eslint-plugin-github which includes a bump of aria-query 5.3.0, one of the jsx-a11y rule is newly flagging:
<td role="cell">
This seems to be unexpected.
We depend on aria-query 5.3.0 in this project for one of our rules, but this is higher than what the latest release of eslint-plugin-jsx-a11y depends on. I think this is resulting in us seeing unexpected behavior which would normally be taken care of by the eslint-plugin-jsx-a11y library. This PR:
temporarily overrides the config for the rule to avoid raising a false positive.
Although I considered reverting 5.3.0, I opted not to since we made a ton of updates in our code in https://github.com/github/eslint-plugin-github/pull/448 in the process of bumping to 5.3.0, that I don't know if it's worth reverting.
However, I think that we should aim to keep the version of aria-query in sync with the version on eslint-plugin-jsx-a11y. I wonder if we should lock down the version, or if there are consequences to this? 🤔
In the latest release of eslint-plugin-github which includes a bump of aria-query 5.3.0, one of the
jsx-a11yrule is newly flagging:This seems to be unexpected.
We depend on
aria-query5.3.0 in this project for one of our rules, but this is higher than what the latest release of eslint-plugin-jsx-a11y depends on. I think this is resulting in us seeing unexpected behavior which would normally be taken care of by theeslint-plugin-jsx-a11ylibrary. This PR:Although I considered reverting 5.3.0, I opted not to since we made a ton of updates in our code in https://github.com/github/eslint-plugin-github/pull/448 in the process of bumping to 5.3.0, that I don't know if it's worth reverting.
However, I think that we should aim to keep the version of
aria-queryin sync with the version oneslint-plugin-jsx-a11y. I wonder if we should lock down the version, or if there are consequences to this? 🤔