Tv

This seems to happen with many activate-on-hover things. Timing (`hide-after-inactive-ms`) and keyboard (`hide-when-typing`) seem to behave the same. Affected: - Chrome and Firefox link hover - VSCode/VSCodium tooltips Not affected:...

Quick glance says maybe the only thing Sway is doing is making the cursor invisible (empty image): https://github.com/swaywm/sway/pull/1979/files

Thank you! I can confirm Niri v25.05 hide pointer works great.

It seems a blank line will also prevent reorder_keys.

You could choose to do what a lot of code refactoring does: assume a full-line comment describes an immediately-following non-comment line.

Yet, it is the assumption made all over the place in code formatters. If you want to write a comment that not about the following line, you put a blank...

Even in the world you are advocating for where comments cannot be tied to uncommented lines, commenting out a line silently breaking reordering sounds unpleasant :-(

Here's another reason full line comments seem superior to me -- this would be hard to express legibly with an inline comment. ```toml [[rule]] include = ["deny.toml"] # WAITING https://github.com/tamasfe/taplo/issues/608...

Even if I get the sha256 myself, it never matches. ```toml [[bans.build.bypass]] crate = "aws-lc-sys" build-script = "e8e389f69f556256f9c00702ead273e942c35aca823274d2e305dcd2f7d6b9ea" ``` ```console $ cargo deny check ban --hide-inclusion-graph error[build-script-not-allowed]: crate 'aws-lc-sys =...

That is totally plausible, I'd be thrilled to demonstrate the same crash with a small C program but I'm not that far in debugging yet.