[css-values-5] Does media-progress() really want the entire <media-feature> production for its progress value?

[weinig]

I am trying to wrap my head around the media-progress() function and am having trouble understanding exactly what kind of <media-feature>s should be supported as the progress value.

The stated production, <media-feature>, is defined to allow a whole host of syntaxes:

<media-feature> = ( [ <mf-plain> | <mf-boolean> | <mf-range> ] )

meaning it can match things like width > 0, orientation: portrait, etc.

The spec does state: "The specified media query must be a valid “range” type query", but I can't figure out how to map that to a spec concept.

From the tests in WPT, which use only width and height as features, making me think that perhaps the grammar should be changed to use <mf-name> rather than <media-feature>.

[cdoublev]

My guess is that the author preferred media-progress(<media-feature> ...) and container-progress(<size-feature> ...) because it better translates the expected function arguments than media-progress(<mf-name> ...) and container-progress(<mf-name> ...).

The parens in <media-feature> should probably be hoisted into <media-in-parens> (#10131). There is also a plan to rewrite the syntax with a generic <boolean> (#10457). Similarly, I suggested rewriting <mf-name> to <condition-name> (#10790).

Obviously, the spec says to restrict <mf-name> to feature whose Type is range in their definition table.

[fantasai]

This looks like an error in the spec, it should be a media query name only, not an entire boolean expression.