Re: [csswg-drafts] [css-color-5] What is the computed value of light-dark(none, none)? (#13866)

The CSS Working Group just discussed `[css-color-5] What is the computed value of light-dark(none, none)?`, and agreed to the following:

* `RESOLVED: light-dark(..., none) computes to image(transparent) if none is the value that was chosen`

<details><summary>The full IRC log of that discussion</summary>
&lt;ydaniv> futhark: light-dark for images supports none value which generates a fully transparent image<br>
&lt;ydaniv> ... but in computed value time it's resolved to [missed]<br>
&lt;ydaniv> ... we need to specify what light-dark(none) computes to<br>
&lt;TabAtkins> q+<br>
&lt;ydaniv> ... anything other than none, don't have a suggestion<br>
&lt;astearns> ack TabAtkins<br>
&lt;emilio> q+<br>
&lt;ydaniv> TabAtkins: we discussed the none be generically for images, lea suggested that we have it computed as itself<br>
&lt;ydaniv> ... agree that it should work like system colors<br>
&lt;ydaniv> ... I'd be ok with resolving it to transparent<br>
&lt;ydaniv> ... prefered that we support the image() with argument transparent<br>
&lt;fantasai> +1<br>
&lt;emilio> +1<br>
&lt;romain> +1<br>
&lt;kbabbitt> +1<br>
&lt;ydaniv> ... it has be in the spec and unimplemeted for a decade<br>
&lt;emilio> So just drop `none` right?<br>
&lt;bkardell> s/has be/has been<br>
&lt;ydaniv> ... so my further suggestion is that we punt all the none colors for image<br>
&lt;astearns> ack emilio<br>
&lt;ydaniv> ... and light-dark(none) resolve to image(transparent)<br>
&lt;ydaniv> emilio: was going to suggest the same thing<br>
&lt;ydaniv> ... I'll be happy kill the none in FF<br>
&lt;ydaniv> TabAtkins: the none kw would still be useful, but it would resolve to image(transparent)<br>
&lt;ydaniv> emilio: I'd be happy to support this, but feels like a weird extension to image()<br>
&lt;ydaniv> ... feels weird that none computes to something else<br>
&lt;ydaniv> ... but not opposing<br>
&lt;TabAtkins> light-dark(url(...), none) vs light-dark(url(...), image(transparent)) is a pretty big typing diff<br>
&lt;ydaniv> fantasai: agree with this plan, I think we already resolve [missed]<br>
&lt;TabAtkins> s/[missed]/to cut down image() to just colors at the f2f/<br>
&lt;ydaniv> PROPOSED RESOLUTION: light-dark(..., none) computes to image(transparent) if none is the value that was chosen<br>
&lt;ydaniv> PROPOSED RESOLUTION: light-dark(..., none) computes to image(transparent) if none is the value that was chosen<br>
&lt;ydaniv> astearns: objections?<br>
&lt;ydaniv> RESOLVED: light-dark(..., none) computes to image(transparent) if none is the value that was chosen<br>
&lt;TabAtkins> I'll cut down the image() function today<br>
</details>


-- 
GitHub Notification of comment by css-meeting-bot
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/13866#issuecomment-4390052390 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Wednesday, 6 May 2026 16:25:03 UTC