- From: CSS Meeting Bot via GitHub <noreply@w3.org>
- Date: Wed, 11 Feb 2026 16:50:30 +0000
- To: public-css-archive@w3.org
The CSS Working Group just discussed `[css-om] Serialization of font-family names: quoted string vs keyword`, and agreed to the following: * `RESOLVED: option 2 for both specified and computed value, if that's web compatible` <details><summary>The full IRC log of that discussion</summary> <ydaniv> fantasai: this is about serializing font-family property and whether value should be quoted or not<br> <ydaniv> ... we can add kind of rule to say whether we quote or not<br> <ydaniv> ... if it matches keywords then you don't need to<br> <ydaniv> ... if it's quoted you can serialize without escapes<br> <emilio> q+<br> <astearns> ack dbaron<br> <emilio> FWIW what Firefox is doing is not _quite_ just doing what the author did: https://searchfox.org/firefox-main/rev/d6dfd9f8cd9717b03750cded8980ef61678fbc8b/servo/components/style/values/computed/font.rs#702-710<br> <ydaniv> ... if you can serialize without escapes then serialize as is<br> <emilio> +1 for both<br> <ydaniv> astearns: is this would be the pattern for the computed value?<br> <astearns> ack emilio<br> <ydaniv> fantasai: yes<br> <ydaniv> emilio: Gecko is not doing exactly what's specified<br> <ydaniv> ... if you have keywords we quote anyway<br> <ydaniv> ... ok with making something conditional<br> <ydaniv> astearns: given it's likely we don't have browsers' interop...<br> <ydaniv> fantasai: I remember writing tests for this<br> <ydaniv> fantasai: quoted if it can't be serialized without quotes<br> <TabAtkins> serialize unquoted if doable without escapes; otherwise quote<br> <fantasai> s/quotes/escapes/<br> <ydaniv> PROPOSED RESOLUTION: serialize unquoted if doable without escapes; otherwise quote<br> <emilio> q+<br> <ydaniv> astearns: this is for both computed and specified values<br> <astearns> ack emilio<br> <ydaniv> emilio: one potential issue is it's not only about escapes, also need to check font-family and CSS-wide keywords<br> <ydaniv> TabAtkins: good clarification<br> <fantasai> https://github.com/w3c/csswg-drafts/issues/5846#issuecomment-3416392705<br> <fantasai> Quote if it matches an existing keyword (already required per spec).<br> <fantasai> Quote if it cannot be serialized as a keyword without escapes, i.e.:<br> <fantasai> contains non-alphanumeric ASCII.<br> <fantasai> contains a word that begins with a number.<br> <fantasai> Otherwise serialize as space-separated idents.<br> <ydaniv> emilio: you need to match it with bunch of strings that are now special<br> <ydaniv> ... some algo that you can do to check, and go back and look at all the keywords, it's a bit unfortunate<br> <ydaniv> ... I guess we need to escape certain keywords<br> <ydaniv> fantasai: they have to be quoted<br> <ydaniv> emilio: is there a case where you need to hard code the special cases?<br> <ydaniv> fantasai: If you keep the quotedness of all is that you need a special state<br> <ydaniv> emilio: that's what gecko does<br> <ydaniv> fantasai: no strong opinion on this, but we'll need to change value<br> <ydaniv> emilio: ok whichever way we're go, but think keep the quotedness is simpler<br> <ydaniv> fantasai: Is it easier to implement? or should we have a dict of words to check?<br> <ydaniv> emilio: [listing cases]<br> <ydaniv> astearns: so are going to allow quoted things where we can?<br> <ydaniv> emilio: would prefer to have it if there's something similar already<br> <ydaniv> astearns: so should we take it back and do impl. research?<br> <ydaniv> ... or take a resolution to unquote where we can?<br> <ydaniv> emilio: ok taking the resolution if others are doing similar stuff<br> <ydaniv> PROPOSED RESOLUTION: option 2 for both specified and computed value, if that's web compatible<br> <ydaniv> astearns: objections?<br> <ydaniv> RESOLVED: option 2 for both specified and computed value, if that's web compatible<br> </details> -- GitHub Notification of comment by css-meeting-bot Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/5846#issuecomment-3885672626 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Wednesday, 11 February 2026 16:50:30 UTC