Re: [csswg-drafts] [css-gaps-1]: Change the initial value for insets property to be 0 (#13137)

The CSS Working Group just discussed `[css-gaps-1]: Change the initial value for insets property to be 0`, and agreed to the following:

* `RESOLVED: change initial value to 0, with a keyword for the current default?`

<details><summary>The full IRC log of that discussion</summary>
&lt;TabAtkins> javierct: this is about default value for the inset property<br>
&lt;TabAtkins> javierct: currently default is 0 / -50%, which makes T-junctions join with each other<br>
&lt;TabAtkins> javierct: we want to propose changing it to just 0, which makes all junctions consistent<br>
&lt;TabAtkins> javierct: reason is, with the default -50%, the value of rule-break doesn't have any visual effect<br>
&lt;TabAtkins> javierct: we got feedback about this being confusing, people tried to change rule-break and nothing changed in the rendering<br>
&lt;astearns> q+<br>
&lt;TabAtkins> javierct: so if we change initial value of insets to 0, that'll give different visual behaviors for all the rule-break values<br>
&lt;TabAtkins> javierct: also either value might be a preferred default to people<br>
&lt;TabAtkins> q+<br>
&lt;TabAtkins> astearns: i'm a little concerned. I get your point about not having a different between the values making it harder to learn<br>
&lt;TabAtkins> astearns: but if people dont' actually want that intersection rendering from "0" value, and have to change rule-break *and* inset to get what they want, that's not great either<br>
&lt;astearns> ack astearns<br>
&lt;dholbert> q+<br>
&lt;fantasai> TabAtkins: If I'm looking at this, I see that 'none' is not what I want, and the other two look the same, yeah.<br>
&lt;TabAtkins> javierct: right, the confusion there is why we are proposing this<br>
&lt;fantasai> TabAtkins: I understand what you're saying, Alan, about if ppl want the previous behavior they have to adjust two properties<br>
&lt;fantasai> TabAtkins: But, this or modifications of this, also looks good. They're both reasonable defaults that people could want<br>
&lt;fantasai> TabAtkins: so I'm ok to do this by default, if it makes easier to learn about<br>
&lt;fantasai> TabAtkins: If strong difference in what ppl want by default ... but seems like a weak preference<br>
&lt;astearns> ack TabAtkins<br>
&lt;florian> q+<br>
&lt;fantasai> TabAtkins: so I'm ok with this being the default and authors just have to learn that to get that 4th behavior you would use -50%<br>
&lt;astearns> ack fantasai<br>
&lt;TabAtkins> fantasai: -50% isn't a very intuitive value for an extremely common behavior. I strongly recommend we have a keyword to get that behavior.<br>
&lt;TabAtkins> fantasai: it can compute to -50%, but would be easier to say<br>
&lt;tantek> +1 fantasai, make the "common case" easy for authors to specify, e.g. with a keyword<br>
&lt;TabAtkins> fantasai: but going back, part of the problem here is the diagram are solid lines. if you used dash, it would be a lot more obvious<br>
&lt;TabAtkins> fantasai: if you use rule-break:intersection, at *every* intersection you'll restart the dashed lines<br>
&lt;TabAtkins> fantasai: that's usually bad, not what you want<br>
&lt;TabAtkins> fantasai: so I can see how this change would help the learnability. but will it do the thing you want?<br>
&lt;TabAtkins> TabAtkins: yes, since the default is spanning-item, you'll get continuous dashes<br>
&lt;TabAtkins> fantasai: okay cool. we should have example with dashes to make this more obvious<br>
&lt;TabAtkins> fantasai: but yeah, add a keyword for the common case<br>
&lt;astearns> ack dholbert<br>
&lt;tantek> +1 fantasai, better to test rules / rule-beak defaults with dashed/dotted to see what's happening and what would authors want as default/common behavior<br>
&lt;TabAtkins> dholbert: echoing fantasai, I see the previous slide with connected lines as being the most common thing people will reach for so want to make it easy<br>
&lt;TabAtkins> dholbert: so a keyword is good<br>
&lt;TabAtkins> dholbert: for dashed lines restarting at gaps, if there's no way to avoid it, that could be a separate issue<br>
&lt;astearns> ack fantasai<br>
&lt;astearns> ack florian<br>
&lt;TabAtkins> TabAtkins: you do get a continuous dashed line in "spanning-item", the default<br>
&lt;TabAtkins> dholbert: okay, that works<br>
&lt;TabAtkins> florian: I also agree a keyword for -50% is great<br>
&lt;TabAtkins> florian: for the proposal to switch initial value, I could probably agree<br>
&lt;TabAtkins> florian: we could maybe have an initial value of "auto" that tmeans 0 in intersection and -50% in spanning-item, perhaps?<br>
&lt;TabAtkins> florian: if we don't do that, I agree with the 0 proposal<br>
&lt;fantasai> PROPOSED: make initial value of `-inset` be zero, come with a keyword for the current initial behavior<br>
&lt;TabAtkins> dholbert: I kinda like florian's proposal to have "auto"<br>
&lt;kizu> +1 to just `0`<br>
&lt;TabAtkins> astearns: so any objections to changing initial value to 0, with a keyword for the current default?<br>
&lt;TabAtkins> RESOLVED: change initial value to 0, with a keyword for the current default?<br>
&lt;fantasai> Keyword ideas mentioned in th room: meet, fill, junction...<br>
</details>


-- 
GitHub Notification of comment by css-meeting-bot
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/13137#issuecomment-3814256054 using your GitHub account


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

Received on Wednesday, 28 January 2026 22:28:04 UTC