- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Mon, 3 Mar 2014 11:18:06 -0800
- To: Daniel Holbert <dholbert@mozilla.com>
- Cc: www-style <www-style@w3.org>
On Sun, Mar 2, 2014 at 2:31 AM, Daniel Holbert <dholbert@mozilla.com> wrote: > On 02/26/2014 01:25 PM, Tab Atkins Jr. wrote: >>> Question: Should "minmax" [...] "auto", and "subgrid" be accepted as valid >>> <custom-ident> values, in the <line-names> list for the >>> grid-template-columns and grid-template-rows properties? > [...] >> Correct, they're not valid. You're right that they'd be unambiguous, >> but I don't think it's necessary to care. > > Followup question: > > When a UA is parsing the grid-template-rows / grid-template-columns > properties **as subcomponents of the "grid" shorthand property**, should > it consider the keywords "rows", "columns", and "dense" as being > disallowed <custom-ident> line-names? > > (These keywords aren't normally disallowed <custom-ident>s, but in the > context of the shorthand, they might have to be, since they are valid > keywords for other sub-properties of the shorthand. There'd be no > ambiguity if we allowed them here, since as <custom-ident> values, > they'd be parenthesized. But I can imagine hypothetical shorthand > properties where there would be ambiguity.) Hm, that's a good question. Technically, the definition of <custom-ident> means that something which is allowed in a longhand is illegal in a shorthand, which isn't great. It means we can't always serialize a shorthand. I wonder if it's better to change the definition of <custom-ident> to not do this? Instead, maybe it would only disallow the keywords that it can be juxtaposed with. That seems hard to define, though. Hm. ~TJ
Received on Monday, 3 March 2014 19:18:53 UTC