- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Wed, 26 Feb 2014 13:23:38 -0800
- To: Daniel Holbert <dholbert@mozilla.com>
- Cc: www-style <www-style@w3.org>
On Tue, Feb 25, 2014 at 6:06 PM, Daniel Holbert <dholbert@mozilla.com> wrote: > Hi www-style, > > (This is about a related but slightly more subtle issue than my previous > email[0]). > > Question: Should "span" be accepted as a valid <custom-ident> value, in > the <line-names> list for the grid-template-columns and > grid-template-rows properties? > > The keyword "span" does *not* appear in these properties' value > definition[1], so it would seem that they should accept it. (based on > the definition of <custom-ident>) > > HOWEVER: The keyword "span" *does* appear in the value definition for a > *different* group of properties -- grid-row-start, grid-column-start, > grid-row-end, and grid-column-end[2] -- and those properties *also* take > a <custom-ident>, which can correspond to the same <custom-ident> used > in grid-template-rows/grid-template-columns. (Both of them can refer to > named lines.) > > So if we do allow 'span' as a <custom-ident> for > grid-template-rows/grid-template-columns, it'll just be guaranteed to > fail to match against any corresponding <custom-ident> in > "grid-row-start" & friends (since it won't be a valid <custom-ident> for > those properties). > > Maybe that's OK? (I tend to think it's OK; I'm just sanity-checking my > understanding, while reviewing a patch to implement this parsing in > Mozilla.) You got it. It's not explicitly disallowed, but ends up being useless due to the way <custom-ident> is defined. If we ever defined some other method of referring to lines that didn't include "span" in the grammar, you could use the lines there. ~TJ
Received on Wednesday, 26 February 2014 21:24:25 UTC