RE: [css3-flexbox] what's the computed value of "width: flex(...)" in a non-flexbox context?

Is there a precedent for value types like that in other places? 

For example, in CSS2.1 percent is not applicable to border-width. Is setting a percent value making the property invalid or is it setting initial value?

± -----Original Message-----
± From: Tab Atkins Jr. [mailto:jackalmage@gmail.com]
± Sent: Thursday, June 09, 2011 2:07 PM
± To: Alex Mogilevsky
± Cc: Daniel Holbert; www-style@w3.org
± Subject: Re: [css3-flexbox] what's the computed value of "width:
± flex(...)" in a non-flexbox context?
± 
± On Thu, Jun 9, 2011 at 12:16 PM, Alex Mogilevsky <alexmog@microsoft.com>
± wrote:
± > ± From: www-style-request@w3.org [mailto:www-style-request@w3.org] On
± > Behalf ± Of Tab Atkins Jr.
± > ± Sent: Thursday, June 09, 2011 11:17 AM ± ± I'm inclined to have both
± > flex() and the 'fr' unit compute to 'auto'
± > ± in non-flex contexts.  We don't really need to do anything smart,
± > because ± it's simply an error to use these units outside of flex-aware
± contexts.
± >
± > By "auto" you mean "initial", whatever it is for the property?
± >
± > I think it should be either invalid or initial. Initial probably
± better...
± 
± Sure, yeah.  The initial value for width/height is 'auto', and flex units
± aren't allowed in any other property, so it's equivalent.  Using flex
± units somewhere else is the same as using an invalid value, like setting
± "padding:blue;" - it just makes the declaration invalid.
± 
± ~TJ

Received on Thursday, 9 June 2011 21:14:55 UTC