W3C home > Mailing lists > Public > www-style@w3.org > December 2017

Re: Shorthands resetting properties they cannot set

From: Gérard Talbot <www-style@gtalbot.org>
Date: Fri, 15 Dec 2017 10:43:40 -0500
To: François REMY <francois.remy.dev@outlook.com>
Cc: Daniel Glazman <daniel.glazman@disruptive-innovations.com>, www-style list <www-style@w3.org>
Message-ID: <8c5bb1437b83ef27d0ff92a233e8d364@gtalbot.org>
Le 2017-12-04 14:49, François REMY a écrit :

> I furthermore disagree with the statement that says it is confusing
> that "border: something" resets "border-image",

I find it unexpected because it is imbalanced and inconsequent.

> I really like when a
> property resets all the properties that have the same prefix, and find
> it confusing otherwise.

Shouldn't "border: 3px solid white" also reset a border-radius 
declaration then?
Shouldn't "font: 20px serif" also reset a previously declared 
I am following your line of reasoning here.

> If as an author you write "border: 3px solid
> white" you want a "3px solid white" border, and not sometimes this
> turn into non-white border because another rule in the document did
> set a border-image previously.

The thing is: border-image covers the border-box area, not just the 
border belt. And border-image paints above the border belt. So, by 
itself and of itself, it is an extra layer *above* the border box. And 
even further than the border box with border-image-outset. border-image 
has little to do with the border belt by definition.

> Regardless of my own preference, I am not sure we have a point strong
> enough to cause all browsers not to follow the spec and have to fix
> their implementation as there is a real cost involved here and I'd
> rather spend the time fixing something useful than tweaking the
> behavior of a property that, as you mention, has minimal use anyway.

I think border-image is incorrect description/identifier for property 
name because of what it does and because of what it can do.

Received on Friday, 15 December 2017 15:44:11 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:09:09 UTC