Re: box-sizing alternative

>From: "Joe Hewitt" <joe@joehewitt.com>
>Date: Tue, Feb 29, 2000, 11:22 AM
>
>> And finally,
>> adding one controlling property ("box-sizing") is preferable
>> to adding two
>> alternative properties ("content-width" and "content-height").
>
> It is very noble of you to try and cut down on the number of properties
> added to CSS.

Yes, simplicity (and ease of authoring) is a core goal of CSS.

> However, in this you are making too many sacrifices just to
> save a few added properties.  You are adding in something that is, IMO, very
> clumsy: "box-sizing". I say that it is clumsy because it is a different way
> of affecting size than the user is used to.

Untrue.  HTML sizing is inconsistent.  E.g. WIDTH on IMG is
box-sizing:content-box whereas WIDTH on TABLE is box-sizing:border-box.

We are making it more consistent, thus more predictable, thus easier to author
to.

> Every other size property is
> set by giving it an explicit pixel size.

or in,cm,mm,pt,%,em,ex and in some cases thin,thick,medium, and in some other
cases, normal.

> The solution I have detailed
> allows the user to continue to use pixel sizes to define specific regions.

So does box-sizing.

> I believe this is simpler than making the user "change the rules" of how
> size is interpreted.

box-sizing was introduced specifically so we wouldn't have to change the rules
of CSS-1 and CSS-2 while introducing a mechanism for specifying the "total
width" of an element including its borders, like, many elements in HTML
behaved.

In HTML, the rules for whether WIDTH includes border or not are tag specific
and therefore inconsistent.  Inconsistent rules might as well not be rules at
all (some legal analogy comes to mind).

Also, we are trying to get away from hardcoded tag specific formatting rules.

> Can you please explain to my why "changing the rules" is better than just
> setting the size of a specific region within the box, or the box as a whole?

It is actually your proposal which advocates "changing the rules" from what
was already documented in CSS-1 and CSS-2.

Tantek

Received on Tuesday, 29 February 2000 14:32:57 UTC