Re: Sloppiness of `box', `area', and `width'

On Tue, 15 Nov 2011 12:45:19 -0800, Tab Atkins Jr. wrote:

> On Tue, Nov 15, 2011 at 11:17 AM, Michael Witten <> wrote:
>> On Tue, 15 Nov 2011 10:13:34 -0800, Tab Atkins Jr. wrote:
>>> On Tue, Nov 15, 2011 at 7:22 AM, Michael Witten <> wrote:
>>>>    * The phrase `content box' means "the box's content area".
>>>>    * The phrase `padding box' means "the box's padding area".
>>>>    * The phrase `border box'  means "the box's border area".
>>>>    * The phrase `margin box'  means "the box's margin area".
>>> This equivalence isn't stated explicitly, but it appears obvious.
>> It's sloppy.
>> The whole section begins by talking about the structure of a box;
>> it's sloppy to introduce the word `box' again with a different
>> meaning (just as it's sloppy to reuse the word `edge' with different
>> meanings).
> It seems unimportant to clarify it, as there is no ambiguity and the
> inference can be made from normal English definitions.

That's a terrible basis for writing a specification.

>> What gets defined is an *area*, not a *box* (which, again,
>> is not the same kind of box as that which is meant to be
>> described by the box model).
> There's no need to use this definition.  In any cases, it's incorrect;
> often when we're talking about "the padding area" we are referring to
> the rectangular box.  The term does sometimes refer to the
> donut-shaped area between the box's edge and the contained box's edge,
> but it's obvious from context which is being referred to.

Then the section should be rewritten to reflect these facts, as noted:

  Subject: RE: Sloppiness of `box', `area', and `width'
  Message-ID: <>

> If you can find any place where there is actually confusion or
> ambiguity due to this somewhat imprecise use of English, please inform
> us.

I have many times now.

`Somewhat imprecise' language is a terrible basis for writing a

>>>  The "width" talked about in this chapter is the distance between
>>>  a box's edge and the nearest enclosed box's edge.
>> Do you see? CAN YOU SEE IT? You basically say what I want to say:
>>          The content edge and the padding edge define the
>>          box's padding area.
>>          etc.
>> Perhaps it will help to point out that `edge' DOES mean `perimeter',
>> as defined by the spec (reasonably, in your opinion), and what we
>> are talking ARE areas. Indeed, my problem with the use `width'
>> stems from these statements in the spec:
>>    * If the padding has 0 width, the padding edge is the
>>      same as the content edge.
>>    * If the border has 0 width, the border edge is the
>>      same as the padding edge.
>>    * If the margin has 0 width, the margin edge is the
>>      same as the border edge.
>> What does `the padding has 0 width' mean? It is complete nonsense!
> No, it's perfectly fine.  You are stuck on a particular definition
> that conflicts with what the spec (and English as a whole) uses, and
> that's causing you trouble.  Anyone familiar with how the 'padding'
> property works can understand that sentence - there are 4 widths, one
> for each side.

As I've stated:

  Subject: Re: Reasonable definitions
  Message-ID: <>

I'm not confused (though I am somewhat amused).

I'm trying to help improve the readability of the specification.
Why are you so opposed to considering the improvement of what
you yourself have admitted is imprecise, conflicting language?

I'm so opposed to leaving the current language as is, because
I think it can be easily improved. Your opposition is bewildering.

If anything, we know that the concept of a "padding box" (etc.)
would benefit from being introduced more explicitly (and with
greater emphasis), and responding to such a suggestion with:

  "Yeah, but, like, *I* understand what the spec currently says"

is just not a proper response.

Received on Tuesday, 15 November 2011 21:09:30 UTC