If you're not using the recommended browsers, you may've come to notice that my page breaks. If you're a webdeveloper, you'll know why, too: I use box-sizing.

For the non-techies amongst you, a quick illustration:

There are two so-called box models floating around on the internet. They disagree about how a block element (likened to a box, hence the name), like a paragraph, or a table, or an image^*, should be measured. Computers aren't human beings, obviously, and as such need to be told precisely what a webdeveloper means when he says, "I want this box to be 100 pixels wide".

That there's any dispute at all might come as a surprise to you. How else can one measure a box except from wall to wall - applied to the internet, thus, from (potentially invisible) border to border? Well, bizarrely, the standards disagree with this intuitive way of measurement.

The standards of HTML define, in their own way sensibly, that the total space a block element takes up on a page is defined by space = width + (2 * padding) + (2 * border_width) + (2 * margin). This means that the width of a box is measured by its content.

The standards box model is both convenient... and a curse. It means the visual space an object takes up changes each time you want to space the padding differently, forcing you to adjust the width to suit.

It also makes percentage designs impossible without nesting block elements: for example, width:100% can't have any padding, lest it extends outside the viewport and forces a vertical scrollbar. The only thing that saves the standard's box model is that for some inexplicable reason, a block element without a width declaration will 'stretch to a hundred percent' as one would intuitively expect it, as changes to the padding will not cause it to stretch out - the width is auto-adjusted so that width and padding together are a hundred percent.

Fortunately, the standards, for CSS3, include an attribute called box-sizing. Internet Explorer, indirect cause of all this mess, as it is the main 'proponent' of the intuitive "border-box" model, long since has learnt to behave when given a DOCTYPE at the top of the page, using this as a switch between "quirks mode" (when absent) and "standards mode" (when present). Opera supports the attribute. FireFox supports the attribute via -moz-box-sizing while it is still in development.

Because it has so little support so far, I fear webdevelopers will shy away from it. "There won't be enough incentive to use it. Other browsers might never support it. And why deviate from the standards we've been using for years, anyway? We know our tricks by now."

But there is incentive.

And hence, ideologically, I choose to use it. It's not necessary - in fact, it means my page will not be viewable in plenty browsers, including Safari - but it makes a statement that's long overdue.

I hope at least someone is listening.