Skip navigation.

exploreopera

| Help

Sign up | Help

Posts tagged with "www"

IE6 is technically very good

, , ,

Over the last year or so, it's occurred to me that IE6, in spite of it's neglect since it was released in October 2001, is really an incredibly capable browser; much more powerful that any other in many ways.

My reasoning is this: how many other browsers at that time could be "upgraded" using client-side scripting to improve their standards support? Current browsers when IE6 was released were: Netscape 6.2, Opera 5.12, IE/Mac 5.0, Konqueror (probably a v1.x). Mozilla, Firefox and Safari didn't exist, not as final versions anyway.

There are many web design features that IE6 does not support. Since its release, smart people have developed scripts to fix many of them: transparent PNGs, :hover on any element, position:fixed, min/max-width/height, <abbr>, SVG, <canvas>, DOM-compliance. Then there's the big-momma of them all: "IE7". The common thing these scripts share is that they are all client-side scripts and download and do their thing without the user being aware of anything; nothing needs to be installed.

I don't think there is any other browser around, even today, that could do that!

Of course, today's browsers don't need to do that! They haven't stood still like IE has. In particular Opera, Mozilla, and Safari have all been striding ahead implmenting fabulous new features for web developers and fixing bugs. They do it smaller, faster, better.

I'm just amazed at how effectively people have been able to drag IE up to even somewhat the level of current browsers! The point of this is how powerful is the underlying infrastructure of IE that allows that to happen!

Following the herd

, ,

In the recent version 9 preview Opera is finally following the herd of browsers that default to margin instead of padding to inset body content.

It was inevitable, given that IE, the Mozilla browsers, Safari, etc all do it that way.

It's also sad, because it's objectively wrong - padding is the only correct answer to insetting body content while keeping the body background out to the edges of the canvas. The reason is that in CSS margin is always transparent, so using it should allow the default (usually white or gray) background colour to show all around the edges of the body which is not what page authors usually want. Padding takes on the background properties of the element, so it is the correct answer.

The reason margin "works" is because browser developers have had to implement nasty special case hacks for the <body> element, and the W3C has had to document nasty special case hacks for the CSS spec to allow that. I really wish the W3C had had the guts to make a stand and say "sorry guys, margin is wrong, use padding".

So, how did IE and Mozilla get to be using margin anyway? Before CSS web pages were styled using HTML tags like <b>, <i> and various attributes, such as "border" on tables. To indent <body>, IE supported the attributes "leftmargin" and "topmargin", while Mozilla supported "marginwidth" and "marginheight". Note that common term "margin". When the CSS spec came along, it also had a margin property, and the IE and Mozilla guys defaulted to using that, without noticing that CSS margin doesn't perform the same function as those attributes.

So now we have a situation where we have ugly hacks in browsers and specs - but only for HTML. XHTML pages have no exemption - they will have to use padding instead of margin for the same effect, or style <html> with the background style that used to be on <body>.

Unless, I suppose, that enough people bitch and moan about "bugs in browsers not handling margin properly" to force browsers and the W3C to mangle their respective products yet again.

That's what worries me the most. I'm probably worrying about nothing, but here we have a situation where the standard bearers are happy to capitulate to the wants of the masses, no matter how wrong that may be. I can't think of anything similar that exists right now, but if it's happened once, it could happen again. What's the point of standards then?
October 2008
SMTWTFS
September 2008November 2008
1234
567891011
12131415161718
19202122232425
262728293031