W3C home > Mailing lists > Public > www-html@w3.org > December 2000

Re: Margins and standards

From: todd fahrner <fahrner@pobox.com>
Date: Wed, 13 Dec 2000 08:56:45 -0800
To: Clover Andrew <aclover@1value.com>, "'www-html@w3.org'" <www-html@w3.org>
Message-ID: <B65CEA4C.9CD%fahrner@pobox.com>
Clover Andrew wrote

>> [marginthis, marginthat] seem to be the only way to achieve
>> this effect (in the real world) even on NN4 and so on.
> 
> Certainly in Netscape 4 you have to use marginheight/marginwidth
> to remove the margins, though you could also try using CSS
> absolute positioning to put things right into the corner of the
> page, which might have the effect you want.

Unless you resize the window, in which case NS4 loses this information on
redraw. So then you stick in some Javascript, and ... *blech*

> If you can be satisfied with supporting IE4+, NN6/Mozilla, Opera,
> and a few others, you can use standard CSS, eg.:
> 
> <style type="text/css"><!--
>   body { margin: 0 }
> --></style>

I agree that this works, but as an academic aside, I believe that padding -
not margin - most closely describes the area we're talking about, inasmuch
as the background of the body element itself bleeds into this area, which is
consistent with padding in the CSS box model. Margin is something else. So
if you want to be "more Catholic than the Pope", you can write "margin: 0;
padding: 0".

>> Does anyone know why they are not valid in the Transitional DTD
> 
> Because they're horrible. :-)

Quite. Most significantly, both of these "enhancements" were foisted upon
the Web *after* CSS was accepted by the vendors in question as the right way
to suggest such formatting. In the case of Netscape, these attributes made
their appearance *after* CSS-1 had become a Recommendation. So these are
hardly "transitional", in the sense that they existed before stylesheet
support had been deployed, and are merely redundant; they are more
pernicious in that they reinvented CSS-1 functionality in a manner that
could only displace CSS (to this day, NS4 does not support the CSS-1 way on
BODY). Other examples of this imposter track from Netscape include, of
course, the LAYER debacle and the 'pointsize' attribute of the already
deprecated FONT tag. These aren't in Transitional either, because they
represent competing alternatives to CSS, not just stopgap precursors.

Netscape didn't get any religion in this area until after Microsoft had
overturned their market share. Now of course it is Microsoft that can't be
bothered to correct and finish its CSS-1 implementation (except in the
relatively insignificant case of IE for Mac). Tantek Celik (MS' sole (or at
least most visible) representative in the W3C UI activity area) is a great
guy, but it is sadly significant that he does not work on Win32 stuff.

> Yes, but as always it's a compromise. The margin attributes
> didn't make it in, maybe because they were too new to be
> that widely used at the time; certainly because they're quite
> easy to replace with (much nicer) CSS. I guess no-one thought
> we'd still be stuck designing for Netscape 4.x in 2000.

We'll be stuck designing for it as long as we coddle its bugs and quirks.
Users of this software typically have no idea how expensive it is to
"support" its retrograde cussedness. Whenever I have any say in the matter,
I let NS4 eat rocks if it can't handle 1996-era W3C UI stuff. Have to start
educating end users.

--
Todd Fahrner

"But IE3 is almost completely extinct, whereas Netscape 4's undead corpse
still shambles about the earth wreaking a horrific vengeance upon the
living."  -- Matt McIrvin
Received on Wednesday, 13 December 2000 11:58:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:44 GMT