Re: em

Eric A. Meyer wrote:
> 
> At 11:23 -0700 06/20/00, Matthew Brealey wrote:
> 
> >Todd Fahrner wrote:
> >   > in many cases,
> >   > authors quite reasonably want a unit system that is relative to user
> >   > preferences alone, *not* dependent upon the document structure. Such
> >   > a system already exists in CSS1, however: the font-size keyword
> >   > system (xx-small - xx-large). It's just that WinIE previously
> >   > destroyed the usefulness of this system by implementing "small"
> >   > rather than "medium" as the initial value, as the spec requires.
> >
> >Very true, but doesn't (even the latest version of) Mac IE do this as
> >well?
> 
>     In "quirks" mode, yes, it does. 

Quirks mode=transitional dtd or no dtd

In other words, for 99% of the pages on the web, and for *all*
commercial pages (commercial sites are *not* written in strict HTML),
the quirks mode is triggered. This is much, much worse than leaving the
bug as it is, because if the bug is left in, there is at least hope that
it some point in the future it will be fixed, but by fixing it only for
those tiny number of pages that include the appropriate doctype, the
bug's status is secured for all time and the web is a worse place for
it.

To make matters worse, the arena in which takeup [of CSS] has been most
limited, namely commercial pages, is the very one that it is affected:
the only pages that are fixed are those written in strict HTML where the
authors had probably fixed it beforehand any way.

But more than that doctype detection is fundamentally misconceived. The
statement at the top of a document that 'this document conforms to this
DTD' in no way entails 'therefore please give me lots more bugs'.

Even assuming that somehow it does, how does an HTML viewer know whether
a DOCTYPE implies bug mode. For example, some WYSIWYG tools (almost of
all of which generate bad code) stick their own doctype in at the top of
the page. Clearly, if ever a quirks mode should be used, this is such a
case, so the browser must use quirk mode if it does not recognised the
dtd. 

But then anyone using a new DTD, such as that of the new ISO/IEC
standard, which is just about the strictest DTD around, will find their
ultra-strict page will be rendered in quirks mode because the browser
was released before the dtd.

> That's because it's in a bugwards
> compatibility mode, in order to avoid breaking old pages.  

Fact (well ok assertion, but it's an assertion I challenge anyone to
disprove): no pages would be broken by MS biting the bullet and fixing
this bug, because:

(a) because of the bug all such pages are broken anyway - they'll either
look too 'big' if authored for (correct in this, if nothing else)
Netscape, or else too 'small' if authored in the (broken) MSIE
(b) because of the above there aren't (m)any pages that use the
keywords, so it doesn't matter anyway.

> In  "strict" mode, IE5/Mac does as it should: 

In other words, 99% of the time it does not.

> This lets standards-compliant authors and pages act as expected.

HTML transitional is no more or less a standard than HTML strict
(arguably more because it has greater acceptance). 

By encouraging the perpetuation of these bugs CSS remains badly broken.
Imagine the scene: 'Hey Mr. Yang [Jerry], why don't you use style sheets
on your site [Yahoo!]?' 
'OK then, how about some small text.' 
'Oh no, sorry Mr. Yang, if you want small text you have to remove all
the HTML formatting from your site, stand on your head and then stick
this DOCTYPE up at the top'
'Forget it.'

There is very little in relation to HTML and CSS that is more arcane
than the concept of the DOCTYPE, and to expect people to get that
'right' before the browser will even condescend to do something as
trivial as to get font sizes right is to condemn CSS to the sidelines.

PS. This is not intended as a flame against you; neither is it an attack
on MSIE in particular; all of my pages now use a strict doctype (against
which the pages do not validate) as a result of the perpetuation of a
Netscape 4 table bug in Mozilla (and hence Netscape 6, and thereby
condemning us to at least another 3 years of a broken web (of course the
concept of a quirks mode only encourages (a) the bugs *never ever* to be
fixed, and (b) the reliance on invalid behaviour)).
-----------------------------------
Please visit http://RichInStyle.com. Featuring:
MySite: customizable styles.         AlwaysWork style 
Browser bug table covering all CSS2 with links to descriptions.
Lists of > 1000 browser bugs         Websafe Colorizer 
CSS2, CSS1 and HTML4 tutorials.      CSS masterclass 
CSS2 test suite: 5000++ tests and 300+ test pages.

Received on Tuesday, 20 June 2000 08:43:41 UTC