why does this happen?

[Ken Williams]

can someone look at this:
http://sites.extremehosting.ca/misc/test.html

its a thin line in IE 6/7 which I expect. In Firefox its a thick bar.
why? anyone?

(e-mail address removed)

 

[dorayme]

can someone look at this:
http://sites.extremehosting.ca/misc/test.html

its a thin line in IE 6/7 which I expect. In Firefox its a thick bar.
why? anyone?

(e-mail address removed)

You are seeing the effect of line-height in the following:

<http://dorayme.890m.com/alt/howLineHeightAffects.html>

that models your case and you can let me know if it is different in IE
to FF. If so, it is almost certainly IE simply getting things wrong yet
again.

 

[dorayme]

You are seeing the effect of line-height in the following:

<http://dorayme.890m.com/alt/howLineHeightAffects.html>

that models your case and you can let me know if it is different in IE
to FF. If so, it is almost certainly IE simply getting things wrong yet
again.

.... or you have a different text size setting on your IE browser from
your FF. The point being that you could well be seeing the effect of the
space allowed for text, to put it simply, where your gif is, the gif
being inline.

 

[dorayme]

In quirks mode Firefox actually copies IE to some extent-- by making the
cell no higher than the image.

I guess, depending on the exact meaning of quirks mode. Nothing but 4.01
Strict HTML and XHTML 1.0 Strict makes the image subject to line-height
and text size settings on my tests.

 

[dorayme]

I use "quirks mode" to mean anything except strict, although there are
degrees of quirks modes in various browsers.

Yes, I know. I have always rather liked your usage. Simple and useful
and unfussy! But I am not quite sure I use it so.

As I have understood it, I think of quirks mode as being what happens
when there is no particular mode dictated by a well formed dtd. In the
case of no doctype, the browser does the best it can and draws on its
memory of past glories, it sweats and puffs and sighs but comes up with
something in the end and then rushes off to the pub to forget what it
has just done leaving the poor author and his audience to face the
consequences.

Whereas "standard" modes are "triggered" by various doctype
declarations, not necessarily, the two I mentioned in my post. In other
words, a dtd, well formed enough in expression, be it transitional or so
called strict, is not quirks though it might result in some similar
effects as quirks.

There is a giant table somewhere in one of the possible worlds in which
all the variations of what actually happens to browsers in all the
possible combinations are outlined. And we need to have some shortcuts
to make sense of it all. Yours is nice and simple. Perhaps if it was a
little more complicated (but not much more) it would also be good.

 

[dorayme]

not necessarily, the two I mentioned

not necessarily the two I mentioned

(...time for a movie I can see, been a long day. The Illusionist with
Edward Norton. I trust that is OK with David Segall.)

 

[Bergamot]

There's only one standards mode as far as I can see

FYI, there is quirks mode, standards mode and "almost standards" mode.
http://hsivonen.iki.fi/doctype/

 

[dorayme]

[...]
In quirks mode Firefox actually copies IE to some extent-- by making the
cell no higher than the image.

I guess, depending on the exact meaning of quirks mode. Nothing but 4.01
Strict HTML and XHTML 1.0 Strict makes the image subject to line-height
and text size settings on my tests.

I use "quirks mode" to mean anything except strict, although there are
degrees of quirks modes in various browsers.

[...]
Whereas "standard" modes are "triggered" by various doctype
declarations, not necessarily, the two I mentioned in my post. In other
words, a dtd, well formed enough in expression, be it transitional or so
called strict, is not quirks though it might result in some similar
effects as quirks.

There's only one standards mode as far as I can see because CSS 2.1 only
defines one standard, except in a couple of places it suggests slightly
different things for XHTML documents (root element backgrounds and what
to do about overlapping spanning table cells for example).

But I don't believe browsers choose "XHTML mode" based on the DOCTYPE,
but instead on the Content-Type.

As I said, I liked your simplification. Perhaps we are at cross
purposes?

I was looking at

<http://en.wikipedia.org/wiki/Quirks_mode>

the other day. The table in it looks like it could be updated but it
looks like a pretty sensible page.

I get the impression that if you don't put a doctype at all you are sure
to be in unknown territory unless you know all the quirks that all the
browsers try to take account of. Whereas if you do put in a doctype that
is not one of the Strict ones, the situation is not quite as bad. Which
is why I am hesitant to adopt your simplification.

 

[Roy A.]

In quirks mode Firefox also adopt the broken box model. In "almost
standards mode" they render everything as in "standards mode" except
from images inside an td element. Other browsers are following
Mozilla.

This is only a problem if you use an PhotoSoup layout with sliced
images. If you use tabular in formations with images, let say an icon,
this is not a problem:

<td><img src="icon.gif" height="16" width="16" alt="New!"></td>
<td>Tabular info as text</td>

Now you can use 'vertical-align' on both the td element and the img
element. E.g there are no difference between "standards mode" and
"almost standards mode".

That only affects PhotoSoups layout with sliced images, so I don't see
your problem. XTHML transitional with Content-Type "application/xhtml
+xml" (or extension .xht or .xhtml) is rendered in "standards mode".
IE 6 and IE 7 is always rendering pages in what Mozilla is calling
"Almost standards mode", as long as you have a DOCTYPE with the word
"html" inside:

<!DOCTYPE anything with the world "html" inside>

In IE 6 and IE 7 this DOCTYPE is rendered in their "standards mode":
Just past "javascript: alert(document.compatMode)" in the address bar
an see for yourself.

I use "quirks mode" to mean anything except strict, although there are
degrees of quirks modes in various browsers.

All browsers is behaving different in "quirks mode". In "standards
mode" and "almost standards mode" all are following the standards.
There is only
one differences between "standards mode" and "almost standards mode"
and that is how sliced images in tabular layouts is rendered within
table cells. In tables with tabular data, IE 6, IE 7 and Firefox are
(trying) to follow the standards.

The difference between "strict" and "transitional" DOCTYPEs is almost
of no importance. They are all rendering in "standards mode" with only
that exception.

In the specifications there are only two modes: "BackCompat" (quirks
mode) and "CSS1Compat" (standards mode/almost standards mode).

 

[John Hosking]

In the specifications there are only two modes: "BackCompat" (quirks
mode) and "CSS1Compat" (standards mode/almost standards mode).

Whose specifications do you mean? "BackCompat" is only mentioned by the
W3C in its HTML5 work. Do you mean to say that HTML5 is relevant at all?
Or have I missed some earlier actual use of "BackCompat" elsewhere? Do
any browsers really know about these modes?

 

[dorayme]

There are a few other differences too. See

http://www.tidraso.co.uk/misc/quirks-mode

On your test 3 in Safari 3 on Mac (but not in FF2), the purple is there
in all 3 modes. And in test 3, all centre the text the same in all 3
modes in Safari 3 (but not on FF2).

 

[Roy A.]

There are a few other differences too. See

http://www.tidraso.co.uk/misc/quirks-mode

for some I have encountered. Test 2 is pretty much the one you're
talking about, but there are other differences in Transitional as you
can see, and they're not all just default stylesheet.

I cant argue with that. I get the same results. I suppose it is this
page that led me jump to conclutions:
http://developer.mozilla.org/en/Gecko's_"Almost_Standards"_Mode

"Almost standards" rendering mode is exactly the same as "standards"
mode in all details save one: the layout of images inside table cells
is handled as they are in Gecko's "quirks" mode, which is fairly
consistent with other browsers, such as Internet Explorer. This means
that sliced-images-in-tables layouts are less likely to fall apart in
Gecko-based browsers based on the rendering engine found in Mozilla
1.0.1 or later when in either "quirks" or "almost standards" mode."

....

"Other than this one difference, "almost standards" and "standards"
modes are exactly the same in terms of layout and other behaviors."

Well, that is not enterly true. Thanks for the links.

 

[Roy A.]

Whose specifications do you mean? "BackCompat" is only mentioned by the
W3C in its HTML5 work.

It seems to be my own specifications. No this isn't in any
recommendations. It is however a convention like DOM0 or innerHtml. No
excuse, the word 'specifications" is completely wrong.

Do you mean to say that HTML5 is relevant at all?

HTML 5 is relevant. It will help IE to destroy the "xhtml 1.1" path,
like they need any help. And like other vendors don't agree. And
without "xhtml 1.1", we have only one path. There is going to be a
html recommendation after "HTML 4.01", but that isn't relevant to me
before it becomes an recommendation. I'm quite happy as long as I can
serve my html as xml if I need too.

Or have I missed some earlier actual use of "BackCompat" elsewhere?

I don't know, I think I have missed it too. I find lots of references
to "BackCompat" and "CSS1Compat", but I don't find the actually
source. And I don't find those terms defined from whomever invented
them (maybe it's defined by Netscape or Microsoft).

Do any browsers really know about these modes?

Yes they do. If your browser support the pseudo-protocol 'javascript'
in the address bar, you can copy and past this sentence "javascript:
"alert(document.compatMode)" into the address bar. If not you can use
this sentence in the onload attribute:
onload="alert(document.compatMode)".

Thank you for commenting on my mistake. I'm not only misleading
myself, but also misleading others.

Sorry. Possessive don't have apostrophes in my language, too. I shall
try to correct that.

 

[Roy A.]

FYI, there is quirks mode, standards mode and "almost standards" mode.http://hsivonen.iki.fi/doctype/

I don't care. As long as the browser is showing my MARQUEE elements.
And they all are showing
it whichever DOCTYPE i'm using. Henri Sivonen, who is that?

 

[John Hosking]

It seems to be my own specifications. No this isn't in any
recommendations. It is however a convention like DOM0 or innerHtml. No
excuse, the word 'specifications" is completely wrong.

Okay, I was just checking. I'm behind the curve on more things than I'm
ahead on, so I just figured this might be another one of the former.

I don't know, I think I have missed it too. I find lots of references
to "BackCompat" and "CSS1Compat", but I don't find the actually
source. And I don't find those terms defined from whomever invented
them (maybe it's defined by Netscape or Microsoft).

MS seems capable of redefining or at least bending the English language:

"When Internet Explorer 6 or later is not in standards-compliant mode,
it is in /compatibility/ mode."

from http://msdn.microsoft.com/en-us/library/bb250395.aspx

Yes they do. If your browser support the pseudo-protocol 'javascript'
in the address bar, you can copy and past this sentence "javascript:
"alert(document.compatMode)" into the address bar.

OK, javascript:alert(document.compatMode) in IE6 gives me "BackCompat".

Thank you for commenting on my mistake. I'm not only misleading
myself, but also misleading others.

Tsk, tsk. Naughty boy. (I do admire your consideration.)

Sorry. Possessive don't have apostrophes in my language, too. I shall
try to correct that.

Well, I wasn't aware I was admonishing you for any actual usage errors
you'd made; it's just in my current sig.

 

[dorayme]

On 5 Sep, 12:34, Ben C <[email protected]> wrote:

[...] ....

There are a few other differences too. See

http://www.tidraso.co.uk/misc/quirks-mode

On your test 3 in Safari 3 on Mac (but not in FF2), the purple is there
in all 3 modes. And in test 3, all centre the text the same in all 3
modes in Safari 3 (but not on FF2).

"And in test 3" should have been "And in test 4"

 

[Roy A.]

Okay, I was just checking. I'm behind the curve on more things than I'm
ahead on, so I just figured this might be another one of the former.





MS seems capable of redefining or at least bending the English language:

"When Internet Explorer 6 or later is not in standards-compliant mode,
it is in /compatibility/ mode."

fromhttp://msdn.microsoft.com/en-us/library/bb250395.aspx

It's cind of crap isn't it? They talk about 'The !DOCTYPE "Switch"',
but their own !DOCTYPE switch is:

<!DOCTYPE anything with the word 'html'>

And there you go, "Standards mode" (or what they call 'standards
mode').

And what do they do if you deliver a perfectly valid "HTML 4.01"
strict
document to their IE 8 beta? Yes, they are displaying an icon that
sugest that you document is broken:

http://news.slashdot.org/article.pl?sid=08/08/30/169218&from=rss