#133 dimensions patch for IE6 using doctypes

[Andy Short]

When looking for the bodys width/height in IE6 using:

DynAPI.document.getWidth()
DynAPI.document.getHeight()

if the document has a valid doctype either XHTML or 4 +
IE6 runs in a CSS1 compatible strict mode when certain
doctypes are present. In that case the documentElement
represents the canvas, not the body element therefore
findimensions() was returning 0 using
this.doc.body.clientHeight and a strict/xhtml doctype

added a check for wether the document was in
strictmode or not and used
this.doc.documentElement.clientHeight

basically if you write HTML or XHTML using doctypes
and you use the documents body's width/height in your
code you will need this for correct display in IE6

[Andy Short]

replacement dyndocument.js

[Richard Bennett]

Logged In: YES
user_id=164855

Hi,
Have you any idea if IE5/MAC has the same difference when
running under "strict" mode?
I added a test for this in the /tests directory in CVS,
here:
<wrap>
http://cvs.sourceforge.net/cgi-
bin/viewcvs.cgi/~checkout~/dynapi/dynapi/tests/doctype_clien
tHeight.html?rev=HEAD&content-type=text/html
</wrap>

Wouldn't it be best to do the doctype-sniffing in
browser.js?
If so, like this:
if(is.strict&&is.IE6)
or like this? :
if(is.IE6Strict)?

Richard.

[Andy Short]

Logged In: YES
user_id=176079

With regard to the strict mode on a mac, not having access
to a mac i couldnt test this behaviour in ie5 , i believe
this behaviour only applies to ie6.

re the check, i put the check within the dyndocument to try
and keep the change to one file, it might be worth adding
this check to the browser.js and use if(is.ie6strict)
method as mozilla isnt affected and believed ie5 on a mac
isnt either, tho this should really be confirmed before
committing these changes to the cvs.

thanks
... sh0rtie

[martin ström]

Logged In: YES
user_id=46384

i've tested on mac ie5:

document.body.clientHeight = 466
document.compatMode = undefined
document.documentElement.clientHeight = undefined

which leads to this is ie6 only.

i think a is.strict is best while other browsers
go into strict mode aswell, and then we can use
is.strict to identify different modes.

[Richard Bennett]

Logged In: YES
user_id=164855

Thanks for testing that.

<<i think a is.strict is best while other browsers
<<go into strict mode aswell, and then we can use
<<is.strict to identify different modes.

That would mean we would have to do if (is.strict&&is.IE6)
Is that the best way?
Bear in mind that the code to test for strict mode in
IE5/Mac, and Mozilla is a lot more involved than for IE6.
Should we add all that in now, or just the IE6 part to
start with?

Also, I noticed that HEX colors have to be started with the
# sign, or they don't show in IE6 Strict mode.
Something else to keep in mind.

[martin ström]

Logged In: YES
user_id=46384

if this.strict could test for mozilla's strict mode aswell
i think this would be the best.

can't we use the test you had on
http://www.richardinfo.com/case_studies/doctypes.html

to detect which mode ie5mac, ie6 or mozilla is in?
and add this to browser.js

[Max Nokhrin]

Logged In: YES
user_id=373738

THANK YOU THANK YOU THANK YOU!!!

There is also a bug in IE6 that returned an incorrect
clientHeight and clientWidth. Your file fixes that error!

Thank you very much!

[Nobody/Anonymous]

Logged In: NO

The Alert box reads:

document.body.clientHeight = 438
document.compatMode = undefined
document.documentElement.clientHeight = undefined

Using IE 5.2.1 for Mac OS X 10.1.5 on
an iMac at 800 x 600 screen resolution

[Nobody/Anonymous]

Logged In: NO

Mac os 10.2, IE 5.2.2...

document.body.clientHeight = 629
document.compatMode = undefined
document.documentElement.clientHeight = undefined

629 is in fact the height of the window's rendering area.