292066 - Page rendered incorrectly on american express web site - pushed too far to the bottom of the page. Works in other browsers

Status: VERIFIED DUPLICATE of bug 294201

(Core :: Layout, defect)

Description

[Chris]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050425 Firefox/1.0+
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050425 Firefox/1.0+

For the past couple of months, when using Firefox (latest nightly, always), my
American Express statement has been shifted to the bottom of the screen by about
an entire page. This doesn't happen in IE.

I've saved a local copy so that non-Amex members can see. Try opening the link
below in Firefox and IE:

http://rtcw.no-ip.org/test/American%20Express-Cards-Account%20Summary.htm


Reproducible: Always

Steps to Reproduce:
1. Visit http://rtcw.no-ip.org/test/American%20Express-Cards-Account%20Summary.htm
2.
3.

Actual Results:  
Page rendered incorrectly

Expected Results:  
Obvious lengthy block should be short

Comment 1

[Ben Fowler]

I read that as the region between "VIEW ACCOUNTS" and "My Account"'s being 
too tall.

Comment 2

[Ben Fowler]

Attachment: Testcase

Slightly shorter testcase which appears to still have the problem

Comment 3

[Uri Bernstein]

Attachment: minimal testcase

You'll have to forgive me for this testcase not looking anything like the
original page. I had to go through a lot of dynamically-generated content and
other fun JS stuff. But this is what it boils down to:

In Gecko 1.7 builds, the inner (blue) DIV is short - it shrink-wraps the text
"LOG IN". In new (1.8) builds, the inner div expands to the entire height of
the outer (red) div.

in the original page, the offsetHeight property of the inner DIV is then used
to resize other elements, which spreads the effect all over the page.

Comment 4

[Steve England [:stevee]]

For me, Uri Bernstein's new testcase in comment 3 looks the same for me in IE as
it does in FF (Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2)
Gecko/20050427 Firefox/1.0+)

Comment 5

[Uri Bernstein]

(In reply to comment #4)
> For me, Uri Bernstein's new testcase in comment 3 looks the same for me in IE as
> it does in FF (Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2)
> Gecko/20050427 Firefox/1.0+)

Hmmm... I'm on a Mac, so I couldn't compare to IE - but I now see that Safari
renderes my testcase like Gecko 1.8, although it renders the original site
correctly. So perhaps I took the wrong path somewhere during the reduction. I'll
try to backtrack and see if I can find where it really breaks.

Comment 6

[Uri Bernstein]

In Safari a completely different code path is being used - scriptflag is set to
false, and therefore alternate_menu() is called, not create_menu(). If I force
scriptflag to true, Safari shows a messed-up rendering for Ben's testcase.

Can someone please check what value scriptflag is getting in IE/Win (according
to the code, it seems that it should be "true", but I just want to verify). You
can test this by putting an alert statement inside alternate_menu() and see if
you get there.

Comment 7

[Boris Zbarsky [:bzbarsky]]

So... IE/Win and current Gecko render "minimal testcase" identically but render
"Testcase" differently.  There's a whole bunch of JS involved, and I wouldn't be
surprised if this all comes down to the site doing some sort of sniffing failing
to deal with the fact that we now render percent heights in quirks mode much
more like IE does.

Comment 8

[Carsten]

Duplicate of 294201?

Comment 9

[Boris Zbarsky [:bzbarsky]]

*** This bug has been marked as a duplicate of 294201 ***