66772 - WRMB: Weird page display caused by overlapping or unclosed tags

Status: RESOLVED WORKSFORME

(Core :: DOM: HTML Parser, defect, P1)

Description

[Alexey Chernyak]

User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; m18) Gecko/20010126
BuildID:    2001012620

Mozilla seems to have trouble displaying Nero download page
http://www.nero.com/en/download.htm the way it was intended to be. Looks REALLY
weird.
The page has no <!DOCTYPE> tag nor any CSS styles in it.
Strangely IE and NS4 have no problems displaying it.
I didn't analyse the page code to try to figure out why it happens.
Hope there's someone who has time and patience to do that. :o)

Comment 1

[Oliver Klee]

Reporter, could you attach screenshots of how the page looks in Mozilla and of
how it looks in NN/IE?

Comment 2

[Alexey Chernyak]

Attachment: Mozilla screenshot

Comment 3

[Alexey Chernyak]

Attachment: Netscape v4.76 screenshot

Comment 4

[Alexey Chernyak]

Attachment: IE v5.5 SR1 screenshot

Comment 5

[Bernd]

The page was better 2001-01-11. The nested tables have been wrong (this has been 
fixed with bug 13553), the page has significantly regressed on 2001-01-12. I am 
looking further on it

Comment 6

[HJ]

I belief that fix was about bug 63445

Comment 7

[Bernd]

This bug was introduced by harishd with his fix for bug 64622
. Backing out the changes fixes the bug. Reassigning. Rising priority.

Comment 8

[harishd]

On my build ( with the patch for bug 64622 ) I don't see this problem. Also, 
2001-01-28-21-Mtrunk does not exhibit this problem.  Which build are u using?

Comment 9

[kmike]

i'm seeing this on 2001-01-28-21-Mtrunk linux build.

Comment 10

[Bernd]

Attachment: weird display 2001-01-29

Comment 11

[Bernd]

Attachment: current cvs build wihtout patch for bug 64622

Comment 12

[Bernd]

Harish, if you convince me, that the attachment from 8:55 is not weird in 
comparison with IE5, Nav4.6 and my attachment from 10:52, I will send you out a 
beer immediately. If your trunk build shows something else could you please 
attach a screenshot.

Comment 13

[harishd]

Attachment: screen shot of what I see.

Comment 14

[HJ]

Build 2001012920 on WinNT4 looks like shit! Try to use a link on that page!

Comment 15

[Alexey Chernyak]

Just tried it on Build 2001013004 Win2K
It actually looks a little bit better than Build 2001012620
I did some code analysing on the page, and it is <FONT> tag indeed that
triggered these problems. On the screenshot you can see 2 problems: inner tables
being aligned weirdly, and the main outer table being broken. The inner tables
problem seems to be fixed.
The page itself I must say is very buggy, with tags overlapping and not being
closed. But that's what we have quirks mode for. Competitive browsers handle
this page very gracefully.
Here is the narrowed down bit of code with overlapping <FONT> tag which I
believe causes the problem.
   <TD>
    <FONT FACE="Arial"><B>Free Nero 5.0 Demo</B><BR>
    <FONT SIZE=-1>
    TEXT
    <TABLE bordercolorr="#000080" border=1 align=right cellPadding=2
cellSpacing=0 width=400>
    <TR><TD>TEXT</FONT></TD></TR>
    </TABLE>
    </FONT>
   </TD>

Comment 16

[Alexey Chernyak]

Attachment: screenshot on Build 2001013004

Comment 17

[harishd]

Thanx alexeyc for narrowing down the problem.

Comment 18

[Stephan Niemz]

updating summary

Comment 19

[Alexey Chernyak]

Attachment: testcase

Comment 20

[Alexey Chernyak]

Attached a simple testcase which is a narrowed down code from Nero page.
if this bug is corrected both tables should look identical.

Interestingly NSv4.76 has problems displaying inner tables of this testcase
because they are align=right, while it displays the actual Nero page, from which
this code been taken out, correctly. weird. :o)

Comment 21

[harishd]

*** Bug 69576 has been marked as a duplicate of this bug. ***

Comment 22

[harishd]

This bug has been marked "future" because the original netscape engineer working 
on this is over-burdened. If you feel this is an error, that you or another 
known resource will be working on this bug,or if it blocks your work in some way 
-- please attach your concern to the bug for reconsideration. 

Comment 23

[Alexey Chernyak]

*** Bug 82644 has been marked as a duplicate of this bug. ***

Comment 24

[rubydoo123]

adding WRMB in summary, adding topembed, cc chofmann, marek

Comment 25

[Andreas M. "Clarence" Schneider]

*** Bug 95918 has been marked as a duplicate of this bug. ***

Comment 26

[Andreas M. "Clarence" Schneider]

Simple testcase: attachment 46353 [details]

Comment 27

[Alexey Chernyak]

Attachment: simplified testcase

Comment 28

[harishd]

*** Bug 56245 has been marked as a duplicate of this bug. ***

Comment 29

[harishd]

Refer to the patch in bug 56245. I'm still working on a better patch due to
inconsistencies in userdefined tag behavior.

Comment 30

[harishd]

--> P1

Comment 31

[Chris Petersen]

*** Bug 96079 has been marked as a duplicate of this bug. ***

Comment 32

[harishd]

I went through all ( I think! ) the duplicates and testcases and it looks like
the fix for bug 56245 has done the trick. What should we do with this bug now?
Mark it   FIXED?

Comment 33

[Alexey Chernyak]

yeah, mark it fixed.
what does WRMB mean?

Comment 34

[Jacek Piskozub]

I believe WORKSFORME is more adequate as there was no specific fix for this bug.

Anyway, WFM with a fresh CVS build (WindowsME) made at 2001082418. Marking
accordingly.

Comment 35

[harishd]

WRMBs are META bugs. I'm not sure if this should be marked WFM. I'm going to
reopen for now.

Comment 36

[Jacek Piskozub]

There are 24 META bugs and 3 WRMB bugs (89245, 93003, 93560) marked WORKSFORME.

Marking WORFSFORME again on the WFM comments  and the existence of the above
mentioned three bugs. If you disagree with this evidence, of course you are free
to reopen the bug.

Comment 37

[Andreas M. "Clarence" Schneider]

*** Bug 97002 has been marked as a duplicate of this bug. ***

Comment 38

[Chris Petersen]

Marking verified on the Sept 28th branch build.