Closed Bug 55623 Opened 19 years ago Closed 3 years ago

html 4.0 compatibility tests

Categories

(Core Graveyard :: Tracking, defect, P3)

defect

Tracking

(Not tracked)

RESOLVED INCOMPLETE
Future

People

(Reporter: buckizard, Assigned: chofmann)

References

()

Details

(Keywords: html4, meta, Whiteboard: [bae:20011129])

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; m18) Gecko/20001006
BuildID:    2000100608

	following tests failed for me when going through this test. 1.3.1   2.1.7   2.2.2
  2.2.3   2.5.1   4.1.2   6.1.3   6.1.9   6.1.10   6.2.6   6.2.9   6.3.4   6.4.1
  6.4.2   8.1.2   11.2.1  11.3.1  11.4.1						

Reproducible: Always
Steps to Reproduce:
	used stated build and os plus was in modern theme						

Actual Results:  	each one failed the specified test						

Expected Results:  		i would think pass					

	don't know if i need special plug-ins for some of these but dont think so. test
was too long.  my computer is celeron600, 32 ram, voodoo 3 16mb card, plenty
free hd, 1024 x 768 w/ high color (16 bit). anyway hope this helps somewhere
OK, I went through the tests and tried to find the bugs associated with each
test failure. Here's what I found:

1.3.1: The problem code was this CSS style, which yielded no background at all:
P.image1back {
  background-image : url(../images/liteblue.gif);
  background-repeat : no-repeat;
  background-position : top left;
  background-attachment : fixed;

Another class using the same background image worked correctly in the previous test.

2.1.7: This test failed because of a lack of BMP support (bug 18502).

2.2.2,2.2.3: These tests tried to display scaled images by creating several IMG
tags pointing to the same image, but with different width and height tags. I
checked the bugs in the ImageLib component, but failed to find a really good match.

2.5.1: This test uses an image map to redirect to another part of the page. This
is supposed to test client-side image map support, but I'm pretty sure the real
reason this fails is the recent page anchor regression (bug 55430).

4.1.2: Works for me on Linux build 2000-10-06-21.(Changing color of visited link).

6.1.3: Attempted to center tables by placing the tables in DIV's and setting
different text-align styles on the surrounding DIV. It didn't work for me, I'm
not sure why.

6.1.9: Works for me. (Highlighting of TH cells).

6.1.10: Creates several tables, each with the "rules" attribute set differently,
i.e. rules=cols. None display any rules at all. I think this relates to bug 21076.

6.2.6: Table and cell border colors are not affected by border-color attributes.
I'm not sure why this doesn't work. All that's happenning is that TABLE and TD
have style="border-color:red" set.

6.2.9: Um, there is no test 6.2.9.

6.3.4: rules=groups is set on a TABLE element, yet no rules are displayed
between groups. Again, I think this is bug 21076.

6.4.1: This test contains a table column group with a style attribute of
background:red, yet the background color of the column group is unaffected. I
believe this is bug 4510.

6.4.2: Table rules set to only draw around column groups. No rules appear, and
again I suspect bug 21076.

8.1.2: This is a VBScript test. Enough said.

11.2.1: This involves Java support.

11.3.1: This page tries to embed another HTML document within itself using the
OBJECT tag, but nothing is displayed where the document should appear.

11.4.1: Tries to use the OBJECT tag to play a sound. I don't have a sound card
and thus can't tell.
2.2.2,2.2.3: 
The style value has a '=' between the property and the value as in:
   style="width=100;height=100"
Is this valid? If the '=' are changed to ':' the images are scaled properly.

6.2.9.
Has the following html, which isn't used correctly:
<COLGROUP>
<COL align="char" ><COL align="char" char=".">
</COLGROUP>
Possibly covered by Bug 51693
"width=100" is illegal in CSS, on two counts. One, '=' should be ':'. Two, the
units are missing. Doesn't sound like a very good test suite if it has errors
that basic in it...
6.1.3 
Has the following html, which validate to 4.0
<html>
  <body>
    <DIV style="text-align : right">
      <TABLE>
        <CAPTION>Right Aligned</CAPTION>
        <TR>
          <TD>XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX</TD>
        </TR>
      </TABLE>
    </DIV>
  </body>
</html>

The words 'Right Aligned' are expected to be right-aligned, i.e. inherit the
align property from its DIV parent.
If the table components are removed, the text is aligned correctly.
As pointed out by Scott Brodmerkle, 6.2.6 is a table border colour problem.
same html that demonstrates this problem is:
<html>
<body>
<TABLE border=3 style="border-color:red;">
<TR> <TD>Red Border</TD> </TR>
</TABLE>
</body>
</html>

The border of the table should, obviously, be red.  On linux the colours is the
default colour.  This may have only recently broken, as I remember seeing this
working before on WinNT.
2.5.1 now works [Named anchors are not working]
Someone should at least glance at this. Marking as New.
Status: UNCONFIRMED → NEW
Ever confirmed: true
assigning to karnaze
for QA: could you look at the text below and create a bug for each single 
problem not covered by a bug?
Assignee: pollmann → karnaze
QA Contact update
QA Contact: petersen → amar
Moving to m1.0
Status: NEW → ASSIGNED
Target Milestone: --- → mozilla1.0
Here's a list of the state of each of the tests.  Can someone look at 11.2.1,
11.3.1 and 11.4.1.  Everything else seems to either Work or be covered.


1.2.1 (blinking doesn't work). Covered by Bug 58788

1.3.1 I've created Bug 84302 for this.

2.1.7 covered by bug 18502

2.2.2 } From Ian: "width=100" is illegal in CSS, on two counts. 
2.2.3 } One, '=' should be ':' Two, the units are missing.  => Invalid test

2.5.1 WFM

4.1.2 WFM

6.1.3 Not sure if it's a valid test.  I've logged bug 84305 for this.

6.1.9 WFM

6.1.10 Covered by Bug 21076

6.2.6 I've logged Bug 84307

6.2.9 Covered by Bug 51693

6.3.4 Covered by Bug 21076

6.4.1 Covered by Bug 4510

6.4.2 Covered by Bug 21076

8.1.2 VBScript => [INVALID]

11.2.1 Java - can someone look at this

11.3.1 can someone look at this

11.4.1 sound - can someone look at this
I'm changing the component to Browser-General, as there are multiple components
involved here.
Also changing Platform and OS to All.
Component: HTMLFrames → Browser-General
OS: Windows 98 → All
Hardware: PC → All
Bug 84302 covering test 1.3.1 has been marked as invalid.
Following commments above, I'm adding those bugs listed as dependencies so
they're easier to track. 
In addition, bug 41368 might also be of interest.

I'll try and locate bug numbers for some of the remaining issues when I get a
chance. 

Adding html4 keyword
Depends on: 4510, 18502, 21076, 51693, 58788, 84305, 84307
Keywords: html4
not specifically a table bug that I can tell. reassiging.
Assignee: karnaze → asa
Status: ASSIGNED → NEW
QA Contact: amar → doronr
Assignee: asa → attinasi
Component: Browser-General → Layout
QA Contact: doronr → petersen
Sending to browser -> layout (see related bug 41368) and reassigning.
No longer depends on: 58788
Are these the remaining issues?

11.2.1 Java - can someone look at this

11.3.1 can someone look at this

11.4.1 sound - can someone look at this

I only put the comments about 11.2.1, 11.3.1 and 11.4.1 as I'm not exactly a fan
of sound and java on web pages, especially when they don't work.
I'm not sugesting there is a problem in Mozilla.



Target Milestone: mozilla1.0 → mozilla1.2
I am not clear how to access the test suite: reporter can you give me a set of 
steps that would get me there?

Also: object element does successfully load and play sound, if the object 
element is coded correctly, java is loaded. As of loading another html document 
within itself, I would have to see the test.

It seems like all other issues are covered by other bgs. If a specific issue is 
encountered with the embedding test case, please open a new bug, either point to 
an appropriate URI or attacha  test case with the code.

Leaving this bug open and marking as a meta bug
Keywords: meta
Whiteboard: [bae:20011129]
Target Milestone: mozilla1.2 → Future
I have added a more detailed comment to bug 4510.
To summarize:

Mozilla1RC1 does not correctly display the sample table in the W3C HTML4.01
specification.  This sample is strictly HTML with NO style attributes.
IE5 does correctly display this sample table.

Mozilla1RC1 does not allow <td> or <th> to inherit HTML4.01 attributes
(including style) from <col/> or <colgroup>.  According to W3C HTML4.01
specification it should.  IE5 does.

Mozilla1RC1 does not attempt to paint the background of <col/> or <colgroup>. 
IE5 does in a way that violates CSS2 only.  At least it tries.  Read my comment
in bug 4510.

I think this should be getting a lot more attention.  I hope it will as soon as
Mozilla 1.0 is final since this is an HTML4.01 issue.
Moving to Tracking.
Assignee: attinasi → chofmann
Blocks: html4.01
Component: Layout → Tracking
QA Contact: petersen → chofmann
Summary: html 4.0 compatability tests → html 4.0 compatibility tests
Marking all tracking bugs which haven't been updated since 2014 as INCOMPLETE.
If this bug is still relevant, please reopen it and move it into a bugzilla component related to the work
being tracked. The Core: Tracking component will no longer be used.
Status: NEW → RESOLVED
Closed: 3 years ago
Resolution: --- → INCOMPLETE
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.