9191 - Default value of border-collapse incorrect

Status: RESOLVED INVALID

(Core :: CSS Parsing and Computation, defect, P2)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Tested on: mozilla.org 1999-06-30-08-M8, Linux.

Test case to be attached.

The default value of the border-collapse property, per CSS2, is collapse.  This
is not happening.  It would be acceptable (although not advisable, I think) for
ua.css to set border-collapse: separate on table.  However, that is not what is
happening.

The test case to be attached shows that the separated borders model is being
used for both table and <div style="display: table"> (using classes...).  If the
separated borders model were used, the first two tables would look the same as
the last two.

I think the border and frame attributes of table (only some values of frame)
should set border-collapse to separate.  It shouldn't be that by default.

I highly recommend that you wait until bug 9024 is fixed before fixing this bug,
since border-collapse:collapse on empty tables can lead to instability (there
are a bunch of asserts on empty tables with border-collapse:collapse, I think).

The display glitch in the bottom row of the test case may be related to bug
2886, but may be separate.  Someone (I?) should investigate further and file
another bug if needed...

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: test case described in text of bug report

Comment 2

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I'm reopening this because I believe (from the newsgroups) that the fix was
backed out to fix bug 11441.  As I said before, I think this should wait until
bug 9024 (and any other collapsing border crashes) are fixed.

I think when you fix this, you should change the style mapping code for the
border attribute and some/all values of the frame attribute to set border-
separate: separate, and leave the HTML default as collapse.

Comment 3

[Peter Linss]

Deferring to M10

Comment 4

[Peter Linss]

Moving non-beta 1 items to M15

Comment 5

[Pierre Saslawsky]

Reassigning peterl's bugs to myself.

Comment 6

[Pierre Saslawsky]

Accepting peterl's bugs that have a Target Milestone

Comment 7

[Pierre Saslawsky]

Pushing my M15 bugs to M16

Comment 8

[Pierre Saslawsky]

Moving table bugs to M15 just to see how many we have.

Comment 9

[Hixie (not reading bugmail)]

Migrating from {css2} to css2 keyword. The {css1}, {css2}, {css3} and {css-moz}
radars should now be considered deprecated in favour of keywords.
I am *really* sorry about the spam...

Comment 10

[Pierre Saslawsky]

Block-moved to attinasi the bugs related to style inside tables

Comment 11

[Matthew Brealey]

> It would be acceptable (although not advisable, I think) for
> ua.css to set border-collapse: separate on table.  

Necessary in fact, because of the way that HTML tables work - cellspacing, etc.

Comment 12

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

No, the alternative is to set it in the attribute mapping code, but only do so
for some border styles.

Comment 13

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Er... values of the rules or frame attributes, not border styles.

Comment 14

[Marc Attinasi]

Karnaze needs to fix the table code to deal with collapsed borders better before 
we can change the default to collapse. Once we can change the default in 
html.css to 'border-collapse:collapse' (or better yet, just remove the rule) 
then we will look at this further to see what the story is with the DIV that has 
a display:table rule. Moving to M16 to match the dependent bug.

Comment 15

[Marc Attinasi]

Moving to M17 to match dependent bug 22897 

(I cannot change the default until collapsed borders work correctly).

Comment 16

[Marc Attinasi]

Now depends on bug 41262 since 22897 was marked a dup...

Comment 17

[Marc Attinasi]

Easy fix once dependent bug is fixed

Comment 18

[Marc Attinasi]

Since bug 41262 was nsbeta3-'d, this bug is getting nsbeta3-'d too. Will change 
the default when collapsing borders work correctly.

Comment 19

[Hixie (not reading bugmail)]

Netscape's standard compliance QA team reorganised itself once again, so taking 
remaining non-tables style bugs. Sorry about the spam. I tried to get this done 
directly at the database level, but apparently that is "not easy because of the 
shadow db", "plus it screws up the audit trail", so no can do...

Comment 20

[Hixie (not reading bugmail)]

there is a proposal in the CSS WG to make this bug INVALID by changing the
initial value in the spec...

Comment 21

[karnaze (gone)]

Reassigning to pierre.

Comment 22

[Artem Saveliev]

Was just working with border-collapse, noticed that it's fixed.

Comment 23

[Pierre Saslawsky]

Reassigned to Ian to do some follow-up with the WG and maybe fix it when a 
decision is taken.

The places where the property is initialized are:
http://lxr.mozilla.org/seamonkey/source/content/shared/src/nsStyleStruct.cpp#743
http://lxr.mozilla.org/seamonkey/source/layout/mathml/content/src/mathml.css#148
http://lxr.mozilla.org/seamonkey/source/layout/html/document/src/html.css#136

Comment 24

[Phillip Oertel]

don't know if i'm 100% on topic here but i got some comments on border-collapse,
too. i don't get mozilla (2001 101117 win2k) to display the
border-collapse:collapse as expected.
the w3c spec is at:
http://www.w3.org/TR/REC-CSS2/tables.html#propdef-border-collapse
also see the testcase png i submitted, comparing mozilla <--> w3c
border-collapse rendering.
phil.

Comment 25

[karnaze (gone)]

border-collapse has been disabled for a long time and is planned to be back in 
m0.9.7.

Comment 26

[Phillip Oertel]

Attachment: comparing mozilla and w3c table rendering for border-collapse:collapse

Comment 27

[Bernd]

this bug has been become invalid. 

see: 

http://www.w3.org/Style/css2-updates/REC-CSS2-19980512-errata.html

Section 17.6 Borders

[2001-10-03] Several popular browsers assume an initial value for
'border-collapse' of 'separate' rather than 'collapse' or exhibit behavior that
is close to that value, even if they do not actually implement the CSS table
model. Therefore the CSS WG proposes to change the initial value to 'separate',
in the expectation that it better matches what users expect it to be.


http://lxr.mozilla.org/seamonkey/source/layout/html/document/src/html.css#151
shows that we are now compliant.

Comment 28

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

What's in html.css is not the default value.  The default value is here:

http://lxr.mozilla.org/seamonkey/source/content/shared/src/nsStyleStruct.cpp#889