65566 - Border is redrawn improperly when fieldset size is increased

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[Antti Näyhä]

Overview description:
  Fieldset border is improperly redrawn when the fieldset's size is increased
  dynamically.

Steps to reproduce:
  1) View the attachment.
  2) Push the button to reveal the hidden paragraph.

Actual results:
  The new, expanded border is drawn, but the previous border is not completely
  cleared.  Result: an ugly extra horizontal line in the right border.

Expected results:
  Nice, clean redraw.  See the attachment on eg. IE 5.5.

Build date & platform:
  2001-01-15-08 (Linux i686)

Additional builds and platforms:
  Mozilla 0.7 (Win32)

Comment 1

[Antti Näyhä]

Attachment: A resizing fieldset test

Comment 2

[Antti Näyhä]

Adding 'html4' keyword.

Comment 3

[HJ]

This is an calculation error caused by the border/padding value(s) in the CSS.
If you use something like this:

  fieldset {
  	display : block;
  	border : 1px solid black;
  	padding : 0px;
	padding-top : 4px;
	padding-bottom : 4px;
  	margin-left : 2px;
  	margin-right : 2px;
  }

then you don't see that problem! If you set border to 2px then you will see 1px
at the right! By increasing the border/padding value(s) the problem get worse,
by that amount of pixels. If you resize the window, the problem is solved. So
this is only a problem after the first rendering. It's a bug, that's for sure!

Comment 4

[HJ]

Attachment: Example of the border/padding issue

Comment 5

[HJ]

b.t.w. that line around fieldset in the default HTML.CSS should be 1px thick and
not 2px because that's ugly!

Comment 6

[[e-mail]]

H-J wrote:
> b.t.w. that line around fieldset in the default HTML.CSS should be 1px thick and
> not 2px because that's ugly!
Actually I would have thought 2px groove threedface would be nicer (c.f.
http://lxr.mozilla.org/seamonkey/source/themes/classic/global/win/box.css)

Comment 7

[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 8

[Pierre Saslawsky]

Reassigned to Layout/attinasi.  I guess there is problem with the reflow of the 
nsFieldSetFrame. It doesn't happen with DIVs for instance (see my new testcase).

Comment 9

[Pierre Saslawsky]

Attachment: testcase fieldset/div

Comment 10

[Marc Attinasi]

Looks more like an invalidation problem, because hiding and re-exposing the
fieldset frame makes it paint OK. CC'ing rods.

Comment 11

[John Keiser (jkeiser)]

The fix to bug 124507 can be applied to this as well.  Taking.

Comment 12

[Adam Hauner]

WFM 2002050806/branch/Win2K

Comment 13

[John Keiser (jkeiser)]

It is passing strange, but this WFM Linux 2002050910 too.  I don't see any fixes
to nsFieldsetFrame, but clear as day the problem is fixed.