213781 - Floating object rendered vertically incorrectly

Status: VERIFIED FIXED

(Core :: Layout, defect, P2)

Description

[MichaelC]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4b) Gecko/20030516 MSIE/0.6
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4b) Gecko/20030516 /0.6

In the page:
http://customwire.ap.org/dynamic/stories/D/DOWNLOADING_MUSIC?SITE=OHCLE&SECTION=HOME&TEMPLATE=DEFAULT
there is an element rendered incorrectly in terms of vertical positioning.  The
incorrectly rendered object is a grey box, bounded in a thin black line,
containing a couple of advertisements.  The box is rendered as top aligned with
the 5'th paragraph, when it should be aligned with the top of the 6'th
paragraph.  I noticed that this error only because the offending item overlays
some text.

Reproducible: Always

Steps to Reproduce:
Load page, start reading, and the error is almost immediately evident

Comment 1

[MichaelC]

My mistake . . . the useragent should have "Mozilla Firebird" in the name.  I
had to delete that and insert MSIE in order to allow the browser to get to sites
that restrict access to Internet Explorer

Comment 2

[Boris Zbarsky [:bzbarsky]]

Hmm.. That content is not actually positioned. It's not clear to me how it ends
up outside its iframe.

What we need here is something like a smallish testcase....

Comment 3

[Bill Mason]

Attachment: testcase

Comment 4

[Bill Mason]

I don't understand what's happening either.  But hopefully to help someone, here
are several things, any one of which I find fixes the problem if you pare it out
of the testcase:

1. Delete the top of page IFRAME
2. Remove align=left from the overall containing TABLE
3. Remove the invalid SPAN tag that is wrapped around all the page content
4. Remove the align=right from the TABLE holding the IFRAME that is the point of
contention of this bug report.

Comment 5

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Looks like a view positioning problem.  I'm guessing we're not copying
NS_FRAME_HAS_CHILD_WITH_VIEW state bits.

Comment 6

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Attachment: simpler testcase

Comment 7

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

(I also wonder why the IFRAME doesn't occupy its full height on the initial reflow.)

Comment 8

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Attachment: patch

Fix the bug by setting NS_FRAME_HAS_CHILD_WITH_VIEW correctly.

Comment 9

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Never mind the IFRAME stuff -- I think the initial incorrect position is
probably coming from the "unconstrained reflow" or something like that.

Comment 10

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Taking.

Comment 11

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Fix checked in to trunk, 2003-07-25 15:04 -0700.

Comment 12

[Bill Mason]

Verified 2003072604 PC/WinXP