282558 - Scaled image dissappears on browser resize

Status: NEW

(Core :: Layout, defect)

Description

[Tony Jacobson]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0

In this html page: http://www.tonyjacobson.com/caseintel/index.html I have a
scaled image (dark blue gradient strip) that scales to fill the width and height
of it's table cell. (image width="100%" height="100%") If i put the URL in when
I first open up a browser window it will display the image. As soon as I resize
the browser the scaled image dissappears, or becomes not visible anymore. As
soon as it dissappears not even page refreshing can cause it to reappear. The
only way I can have it appear again is to close and reopen the browser.

Reproducible: Always

Steps to Reproduce:
1. Open a new firefox browser window.
2. past in this url: http://www.tonyjacobson.com/caseintel/index.html
3. The scaled image (dark blue gradient strip on the far right) scales to fill
the space.
4. Resize the browser window and the scaled image dissappears.

Actual Results:  
As soon as I resize the browser the scaled image dissappears, or becomes not
visible anymore. As soon as it dissappears not even page refreshing can cause it
to reappear. The only way I can have it appear again is to close and reopen the
browser.

Expected Results:  
The image should scale to fill the space in it's contained in.

Comment 1

[alanjstr]

WFM Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041109
Firefox/1.0 (MOOX M3)

Comment 2

[David Tenser [:djst]]

Could you include a screenshot of how it's supposed to look like?

Comment 3

[Steve England [:stevee]]

Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2) Gecko/20050330
Firefox/1.0+
I also see incorrect behaviour for the URL. (Resizing the browser window causes
elements of the webpage to disappear and reappear and not get drawn properly)

Comment 4

[Daniel Ross]

http://members.cox.net/oceantides/ also suffers this issue; refresh the page to
make the left frame disappear. The page does not validate, but I have in the
past created a cleaned up version that did, and it continued to suffer the same
issue.

I have tested this issue and found it on Linux as well as XP. Somebody please
change the OS to All.

Comment 5

[Scott MacGregor]

no patch, not a regression over 1.0 and not serious enough for us to stop the
release. Hopefully someone will work on this for the next release. 

Comment 6

[Daniel Wang]

reporter (Tony), has the page changed? Can you attach a reduced testcase here?

Comment 7

[Samuel Sidler (old account; do not CC)]

I still see this as well using Firefox 2.0.0.3 on Mac OS X.

Comment 8

[Daniel.S]

Attachment: What is should (?) look like

Comment 9

[Daniel.S]

Attachment: testcase

Basically the 1x508 Pixel image (1x100 Pixel in the testcase) is streched equally vertically and horizontally.
If the browser window gets resized vertically, the image height shrinks.

Opera, Safari and IE7 keep the intrinsic height of the image while we and IE8 expand the image 1:1. That's true for Quirks and Standards Mode.

May be dupe, not a bug at all, but I'm unsure.

Comment 10

[dolphinling]

I think there's another part to this bug that the testcase doesn't show.

On Minefield (20081224 linux), with the page in the URL field, when my browser is narrower than the 911px that the left table cell is specified as, I see the image that's in that cell. As I resize the browser so the right cell can show, the image in the right cell expands horizontally to fit its container and vertically to keep its aspect ratio, thus stretching the page vertically to absurd heights. The content of the left cell is still on the page, just in the middle, and you need to scroll way down to see it. This is what comment 9 is about.

However, there is also an issue with the image in the right cell. When the height of the right cell is over 65536 (or somewhere around there because the image isn't the only thing in the right cell), the image disappears, and even before that, when the height of the right cell is between 32768 and 65535px (again, or somewhere near there) the image stops displaying at around 32kpx.

I believe that in the original report, a reflow bug was causing the table cell to not expand its height until the browser was resized.

Comment 4, amazingly enough, suffers exactly the same issues--images that scale horizontally and thus become extremely tall, and also get cut off above 32kpx and disappear above 64kpx. It has a slightly different page layout, and also the actual content has to be removed to even get the images under 64kpx height.


The first issue here is as comment 9 says maybe not a bug at all. The second issue is definitely a bug, but likely a dupe.

So, analysis complete, someone needs to decide what to do with this.