Damian Shaw [Quan] Reporter
	Description • 17 years ago

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9b4pre) Gecko/2008021704 Minefield/3.0b4pre
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9b4pre) Gecko/2008021704 Minefield/3.0b4pre

Though not a prerequisite of passing the Acid2 test it does seem to indicate that zooming doesn't work well enough.

Reproducible: Always

Steps to Reproduce:
1. Take the acid2 test
2. Zoom in or or out
Actual Results:  
Smiley face goes wrong, not too much for zooming in bit by bit, but more when zooming out. Results are amplified when zooming out then zooming in.

Expected Results:  
Should continuously see a nice smiley face.

	Damian Shaw [Quan] Reporter
	Comment 1 • 17 years ago

Here's what the acid2 guys have to say on zooming in Opera 9: http://www.webstandards.org/2006/07/20/acid2-and-opera-9-clarifications/

The basic gist of which, is that while zooming should keep a nice smiley face it's not a prerequisite of the test.

But that's not really the point here, people who 'NEED' zooming to use the browser should have it work well even on incredibly awkward pages.

	Damian Shaw [Quan] Reporter
	Comment 3 • 17 years ago

Glad to see someone else has spotted this, but I'm basically asking Acid2 to work at all zooms to provide accessibility for people who need zooms, not just zooming out then back in.

Not to pass the Acid2 test, but to make sure zoom works properly. I don't know how 'close' a bug needs to be, but I am asking something different... Should I make comments on that bug for its remit to be widened? 

	Robert Longson [:longsonr]
	Comment 4 • 17 years ago

I would think that you really ought to find real world web sites that break with zooming if you want to have a wider remit bug.

	Damian Shaw [Quan] Reporter
	Comment 5 • 17 years ago

Thanks, I've come across a few, I'll leave this bug dead and start making a list of websites and make a new bug :-).

Like I say, it's not really a matter of Acid2 test, just awkward web pages in general.

	Jesse Ruderman
	Comment 6 • 17 years ago

Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.4; en-US; rv:1.9b4pre) Gecko/2008021801 Minefield/3.0b4pre

As long as I shift-reload after zooming (to fix the scroll position), the only problem I see with the acid 2 face is that the background color behind the eyes is a pattern of yellow/orange/red instead of solid yellow.

	Jesse Ruderman
	Comment 7 • 17 years ago

/* the two images are identical: 2-by-2 squares with the top left
      and bottom right pixels set to yellow and the other two set to
      transparent. Since they are offset by one pixel from each other,
      the second one paints exactly over the transparent parts of the
      first one, thus creating a solid yellow block. */

Crazy stuff; I'm not surprised that zooming breaks it.

	Jesse Ruderman
	Comment 8 • 17 years ago

Opera 9.5 does much the same thing.  Hixie claims that this is a bug:

"The eyes go orange if you view the test zoomed in or zoomed out: This is a bug."

http://www.webstandards.org/2006/07/20/acid2-and-opera-9-clarifications/

	Jesse Ruderman
	Comment 9 • 17 years ago

Reopening: let's make this be about viewing the test zoomed, not about zooming an already-loaded test (which is what bug 415446 seems to be about).

	Robert O'Callahan (:roc) (email my personal email if necessary)
	Comment 10 • 17 years ago

That test seems to be claiming that any kind of image scaling that uses interpolation is not allowed.

	Jesse Ruderman
	Comment 12 • 17 years ago

That looks like it's just the scrolling issue (you need to shift+reload), which isn't a bug.

	Damian Shaw [Quan] Reporter
	Comment 13 • 17 years ago

shift+reload doesn't fix it

	Miroslav Mazel
	Comment 14 • 16 years ago

A problem is that when you zoom out too much, and then reset the size (without reloading), the page now does not pass the Acid2 test (you have to reload it to pass)...

	juan becerra [:juanb]
	Comment 15 • 16 years ago

I observed this problem in 3.0.6build1 while going through a BFT. As other people mention, when you shift reload the page, the image is rendered correctly. My expectation, however, is that if you zoom in the reference image, and you do the same in the test image, I expect both to look the same.

	d
	Comment 16 • 15 years ago

To me, it still breaks in 3.5.

	Damian Shaw [Quan] Reporter
	Comment 17 • 15 years ago

Any reason you've decided to change the title? That isn't the only problem with zooming...

	Jesse Ruderman
	Comment 18 • 15 years ago

If there are multiple problems, there need to be multiple bugs...

	shirish
	Comment 19 • 15 years ago

And this still breaks in 3.6b2 nightly build as well.

	Damian Shaw [Quan] Reporter
	Comment 22 • 12 years ago

Invalid case, test was created to trip browsers up in normal zoom who didn't implement CSS spec correctly. Anyone who wrote an actual web page like that and expected zoom to work would be told to write cleaner webpages.