Working together for standards The Web Standards Project


Acid2 and Opera 9 Problems?

By Molly E. Holzschlag | July 13th, 2006 | Filed in Acid2

We’ve received some reports here at WaSP that Opera 9 is not passing Acid2 under certain unique scenarios. We’d like to hear from you in comments if this is the case.

Skip to comment form

If you’re running across any problems with the rendering of Acid2 in Opera 9, please comment below and let us know the following:

  • OS and version
  • Screen resolution settings
  • Does the test resolve on refresh?

If you can do a screen grab of your results as well, and pop ‘em up on a server somewhere and provide the URL, that’d be great too.

Your Replies

#1 On July 13th, 2006 1:12 pm Bryan Rieger replied:

Seems to work fine for me – the test resolved on the initial load, although there was a small glitch for a second while it loaded.

Opera 9.00 Build 8501
Windows XP SP2
1280 x 768

Screenshot:
http://bryanrieger.com/resources/images/opera9-acid2-test.png

#2 On July 13th, 2006 1:14 pm Bryan Rieger replied:

Sorry, didn’t realise the ‘yellow + black bars/hat’ aren’t supposed to be there – oops! Refreshing the page does indeed make them go away.

#3 On July 13th, 2006 1:48 pm Jules replied:

I found a difference in Opera 9 (Windows 2000) depending on whether Fit to Width is ON or OFF (it seems to me that by default, this setting is on in Opera 9).

http://flickr.com/photos/[email protected]/188908390/

#4 On July 13th, 2006 2:34 pm unwiredbrain replied:

Opera 9.00 Build 8501
Windows 2000 (SP4, ITA)
1024 x 768

Looks exactly as the reference, even on refresh. Good.

#5 On July 13th, 2006 2:40 pm Emil Stenström replied:

9.00 Beta (build 8414)
Windows XP (all updates applied)
1152 x 864

It seems to have something to do with the screen width. The page it looks fine at the first load. If I then resize the page I get the “hat” like Bryan Rieger showed above. A reload fixes it.

#6 On July 13th, 2006 3:02 pm Jason Brunette replied:

Resizing the browser window (or adjusting the zoom level in the page) causes “corruption” because the “Standards compliant?” text at the top of the page wraps (or unwraps) depending on the width of the browser window, causing part of the text object to move accordingly, but some of the objects do not (“the hat”).

#7 On July 13th, 2006 3:05 pm Jason Brunette replied:

Er…

“causing part of the TEST object to move accordingly”.

#8 On July 13th, 2006 3:18 pm Jason Brunette replied:

Thinking about it more, the “hat problem” is by design. Part of the test includes absolutely positioned elements that exist in the same place on the browser viewport no matter how far the page has scrolled. Those absolutely positioned elements are hidden under the “Standards compliant?” box at the top of the page. The “Take the Acid2 test” link simply scrolls the page down so that the absolutely positioned elements are perfectly aligned with the rest of the test elements. Changing the “height” of the document content (by adjusting the page width as described above) causes the absolute positioning to no longer work.

Perhaps a simple “fix” (if even deemed important) is to prevent the text in the “Standard Compliant?” box at the top from wrapping, or make it an image, or whatever.

Of course, this is all assuming that this is the problem Molly is asking about. :-)

#9 On July 13th, 2006 3:22 pm Jon replied:

Opera Version 9.00 Build 344
Ubuntu Linux 6.06 i686, 2.6.15-26-386
Dual monitors at 1280×1024

Renders correctly initially.
Renders correctly on refresh.
Renders correctly on minimize then maximize.

If window is resized to smaller than the smiley face, an artifact appears directly over the top.
http://www.mondaybynoon.com/images/opera9/resize.png

If the scroll wheel is used, double artifact appears. See screenshot:
http://www.mondaybynoon.com/images/opera9/scrollwheel.png

Spanning dual monitors is interesting:
http://www.mondaybynoon.com/images/opera9/dualspan.png

#10 On July 13th, 2006 3:40 pm Justin Thorp replied:

It doesn’t look like it is passing for me.

Opera 9.0 Build 3447
Mac OS 10.4.7
1024 x 768

Screenshot:
http://www.mycapitalweb.com/justin/acidtest.jpg

It doesn’t fix on refresh.

#11 On July 13th, 2006 3:48 pm Jason Brunette replied:

Justin Thorp:

That’s a zoom bug in Opera 9. It appears you have the zoom setting at 90%, which causes Opera to improperly render the Acid2 test. Set the zoom setting to 100% (View menu : Zoom : 100%) then refresh the page and it should look normal.

#12 On July 13th, 2006 5:02 pm Nathaniel replied:

I experience two odd bars, however this only occurs after I scrolll…..should I even be able to scroll? The little smiley guy moves down, and the bars stay in place. This didn’t happen in Konquerer. After a refresh, the page moves back to the normal location (but I can still scroll and reproduce it again).
http://aeonofdarkness.com/~pics/opera-acid2.png

#13 On July 13th, 2006 5:03 pm Nathaniel replied:

I forgot to include platform info:
WinXP SP2
1280×1024
Opera 9 build 8501

#14 On July 13th, 2006 5:18 pm Chriztian Steinmeier replied:

Opera 9.00 (Build 3447)
Mac OS X 10.4.7

When resizing to 862px width or less: Something drops…

When resizing to 862px width or less: Even more goes…

#15 On July 13th, 2006 8:29 pm Ben Boyle replied:

Everything is exactly as described in previous comments (for XP, Opera 9 build 8501 @ 1280×1024).

Additionally, the nose goes blue when hovering over the middle section (I don’t know if that’s part of the test).

#16 On July 13th, 2006 9:36 pm nullpage replied:

Win2000
Opera9 build 8518
1024×768

http://www.geocities.jp/nullpage_vwxyz/opera9_acid2.jpg

#17 On July 13th, 2006 9:41 pm Sharif Sayed replied:

Ubuntu 6.06 LTS & Windows XP Home SP2
Opera 9.00 Build 344
1680×1050

for some reason two bars seem to be moving whenever i scroll up and down.

#18 On July 14th, 2006 1:08 am karaj replied:

WinXP
Opera9.01 b8518
1280×1024
Refresh does not help

http://w3.enternet.hu/puki33/opera_acid2_bug.png

#19 On July 14th, 2006 1:22 am Alexander replied:

Perfectly (on refresh too), but the two horizontal bars on the top of the head moving when scrolling page top-down too…

WinXP
1152×864
Opera 9 8501

#20 On July 14th, 2006 2:12 am Rolf replied:

Version: 9.00
Build: 3216
Platform: MacOS X
System: 10.4.6

Some problems there. Refreshing didnt solve them.

http://rolf-thomas.de/test/acid2.jpg

#21 On July 14th, 2006 2:19 am chaals replied:

Nope, you cannot scroll, cannot resize your window or text (except perhaps in certain particular magic jumps), cannot set a minimum font-size at anything like someone who needs it would use, or the test will not render “correctly”.

That’s the way it was written. (After all, there aren’t that many people who are going to try and read the content of the test right? It’s just to see if it works or not…)

cheers

Chaals

#22 On July 14th, 2006 2:47 am edwin replied:

i experience the same problem with nathaniel

“” I experience two odd bars, however this only occurs after I scrolll…..should I even be able to scroll? The little smiley guy moves down, and the bars stay in place. This didn’t happen in Konquerer. After a refresh, the page moves back to the normal location (but I can still scroll and reproduce it again).
http://aeonofdarkness.com/~pics/opera-acid2.png “”

Version 9.00 Build 8505
System Windows XP
Screen Resolution 800 x 600

doesn’t resolve after refresh.

#23 On July 14th, 2006 3:22 am Andrew Gregory replied:

I think folks should have a read of:

Acid 2 in major browsers

It’s very informative!

#24 On July 14th, 2006 3:59 am Frederik Vanderstraeten replied:

Looks fine, except if I resize or change text size. It works again however when clicking the url bar and pressing enter (ask Opera to scroll to top again).
However after doing this a few times strange red lines appear.

Sorry guys, Opera doesn’t pass. You have to do more then just rendering fine just after loading!
The css of web pages shouldn’t break after a resize!
And don’t say it’s just the acid test. My web site also breaks in Opera after a resize. (In Opera 8, this issue is fixed in Opera 9.)

You should probably do like Firefox. Show the starting screen with the “Take The Acid2 Test” box again after a resize.

Almost there! Good luck!

#25 On July 14th, 2006 4:28 am Robert Græsdal replied:

@Fredering Vanderstraeten:
What you are seeing is a zoom bug and has more to do with the displaying of images than anything CSS related.

The point is that Opera pass all the conditions set forth by the test. Zooming of images, and the entire page for that matter, is definately not part of the test therefore Opera pass.

#26 On July 14th, 2006 4:28 am Frederik Vanderstraeten replied:

I just noticed this, maybe you already knew it, but Opera’s View Source syntax highlighting doesn’t pass the test!

It doesn’t render the

near the end right.

#27 On July 14th, 2006 4:30 am Frederik Vanderstraeten replied:

Oops I forgot this blog supports html.

It doesn’t render the
<!– ->Error<- –>
near the end right.

#28 On July 14th, 2006 4:34 am TarquinWJ replied:

Several of you mention problems when you zoom, resize, scroll, etc. These are all expected, and are NOT part of the test. The test itself is not designed to pass except under default setups. I will repeat the points from my article:, since some of you (such as Frederik) need to read this before posting screenshots or commenting.

* The forehead is a fixed position element. It will stay still when you scroll. That is correct behaviour. The test only works when it is at the default position after clicking the link. That is the correct response.

* The test may not pass if you make your minimum font size too big, or if you zoom, etc. The test itself is intentionally designed to only work assuming a ‘normal’ setup, so this is expected behaviour. Complain to the authors if you do not like it :)

* Resizing the window may cause text at the top to wrap, and scroll the page a little, so you will need to click the link again – see note 2.

* Making the window very narrow will make parts of the test become too narrow (such as the chin) – this is because they use the table layout algorithm, and again, this is expected behaviour.

* The test will not work if you enable small screen rendering or Fit To Width (with a narrow window). That is intentional behaviour.

If you need more information about this, please see the whole article:
http://www.howtocreate.co.uk/acid/

Note that it will also not work if you disable images, CSS, run user JavaScripts that alter it, or any number of other things (zoom, in particlular, will show odd responses as it has to compute scaled sizes that often end up being less than a pixel – so the hash images may not scale perfectly, and odd patches of colour may appear as a result).

If you have changed any of these things, then your screenshot is useless. Undo your changes, and try again. If it still goes wrong, then perhaps you have found a real problem.

#29 On July 14th, 2006 4:36 am Rijk replied:

@Fredrik, what does View Source have to do with Acid 2? Please stop slandering Opera, if you don’t understand the way the test works. It does *not* say the display should be unaffected by scrolling, zooming, resizing, etc. Read the link Andrew gave, and the explanation given by Chaals!

“Do like Firefox” currently means not passing the test, doesn’t sound like a swell idea…

@Rolf: please use Opera 9 ginal to do the test, not a preview version…

Molly, what was the purpose of this thread? It is no secret that the test ‘fails’ (in all browsers) after zooming (if thre browser supports full pagezoom), resizing, scrolling etc. Please consult with Hixie if you think there are problems.

#30 On July 14th, 2006 4:42 am Joel replied:

*DO* read the notes above before whinging. If you looked at the site posted by Andrew, you’d see that the Acid2 should be expected to do exactly what you suggested happens.

#31 On July 14th, 2006 9:06 am Nathaniel replied:

On the explanation of the Acid 2 test, this is said, “If the Acid2 page is scrolled, the scalp will stay fixed in place, becoming unstuck from the rest of the face, which will scroll.” With a picture of a black rectangle. There is no other mention of scrolling of the page. Unless I’m missing something then, the distortion shown here: http://natmaster.com/~assets/images/acid2/opera-acid2.png is still wrong.

#32 On July 14th, 2006 9:15 am TarquinWJ replied:

@Nathaniel

You are missing something. Please read this article:
http://www.howtocreate.co.uk/acid/

If you scroll, you invalidate the test. What you are seeing is what is meant to happen. If you still need clarification, please check the source code of the test yourself and check out the CSS fixed positioning that is used.

#33 On July 14th, 2006 12:10 pm Reynar replied:

We’ve received some reports here at WaSP that Opera 9 is not passing Acid2 under certain unique scenarios.

“Certain unique scenarios”?. Acid2 should not work under “unique scenarios” or exotic configurations; you are not supposed to mess around with zoom, fit to width, stylesheets, … Aren’t you aware of this?, shouldn’t you be aware of how Acid2 works?.

This entry cannot be more vague and misleading. Sorry but this is FUD material at best.

#34 On July 14th, 2006 12:18 pm Robert replied:

The instanity is spreading to digg.com too:

http://digg.com/software/Opera_s_Passing_of_Acid2_Test_Called_Into_Question

This makes me embarrassed to be a Firefox user.

Opera passed before Firefox. Big deal. Move on.

Don’t make a fool of the entire Firefox community by spreading FUD about Opera.

#35 On July 14th, 2006 12:21 pm Some questioning whether Opera 9 passes the Acid2 Test » Opera Watch replied:

[...] Some questioning whether Opera 9 passes the Acid2 Test By Daniel Goldman July 14th, 2006 1:20 PM EDT A post on the Web Standards Project (WaSP) blog seems to question whether Opera 9 passes the Acid2 test. [...]

#36 On July 15th, 2006 3:18 am Frederik Vanderstraeten replied:

@Rijk: The View Source has nothing to do with the acid2 test. I just thought I’d say that so someone could correct it.

You’re right, Opera 9 passes the acid2 test. I would say congratulations, but I already did when it became known that it passed the test.

But you should also look beyond the test. Just passing the test doesn’t mean you’re standards-compliant.

The issue of the test breaking when you resize the window, is because of what I experience as a usability problem (probably has nothing to do with standards, but it is annoying). When you resize a window, you scroll to somewhere else. It happens on pages where the width of the text is relative (80%, but also the standard 100%). If you’re in the middle of reading a long document, and after resizing, you have to find out where you were, that’s quite annoying. Of course this happens in all browsers up to today. (I know why this happens, still it can be fixed.)

I guess I just wanted to say, that altough you did a great job at passing the test, your browser still isn’t perfect, but keep going like this, and it soon will be!

#37 On July 15th, 2006 4:19 am Robert replied:

“Just passing the test doesn’t mean you’re standards-compliant.”

Straw men will get you nowhere.

“I guess I just wanted to say, that altough you did a great job at passing the test, your browser still isn’t perfect”

No one claimed that it was. This is a discussion for another place and another time. In this blog post, we are discussing the false claims that Opera was lying about passing the Acid2 test.

#38 On July 15th, 2006 4:38 am Nikko replied:

Windows XP (no service packs)
Opera 9.01 build 8518
800×600 monitor

Zooming in shows red grains in the eyes. Zooming out misplaces the lines in the head and shows a line in the face between eyes and nose.

I just want to say that Opera did a great job for making the face appear correctly but still it’s not perfect, nobody’s perfect and Opera is the best browser around even though it has only a niche of users.

I don’t think that Opera is lying and I think that they should just tell that they “almost” passed the test so this kind of issues would come out.

#39 On July 15th, 2006 5:11 am TarquinWJ replied:

@Nikko

Allow me to repeat myself :)

“zoom, in particlular, will show odd responses as it has to compute scaled sizes that often end up being less than a pixel – so the hash images may not scale perfectly, and odd patches of colour may appear as a result”

That explains the red grain around the eyes. Zooming to 50% usually (but not always) works.

Zooming also changes the size of the text at the top, and therefore changes how it wraps. In doing so, it changes how the page is scrolled, and makes those two black+red and black+yellow fixed position elements to become offset from the rest of the face.

The test is not designed to be zoomed, and if you do zoom, you invalidate the test. The test must only be used at 100% scale – that is the way it was designed to be used.

Please see this article for more information:
http://www.howtocreate.co.uk/acid/

#40 On July 15th, 2006 5:13 am Reynar replied:

@Nikko
“Zooming in shows red grains in the eyes. Zooming out misplaces the lines in the head and shows a line in the face between eyes and nose.”

Way to go!! You almost got it. Now, go read the source code and the comments in this page and come back when you understand what it is suppossed to happen to Acid2 when you Zoom the page.

“I just want to say that Opera did a great job for making the face appear correctly but still it’s not perfect, nobody’s perfect and Opera is the best browser around even though it has only a niche of users.”

Patronizing people is just a bad idea, even more when you clearly show that you don’t understand what is Acid2.

“I don’t think that Opera is lying and I think that they should just tell that they “almost” passed the test so this kind of issues would come out.”

They should tell nothing, Opera passes Acid2 just fine, there are no “issues” other than you don’t knowing how Acid2 works. If some people don’t understand the test or want to spread FUD about Opera with the help of WaSP it’s not Opera’s fault.

BTW, the author of this post should apologize to Opera for posting such an unprofessional and misleading entry.

#41 On July 15th, 2006 6:07 pm Christian Kirbach replied:

test fails on Linux, screen resolution 1280×1024
Ubuntu Linux 6.06

The top of the head is misplaced a bit.Reload does not have any effect.

Opera details:
Version 9.00
Build 344
Platform Linux
System i686, 2.6.15-23-686
Qt library 3.3.6

#42 On July 16th, 2006 11:08 am Robert replied:

That’s strange considering that it renders correctly for me. (And also did on the 2.6.15-23 kernel build).

Version 9.00
Build 344
Platform Linux
System i686, 2.6.15-26-386
Qt library 3.3.6

Could you please post a screenshot and check your zoomlevel, minimum font-size setting and if you scrolled the page or not (as mentioned earlier this does affect the result).

#43 On July 16th, 2006 3:09 pm Rory replied:

Works fine for me !!

Opera 9
Win XP Pro
1024×768
16bit color depth

#44 On July 17th, 2006 12:35 am Dennis Dolan replied:

http://www.dennisdolan.com/screencapture001.jpg The screen capture shows the effect of scrolling down a couple of lines.

http://www.dennisdolan.com/screencapture002.jpg shows the results of scrolling the face completely off the screen (same result whether scrolling up or down).

Reloading the page restores the face, but the error recurs with a vertical scroll.

OS: Windows XP Media Center Edition, version 2002, Service Pack 2

Opera: Version 9.00, Build 8502

Display: 1440 x 900, 32-bit, Opera is in a window about 1024 pixels wide

#45 On July 17th, 2006 5:08 am Ivan replied:

Opera 9.00 Build 8501
Java Sun Java Runtime Environment version 1.5
Win XP Pro SP2
1152×864 x32bit

#46 On July 17th, 2006 8:01 am TarquinWJ replied:

@Christian Kirbach

Sounds like your minimum font size it set too high – probably set to 13px. That is another thing that invalidates the test. If you set it to its default value of 9px, it should work.

Tools – Preferences – Advanced – Fonts – Minimum font size (pixels) = 9

If not, I will need to see a screenshot to work out what you have changed that causes it to fail.

#47 On July 17th, 2006 1:13 pm Bob replied:

It renders correctly when minimum font size is set to 12 and below and fails when it is 13 and above.

This is actually illegal under disability discrimination laws in the UK, and quite probably in other countries as well.

The Acid2 test should be thought out again.

#48 On July 17th, 2006 1:34 pm Nathaniel replied:

“The Acid2 test should be thought out again.”
Or maybe just make an Acid3?

#49 On July 17th, 2006 5:20 pm Kelson replied:

This is actually illegal under disability discrimination laws in the UK, and quite probably in other countries as well.

How? Acid2 isn’t meant for human consumption, it’s meant to test a browser’s capabilities. Acid2 no more needs to be accessible than an XML fragment intended for data exchange.

I think what’s getting lost in this discussion is that displaying the pixel-perfect happy face isn’t the point of the test. It’s an indicator, which should appear pixel-perfect under default conditions (mostly listed in the guide, except that it doesn’t describe “common values” for the “default stylesheet”) that the browser handles a set of features and errors in the correct fashion. Change those conditions, and the rendering will change as well.

A minimum font size is a user setting that overrides the author’s design. It’s the equivalent of using a user stylesheet. I can put all kinds of things in a user stylesheet that prevent the test page from showing the happy face, but that doesn’t change whether Opera handles data: URLs correctly, or fixed positioning, or transparent borders, or invalid CSS properties, or any of the other items being tested.

#50 On July 17th, 2006 7:04 pm Ben Buchanan replied:

Renders fine for me – tried Opera 9.0 (build 8501) on both WinXP and Win2k.

Yes, it does “break” if I scroll or zoom but I was never under the impression that the Acid2 test was supposed to scroll or zoom.

As far as I can tell from the Acid2 guide, the successful rendering means it passes the test. Zoom is supposed to be 100% and the scalp is *meant* to stay where it is. The red artefacts thing is irrelevant as you have to go outside the test to produce it. Just my take on it, some of the wording is ambiguous under this level of scrutiny.

…and we should all take a deep breath! :)

#51 On July 18th, 2006 7:07 pm Bob Harrison replied:

Did not work for me – a bar across where the eyes should be (I don’t have a website but could email the screen grab to some-one)
System
XP2 fully up-to-date
1024 x 768 (also does not work on 1280 x 1024)
32 bit colour
Min font size 9 pt
Also does not resolve on refresh

#52 On July 19th, 2006 12:46 am Bob Harrison replied:

Whoops! forgot to mention Opera version
9.00 build 8502

#53 On July 19th, 2006 3:09 am Bob Harrison replied:

Well, well – on another machine (my laptop rather than the work desk top) Opera 9 is fine! – passes Acid2
XP – SP2
1024 x 768 min, 32 bit colour, font 9 pts
9.00 build 8502

Any explanation?

#54 On July 19th, 2006 4:26 am TarquinWJ replied:

@Bob

Sounds like you have disabled images or possibly changed the zoom. (And again, doing either of these invalidates the test).

#55 On July 19th, 2006 5:35 am Bob Harrison replied:

Tarquin
Nope – I did exactly the same on both systems. Can a firewall have any affect? I have a hard-wired one on my ADSL modem at home – work has restrictions on things like flash and maybe other things. Perhaps that is limiting some of the commands – I am not familiar with the coding

#56 On July 19th, 2006 11:22 am Event Horizon of D-Dz.Net » Blog Archive » Opera9 and Acid2 Problems replied:

[...] WaSP posted an entry asking readers if they encounter any problems while rendering the Acid2 test page with Opera 9. Yes, I do encounter problems while resizing the browser and more problems when zooming in and out. The smiley face rendition became disjointed and sometimes, a refresh would fix it. [...]

#57 On July 19th, 2006 1:59 pm Bradley Momberger replied:

Opera 9.00 release, build 8501
WinXP SP2
1024x768x32b

On first access, the line on which the “eyes” were supposed to show up had a red-tinged bar going from the beginning of where the left eye should be to the right edge of the face. Refresh fixed it.

I’ve noticed similar issues on a lot of pages with Opera 9. Some pages render completely blank when refreshing my session on program startup, others have minor rendering problems. Both can be remedied through refreshing the respective pages.

#58 On July 19th, 2006 11:23 pm Bob Harrison replied:

One final test was to use the lap top at work – exactly the same setup as at home except using a proxy server connection to allow external web access at work (this is not necessary on the ADSL at home). The test failed – giving the same errors as the work desk top. Conclusion – the work network is affecting the results in some way. Does this happen for others? Some clever people out there must have and idea of why this is – it does not appear to be a fault of Opera.

#59 On July 20th, 2006 8:00 am chaals replied:

The Acid2 test should probably have been designed so that people with visual disabilities could still run the test effectively themselves. Given that it is about rendering, it is very hard to ensure that this will always be the case. And it is always possible to do an automatic comparison based on checking actual images without seeing the thing at all.

But that’s part of history and learning. Maybe Acid3 will be different, but who knows.

#60 On November 14th, 2006 9:21 am KJ replied:

I’m having an issue which might be related to this and i’m hoping someone can advise…

I have a really strange rendering issue in Opera 9. My site renders fine in every other browser but Opera 9. For some reason when I view a page that requires a vertical scroll bar the page shifts its position down by a 100 pixels or so but when I hit refresh the page renders in the correct position.

You can see the page here. If you resize the browser window the page will jump down, then refresh and it moves back up again.

Is this related to the acid2 test?

#61 On January 6th, 2007 1:31 am Jon Allen replied:

Works fine for me

Opera 9.10 Build 8679
Windows Vista (RC1 Beta build 5600)
1280X1024

Return to top

Post a Reply

Comments are closed.


All of the entries posted in WaSP Buzz express the opinions of their individual authors. They do not necessarily reflect the plans or positions of the Web Standards Project as a group.

This site is valid XHTML 1.0 Strict, CSS | Get Buzz via RSS or Atom | Colophon | Legal