Project: chromium Issues People Development process History Sign in
New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.
Starred by 111 users

Comments by non-members will not trigger notification emails to users who starred this issue.
Status: Fixed
Owner:
Last visit 16 days ago
Closed: Jul 2009
Components:
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 1
Type: Bug
M-3

Restricted
  • Only users with EditIssue permission may comment.



Sign in to add a comment
acid3 - Acid test 3 fails on chrome
Reported by duncan.l...@gmail.com, Sep 2 2008 Back to list
Product Version      : 0.2.149.27
URLs (if applicable) : http://acid3.acidtests.org/
Other browsers tested:
Add OK or FAIL after other browsers where you have tested this issue:
     Safari 3: fail
    Firefox 3: fail
         IE 7: epic fail

What steps will reproduce the problem?
1. Visit: http://acid3.acidtests.org/
2. Wait for the test to run
3. Note that the result doesn't match the reference rendering: http://acid3.acidtests.org/reference.html

What is the expected result?
The test result should match the reference rendering here: http://acid3.acidtests.org/reference.html

What happens instead?
It fails with 63/100 test passed, when I tried it.

Acidtests.org appears to be very bust at the moment (odd that :-) so this might 
be affecting things a bit. I though that the webkit dev version passed acid3?

Yeah - here we are: http://webkit.org/blog/173/webkit-achieves-acid3-100100-in-
public-build/ - webkit blog entry from March 08 about how the webkit nightlies 
now pass acid3. What's happened since then?
 
 
The latest Chrome 0.2.149.27 Scores a 79/100 in ACID 3. I have attached a detailed 
report.
Chrome-acid3.PNG
35.4 KB View Download
Status: Untriaged
What's happened is that we aren't on the WebKit trunk, we're on a version of WebKit 
that's roughly equivalent to Safari 3.1.

Confirmed, but this is basically Upstream.  We'll get nearly all of this for free 
once we sync to WebKit tip of tree.
When and how often are you planning to synch with webkit tip to tree?
[NOT A BUG]
Failed on 79,78,77%
5 measures, not enough for statistics

Comment 6 by Deleted ...@, Sep 4 2008
i too would like to know when you plan on syncing to "tip." though I can understand
you wanting to use the most stable 3rd-party software as possible, so you can focus
on bugs in Chrome itself as opposed to its HTML renderer...
The sync might not produce 100/100 on acid3 test due to differet JavaScript 
(EcmaScript) engine. Webkit uses JavaScriptCore, but Chrome uses V8 therefore there 
still might be bugs in V8.

Not switching to chrome until it passes acid3 test.
If this browser is suppose to embrace new technology, it should be compliant with W3 
standards.  It needs to be able to pass the acid test.  Failing acid3 means it's not 
compliant with JavaScript W3 standards.  I am not sure if it's an issue with using an 
older version of WebKit, or the new V8 JavaScript engine you all coded into it.
Comment 9 Deleted
They are most likely not going to sync with webkit trunk before it's marked /
branched of as stable.

If you want ACID 3 100% score so bad, use webkit nightly build or Opera wingogi
preview. Both trunk / in development / contains a lot of bugs..

But I do question why they aren't on pair with Safari 3.1.2(525.19), I would guess
that would be the 'Safari-3-1-branch' ?

So stop whining! No other browser support this in a stable browser yet..
Comment 11 by igi...@gmail.com, Sep 7 2008
@Cody.Petry: Correct me if I'm wrong, but AFAIK acid3 tests a lot of *proposed* new 
web standards.  Not all of the tests have been formally adopted as standards.  But I 
might be wrong.
If it gets the frowning kitten on ACID3, it's falling behind other browsers already
out there.
chrome.png
53.7 KB View Download
Comment 13 Deleted
Opera 9.5                 : 84
Chrome 0.2.149.29         : 79
Safari  3.1 (windows+mac) : 75
Firefox 3                 : 71
IE 7                      : 5
The above comparison is not fair, you need to compare the beta version of Chrome with
the beta version of the other browsers:
Safari 4          : 100
Opera             :  99
Firefox 3.1       :  85
Chrome 0.2.149.29 :  79
IE 8 beta 2       :  21

Having said that, compared to the browsers available, Chrome is hardly a disaster and
the upgrade to the latest version of Webkit need not necessarily be urgent because of
this bug.
Comment 16 by Deleted ...@, Sep 16 2008
Google Chrome Beta: 78
Internet Explorer 7: FAIL 12
Netscape Navigator 9: FAIL 52


since Acid3 mainly tests ECMAScript, and Chrome uses a different ECMAScript engine 
than WebKit, so even syncing to the tip of WebKit tree doesn't include the 
JavaScriptCore right?
@igitur: You are incorrect. Acid3 tests standards that are at recommendation level 
are beyond. It is not testing proposed new standards as you suggest.
Labels: -area-unknown Area-Misc
The latest chromium dev build ( 2786 ) Scored a 98/100 for me, web-kit has been 
updated to version 528.4 now according to about:version.
Comment 21 Deleted
Comment 22 Deleted
The latest chromium buildbot (3625) passes the Acid3 test with a score of 100/100 but  
it still says "linktest failed".
Comment 24 by jon@chromium.org, Oct 21 2008
Labels: -Area-Misc Area-WebKit Mstone-1.1
Status: Upstream
The relevant test is:

    function () {
      // test 48: :link and :visited
      var iframe = document.getElementById("selectors");
      var number = (new Date()).valueOf();
      var a = document.createElement('a');
      a.appendChild(document.createTextNode('LINKTEST FAILED'));
      a.setAttribute('id', 'linktest');
      a.setAttribute('class', 'pending');
      a.setAttribute('href', iframe.getAttribute('src') + "?" + number);
      document.getElementsByTagName('map')[0].appendChild(a);
      iframe.setAttribute("onload", 
"document.getElementById('linktest').removeAttribute('class')");
      iframe.src = a.getAttribute("href");
      return 3;
    },

As you can see it lacks assertions of expects like the other tests so I don't think 
it is a valid test.  Either way, we should pick up the fixed from WebKit and Acid3 
test failures should be entered there.  See https://bugs.webkit.org/show_bug.cgi?
id=17064
Comment 25 Deleted
Comment 26 by phistuck@gmail.com, Oct 30 2008
Now, it scored 77-79/100. What the hell happened? :(
Version 0.3.154.6-9 (same results for both).
Comment 27 by phistuck@gmail.com, Nov 25 2008
0.4.154.25 - still 79%. Bad, bad, bad. Regressed!


PhistucK
Comment 28 by sutea...@gmail.com, Nov 25 2008
windows XP
Thailand
Opera 9.62 build 10467 :: 84-85 
Chrome 0.4.154.25 :: 77 - 79
IE 8 beta 2 :: 12-21 
Firefox 3.0.4 :: 70-71

about 10 times per browser.
i dont want to install safari on windows. :P
why ppl still use ff and IE
 Issue 6208  has been merged into this issue.
Comment 30 Deleted
100/100 with actual dev-chrome but "Linktest failed"
Status: Started
Let me see what I can do.
I've just noticed that the new version (2.0.169.1) always passes test 69. Previous 
versions did pass it, but by clicking on the A it said that it took too many 
iterations; now it's working perfectly. Great work!
Comment 34 by jon@chromium.org, Mar 19 2009
Labels: -Pri-2 -Mstone-1.1 Pri-1 Mstone-2.0
The new BETA version of chrome passes with 100/100 however it says Link Test fail, 
not a bug
Comment 36 by bra...@gmail.com, Mar 20 2009
@thydavidcome, it plainly says that one of the requirements to pass acid3 is that
"the final page has to look exactly, pixel for pixel, like this reference rendering".
 Seeing as how the reference rendering does not contain the text "Linktest failed",
the acid3 test is failing and this definitely is a bug.
For those of you saying you get 100/100: do you ALWAYS get it? Because on my pc it's 
pretty random: sometimes I get 100/100, but other times I get 99/100, and test 80 
fails ("Test 80 failed: kungFuDeathGrip was null"), apart from the other problems 
(clicking on the A it says "Test 26 passed, but took X ms (less than 30 fps)", and 
there's also the linktest failed problem). Also, by comparing my result with the 
reference rendering, the characters that form the word "Acid" are slightly different. 
It's a matter of a few pixels, and it couldn't be noticed otherwise, but if you click 
on the tab where you have the results, and then on the one with the reference 
rendering, you can notice it. Can anybody else reproduce these problems? Especially 
the random 99-100 behaviour...

I was almost forgetting it: I'm using 2.0.170.0 (build 11755).
@ fabio.tu...@infinito.it

The same things happen to me too.
Should the «LINKTEST FAILED» notification in the Acid3 Test be regarded as indicating
a bug in Chrome - and therewith something to be resolved - or not ?...

Henri
Linktest is test 48, which we currently fail. The nature of the test is such that 
there's no easy way to report success or failure, so it's always counted as passing 
in the 100/100 count and the only indication of failure is the visible LINKTEST 
FAILED text. So we are failing it.

The solution is fairly straightforward. I will try to fix this next week.
Comment 41 by mooks...@gmail.com, Mar 20 2009
Sometimes I get 99/100, but usually I get 100/100 (w/ the linktest failed).  But this 
also happens in Opera 10 and Safari 4 - Every so often, on a refresh, it fails one 
test.  Probably not a big deal.

One thing I have noticed, however, is that the 'Acid3' text in Chrome is not a pixel-
perfect match when comparing it to the reference drawing.  It looks like there's some 
sort of font smoothing occurring that does not happen in either Safari 4 or Opera 10, 
and it gives me the impression that the pixels have actually shifted slightly.  (This 
is on build 2.0.171.0 12184)
Dglazkov, it would be wonderful if you could fix this minor bug - plus the rendering
inconsistencies that some users on this thread have reported - next week, as you have
indicated. This sort of thing is important to many of the early adopters who are
attracted to Chrome, and whose reports on the browser will be influential in getting
others to use it....

Henri  
Even without the linktest failure message I wouldn't have a pixel-perfect rendering 
because I see a slight difference in the font of the "Acid3" text.

Chromium 2.0.171.0 (12225) and Chrome 2.0.170.0
Hope the "linktest failed" issue is fixed this week.
Please stop posting comments here. 60 people have starred this issue and will get a 
mail if someone replies.
Why is this a P1? Or even a mstone-2.0 bug? As far as I can tell the failures here have 
close to zero impact on site compatibility.
Comment 47 by dhh...@gmail.com, Mar 30 2009
Because of PR reasons.  Passing Acid3 is good PR for Chrome.
Comment 48 by jon@chromium.org, Mar 31 2009
Labels: Waiver
I don't understand why passing the Acid3 test is P1, whereas passing the Acid2 is 
P2 :-)

I think the Acid2 should be more important. Also wikipedia is indicating that Chrome 
does not pass Acid2.

http://code.google.com/p/chromium/issues/detail?id=7734
Comment 50 by jon@chromium.org, Apr 6 2009
Labels: -Waiver FixIt
Wikipedia also says Chrome fails the Acid3 test and Safari passes it.
As I write this, I see 100 / 100 *BUT* the linktest fails. Not sure why but, that's the 
only issue I see in version 2.0.172.2 (Official Build 13060) everything else is 
perfect.
additionally if you do history backward and forward the test only show 96/100 or 99/100 
with linktest fail.
Odd that this issue, concerning which dglazkov wrote «The solution is fairly 
straightforward. I will try to fix this next week.» on 20 March, still has not been 
resolved ! Moreover, the issue of Chrome compliance with web standards seems to be 
more  complex than that of whether the browser passes the link test in the Acid3 
Test. At first glance, the visual artifact that prevented Chrome from being fully 
compliant with the Acid2 test - the dark bar near the nose of the «Hello World» face 
reported by subvertigo on 14 February - no longer appears in the latest version 
(2.0.174.0) and the reference version on the left looks identical to the test version 
on the right, as seen in the first attachment below. However, if one «squeezes» the 
two pictures laterally, they react very differently, as shown in the second 
attachment below. It would seem that Chrome developers need to take a further look at 
standards compliance issues....

Henri
Acid2 Test på Chrome 2.0.174.0 20090418 Squeeze.JPG
71.6 KB View Download
Acid2 Test Chrome 2.0.174.0.JPG
87.2 KB View Download
Comment 55 by phistuck@gmail.com, Apr 18 2009
@mhenriday - Even though a little off topic (since this issue is regarding ACID 3 and 
not ACID 2), I do not think the resizing issue act as grounds for bad results.
FireFox 2 and IE 6 and 7 (did not check 8) do not even pass the test, FireFox 3 (kind 
of, read below) passes the test, but resizing screws the results also.
The ACID test guys (apparently, I did not check the code though) presume that your 
display resolution has a minimum of 800x600.

On the other hand, both FireFox 3 and Chrome (whatever version) pass the test, with a 
slight error (the nose thing). Not only Chrome.
Phistuck, that FF 3 and Chrome are far more standards-compliant than IE 6, 7, or 8 is
well known, and to my mind requires no further comment here. My reasons for
mentioning Chrome's deviation from the Acid2 reference rendering in my comment 54
above are 1) subvertigo's comment 49, in which he questions making compliance with
the Acid3 linktest the main priority, and 2) that the deviation I describe might, in
fact, have consquences for users on notebooks with small screens. Let us wait and see
if we get some feedback from Chrome developers like dglazkov....

Henri
I get 98/100, possible do to my own personal settings.
As the browser needs to be at it's default for the test to work.
I am running the Developer alpha version or whatever, 2.0.174.0 (Official Build 13645)
Comment 58 Deleted
@ mhenriday

The acid2 bug is corfirmed also with the last official dev revision (2.0.174.0).
The strange thing is that Safari/WebKit have not this problem.
I've just noticed that with 2.0.174.0 test 80 is always passed (I've tried something 
like 30 times, and it has never failed) and the fonts used in the word "Acid" are 
perfect, because they match the reference rendering pixel by pixel. Excellent job! 
Now the linktest failed is the only problem left, and then Chrome will be fully Acid3 
compliant.
The only other thing that I've noticed is that, the first time that I've run it, test 
69 was slow (it was passed, but by clicking on the A it said it took too many 
iterations). All the other times, when I pressed F5, this problem didn't show up.

Btw, mhenirday, for Acid2 I still see the nose as slightly different.
Comment 61 by Deleted ...@, Apr 24 2009
Due to latency, it can take over two seconds to download all the little files the page calls for. Because 
of that, the test passed-but-is-too-slow. I'm sure they took that into account, which is why we still get 
100/100 even if the animation isn't smooth. I had about 2000 milliseconds of total latency, even after 
running it a couple times, at the points where I saw the animation jam up a bit.
The page itself is also a couple hundred KB, and it hadn't finished loading until after that first step of 
latency (empty.css).
For the record, my internet connection has some incredible latency.
But see what your document inspectors say about the time.
Comment 62 by jon@chromium.org, Apr 28 2009
Labels: -Mstone-2.0 Mstone-2.1
Changes for the linktest require some extensive rethinking of our history.  That is not 
going to happen in mstone:2.0 so moving to 2.1.
Safari and even Opera/Firefox pass without the "linktest failed" message.
As of now, with my own present build (2.0.182.0 (16468)) there are two minor things 
when trying Acid3:

1. The failed linktest
2. The purple X in the top right corner

On the positive side it seems that we are closing in on the test 26, on my computer 
(2.4 GHz P4 with 1 GB) it is just over 100 ms. 

As of now it looks like Chromium is beating all current browsers when it comes to 
speed. Sunspider scores here 1728 ms for Chromium compared to 2995 for Firefox 3.6a1 

Comment 65 by fam....@live.nl, May 21 2009
3. When you move your mouse above the last two 'squares' (the blue and purple) you also 
get a 'question-mouse' and when you click there you get the messagebox as well. 
On 2.0.181.1, I don't see the mouse-over «question mouse» mentioned in Comment 65 
above, but I do see the «LINK TEST FAILED» message and the «X» at the top right....

Henri 
Comment 67 by fam....@live.nl, May 21 2009
I also have 2.0.181.1. 

for bug part 3:

go to the acid3 testpage, wait for the test to finish and then move your mouse in the 
marked area you'll see on my screenshot. If your mouse gets in this area the cursor 
will change into an 'question-mouse'. 
acid3failure.png
13.0 KB View Download
part 3 is not a bug. look at the source code of acid3 website. the message box tells 
you which test failed
Labels: -mstone-2.1 mstone-3
Comment 70 by luc...@telenet.be, May 28 2009
In Dev version 3.0.182.2

I do not get 100/100 anymore on the acid3 test  :-(
My score is 94/100 with a purple x in the top right corner.
buildbot snapshot 17071/chrome-mac (3.0.183.0) now shows 100/100 with "linktest 
failed", same as described a while ago in comment 52 above.
Comment 72 by Deleted ...@, May 28 2009
Chrome 2.0.172:
     99     Linktest failed
FireFox 3.01:
     70
IE 8.0.6001:            
     20     
IE8 Comp mode:
     13


Comment 73 by jon@chromium.org, Jun 1 2009
Labels: -mstone-3 Mstone-4
Comment 74 by temp4...@gmail.com, Jun 13 2009
Version 3.0.187.1 seems to give 100/100 and sometimes 99/100 with linktest failed and 
and an X at the top right corner.

The report for the test:
Failed 0 of 100 tests.
Test 26 passed, but took 61ms (less than 30fps)
Test 69 passed, but took 42 attempts (less than perfect).
Total elapsed time: 4.36s
Chrome ‪3,0,187,1‬acid 3.png
16.1 KB View Download
And as I say in comment #53 if you visit the acid3 web from another previous web when 
you do history backward and then forward the test fail with 96/100 and the new chrome's 
version shows a white box in the middle.
Comment 76 by dhw@chromium.org, Jun 22 2009
 Issue 14027  has been merged into this issue.
 Issue 14054  has been merged into this issue.
Comment 78 by dhw@chromium.org, Jun 24 2009
 Issue 15124  has been merged into this issue.
Yes still 100/100 with LINKTEST Fail.
Labels: -Mstone-4 Mstone-3
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=19910 

------------------------------------------------------------------------
r19910 | dglazkov@chromium.org | 2009-07-03 10:38:39 -0700 (Fri, 03 Jul 2009) | 14 lines
Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/expire_history_backend.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/expire_history_backend.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/expire_history_backend_unittest.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/history.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/history_database.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/history_database.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/redirect_uitest.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/visit_database.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/history/visit_database.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/profile.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/profile.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/browser_render_process_host.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/browser_render_process_host.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/mock_render_process_host.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/mock_render_process_host.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_process_host.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/test_render_view_host.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/tab_contents/tab_contents.cc?r1=19910&r2=19909
   A http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_event_listener.cc
   A http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_event_listener.h
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_master.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_master.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_perftest.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/visitedlink_unittest.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/chrome.gyp?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/common/render_messages_internal.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/common/visitedlink_common.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/renderer/render_thread.cc?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/renderer/render_thread.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/webkit/glue/webview.h?r1=19910&r2=19909
   M http://src.chromium.org/viewvc/chrome/trunk/src/webkit/glue/webview_impl.cc?r1=19910&r2=19909

Fix Acid3 Test 48: LINKTEST, Chromium side....

R=brettw
BUG= http://crbug.com/231 
BUG=http://crubg.com/5160
TEST=ExpireHistoryTest.ArchiveSomeOldHistory
TEST=ExpireHistoryTest.ExpiringVisitsReader
TEST=VisitedLinkTest.Listener
TEST=VisitedLinkTest.Resizing
TEST=VisitedLinkRelayTest.*
TEST=VisitedLinkEventsTest.*


Review URL: http://codereview.chromium.org/113591
------------------------------------------------------------------------

Status: Fixed
Still getting Linktest Failed and a Purple/Red box with a white X in the upper right-
hand corner in Chromium 3.0.192.0 (19910)
Chromium 3.0.192.0 (19910) with a new emtpy profile: Linktest passed, only Purple/Red 
box with X
The Purple/Red X failures are due to downloadable fonts explicitly disabled due to 
security concerns. Run with " --enable-remote-fonts" option to get full pass.
Oh - interestingly - Linktest Failed only occurs in Incognito mode (which is my default 
here at work).  In regular browsing mode the linktest passes, but the red/purple box 
with the X persists in either mode.
The LINKTEST test will fail in incognito mode, because the history items (which is what 
is being tested) are not stored.
Understood, and thanks for the remote fonts info, as well as the 100% fix.
The LINKTEST test fails in Firefox's private browsing mode too.

Many thanks for fixing this dglazkov@
Comment 90 by zr40...@gmail.com, Jul 3 2009
The linktest failure in incognito mode doesn't really count as the test only applies to browsers in default settings.
Comment 91 Deleted
Comment 92 Deleted
Comment 93 Deleted
Comment 94 by keic...@gmail.com, Jul 4 2009
Chromium	3.0.192.0 (Developer Build 19918)
WebKit	531.3
V8	1.2.11.1
User Agent	Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US) AppleWebKit/531.3 
(KHTML, like Gecko) Chrome/3.0.192.0 Safari/531.3

First try without cache ( first visit to acid3.acidtests.org )
test 01 passed but took 37ms (less than 30fps)
test 05 passed, but took 45ms (less than 30fps)
test 26 passed, but took 255ms (less than 30fps)
test 57 passed, but took 42ms (less than 30fps)
Test 69 passed, but took 206 attempts (less than perfect)

second and more tries, this 2 tests doesnt disappear
test 26 passed, but took 277ms (less than 30fps)
test 69 passed, but took 50 attempts (less than perfect)

Linktest failed in incognito mode and the 2 consistent errors
test 26 passed, but took 277ms (less than 30fps)
test 69 passed, but took 50 attempts (less than perfect)

with --enable-remote-fonts parameter 
first visit chrome normal
test 00 passed, but took 62ms (less than 30fps)
test 26 passed, but took 230ms (less than 30fps)
test 40 passed, but took 37ms (less than 30fps)
test 46 passed, but took 38ms (less than 30fps)
test 69 passed, but took 69 attempts (less than perfect)
no more pink X at the upper right

second and more tries, this 2 tests doesnt disappear
test 26 passed, but took 277ms (less than 30fps)
test 69 passed, but took 50 attempts (less than perfect)

first try Linktest failed in incognito mode and the 2 consistent errors
test 26 passed, but took 277ms (less than 30fps)
test 40 passed, but took 34ms (less than 30fps)
test 69 passed, but took 50 attempts (less than perfect)

second and more tries, Linktest failed in incognito mode and the 2 consistent errors
test 26 passed, but took 277ms (less than 30fps)
test 69 passed, but took 50 attempts (less than perfect)

summary: there are many test passed, but not coded quite well or optimized because
if done so, the logs of acid3 will not tell those reported above. and its better to
fix too incognito mode and chrome must test for security the downloaded fonts so it 
will not be manually --enable-remote-fonts option to download those fonts

acid3 states that
To pass the test, a browser must use its default settings, the animation has to be 
smooth

i think it doesnt passed the acid3 test because the animation is not smooth

@keichee and in general -
Incognito mode does not remember the visited links *in order* to increase the 
privacy, that is exactly why the FireFox Private browsing mode fails too. It is a 
feature and not a default one.
Besides, Incognito\Private\InPrivate\whatever names this mode has in other browsers, 
is not a default mode, which is the prerequisite for the test.

@keichee -
About the slow tests - that is another thing and they should be perfect, not anything 
less, I guess.
I am running the latest Dev release, so I have not seen the success of the linktest 
yet, but for me, in the first visit to the test, test 26 took much less than in your 
version (still not good according to the ACID3 measurements) - 62ms, while for you it 
took over 200ms. All of the other tests (besides test 69) were successful, apparently 
(they did not show up in the result as not passing or flaky).
So I do not know why you got far more flaky tests (0, 1, 5, 57, 40, 46)... I will 
check in the next Dev release.
Comment 96 by fam....@live.nl, Jul 4 2009
I run the latest dev version (.190.4) on vista and the only failed test (in non-
incognito after the second visit) is test 26, with 94 ms. (and the purple X and the 
linktest failed)

If I run it the first time (cache removed) I get test 26 in 98 ms AND test 69 in 35 
attempts (and the X and the linktest)

An other thing I noticed is that when I ran it the first time is that the first halve 
second you see 'JS?' and at the bottom something like you have to enable scripts.

So in the new version (the 3.0.192.0) there are much more tests failed than in the 
'current' newest release
Comment 97 by prog...@gmail.com, Jul 4 2009
when i test it with 3.0.192.0 (Developer Build 19918)
for the first time (cache cleared)
i get test 99/100
test 26 passed, but took 47 ms (less than 30fps)
test 80 FAILED, kungFuDeathGrip was null

after a refresh:
test 26 passed, but took 47 ms (less than 30fps)
test 69 passed, but took 47 attamps (less than perfect)

i'd be lying if i say it bothers me that much ;)
Comment 98 Deleted
Comment 99 Deleted
@keichee -
(Regarding the comment you have just deleted -)
Then you got Incognito mode a little bit wrong.
It does not keep the history, not even in the same session, supposedly (go to the 
History page in Incognito, no entries are added, for example).
incognito mode just same as the normal version, its just deletes the history on 
process exit.

please see the picture attached, im using a new netbook. i submitted an issue also 
that chrome doesnt utilize fully the multiprocessor of a pc, as its uses only 50% of 
the cpu usage. and only threads on one processor. 

Yes the processor speed is slow, but when 2 cores of a multiprocessor pc are 
utilized, i wont get slow speeds just like i posted earlier.

A fully optimized code will use 100% if not more of the cpu capabilities rather than 
suppressing the error by means of using high end computers imho.
chromebug02.JPG
180 KB View Download
@phistuck

yes thats based on coding on incognito mode, but when the users, not developers uses 
it, its just the same goal, to supress the history.
@keichee - not really, keeping history entries during the session may decrease privacy.
incognito states:
You've gone incognito. Pages you view in this window won't appear in your browser 
history or search history, and they won't leave other traces, like cookies, on your 
computer after you close the incognito window. Any files you download or bookmarks 
you create will be preserved, however. 

so if i am right, the words 'after you close the incognito window' will only do 
states above when the corresponding window closes, but when testing the acid3, i have 
not yet closed the corresponding window yet, thus 'Pages you view in this window 
won't appear in your browser history or search history, and they won't leave other 
traces, like cookies, on your computer' is not deleted yet so theoretically acid3 can 
access that specific history until the 'after you close the incognito window' have 
been executed.
when surfing private stuff using incognito and suddenly someone sees the opened private 
stuff in your browser because you have not closed it yet, then your privacy is 
breached. 

why not make the history available to that url only and limits its access to the files 
to that specific url only for acid3 to work. 
@keichee
Actually, that may be relate to only the last part of the sentence, but it does not really 
matter.
Incognito is no default, immediate (or not immediate) visited link marking (which is the 
needed 'feature' here to pass the test, as far as I know) will be making the private 
browsing less private.
No need to fix this.
"after you close the incognito window"  only applies to  "won't leave other traces, 
like cookies, on your computer"


the normal chrome perhaps needed a fix, so that there will be no warnings in the acid3 
tests. not less
There are a shit pink-red  in top-left corner but 100/100 and none linktest-fail with 
the chomium build n°19938

Escuse me for my poor english cause I'm french !
why this issue is not yet corrected in latest dev version? linktest still fail with 
3.0.191.3
Passes in newer builds, but with red x in top right corner and it says passed slowly.
See screen shot.
by William Cody Winter on 2009-07-07 at 00.50.56
359 KB View Download
p.s the build I tested above with was 3.0.193 (193.0)
It still fails one test from time to time:

Failed 1 tests.
Test 26 passed, but took 50ms (less than 30fps)
Test 80 failed: kungFuDeathGrip was null
Total elapsed time: 0.23s

rev. 19913

Comment 114 by fam....@live.nl, Jul 7 2009
In 3.0.191.3 I get an other bug extra. See the screenshot. It only happens at 62 
percent, at 63 everything is normal again.
I hadn't have this on 3.0.190.4
acid3-62percent-bug.png
27.6 KB View Download
Sorry, this cannot be considered fixed.

2 tests are slow (test 26, test 69), the progress is not smooth, and the biggest thing 
- there is a huge white cross on a pink background in the top right corner, which 
doesn't exist in reference rendering.
Comment 116 by dhw@chromium.org, Jul 8 2009
The white X will be there for now unless you run with --enable-remote-fonts or  Issue 9633  
gets resolved.  For details, see:
http://groups.google.com/group/chromium-dev/browse_thread/thread/c3b3db4339d3642c#
Still. As long as the X is there this issue can i no way be considered "fixed".
There area also the issues with test 26 and 69 being slow.
Comment 118 by keic...@gmail.com, Jul 19 2009
this will be considered fixed if the community see the procedure flawless without any 
parameters,not by a technical fix only, hope somebody will agree.
As I said before do this tests.

=== First ===
1) Visit http://acid3.acidtests.org/ (100/100)
2) Visit www.google.com
3) Visit again http://acid3.acidtests.org/ (99/100)

=== Second ===
1) Visit http://acid3.acidtests.org/ (100/100)
2) Visit www.google.com
3) Press history back button (96/100 with a white box in the center)
Comment 120 by zr40...@gmail.com, Jul 19 2009
Re comment 119:

With a fresh tab for each sequence:
The first sequence results in 100/100 both times.
The second sequence results in 100/100 both times.

I did notice that occasionally test 80 fails with "Test 80 failed: kungFuDeathGrip 
was null". The test passed when performing the same sequence again in another fresh 
tab.

This is using 3.0.193.1 (Official Build 20697).
Yes, but then Chrome doesn't clean the previous page correctly or the history back 
button don't show the previous page as you seen before.
Looks to me as if the latest Chrome for Linux (3.0.194.3-r20926) now passes the Acid3 
Test - 100/100 with no «FAIL» message. Splendid ! Now if only we could get it to work 
well with SCIM !...

Henri
When I tried the following on version 3.0.193.1 on Windows:

=== First ===
1) Visit http://acid3.acidtests.org/ 
2) Visit www.google.com
3) Visit again http://acid3.acidtests.org/ 

=== Second ===
1) Visit http://acid3.acidtests.org/ 
2) Visit www.google.com
3) Press history back button 

I get 100/100 every time. However, the first time (presumably with an empty cache) 
there are a few tests that take more time than expected. The second time, everything 
appears to be fine, except the pink/white X in the top right corner which is 
displayed every time the test is run.
Comment 124 by keic...@gmail.com, Jul 24 2009
need to patch the flaw of the downloaded fonts so that chrome will not execute viruses 
when a font is downloaded, thats the reason why theres an X at the top right corner.
For the past several builds (~about 22200+) Acid3 has been pretty messed up. I’ve 
attached a screenshot of what it looks like on 3.0.198.0 (22483).

Acid3.PNG
32.3 KB View Download
 Issue 24667  has been merged into this issue.
this should be reopened, the test now fails again, the page is messed up.
i am getting 100/100. 
I usually get 100/100 (sometimes 99/100), but the link test always fails....

Henri
@karen

Click on the "A" in the test to see where the test didn't perform as perfectly as it 
could.

Also, as of right now, a pink "X" is rendered at the top right of the test, which if I 
recall correctly, has to do with Chrome not yet supporting embedding of fonts.

So Chrome still has a little bit more to go before I'd call this fixed.
i clicked on the A and here's what i get:

Failed 0 tests.
Test 00 passed, but took 75ms (less than 30fps)
Test 26 passed, but took 47ms (less than 30fps)
Test 46 passed, but took 44ms (less than 30fps)
Total elapsed time: 0.69s

and i clicked on the link too and my image matches 100%. I am on the mac. 

the pink X is there however.
apparently the issue only happens on windows xp and windows seven. I'm not talking 
about the result of the test (that is always 98/100 for me by the way...) but the fact 
that the page is completly messed up. I opened a bug that is confirmed by a member of 
google. tracked as issue http://code.google.com/p/chromium/issues/detail?id=25031


mhenriday, don't get missleaded by the "100/100". To pass the test, it is written you
must compare to the reference rendering, and the reference rendering not only shows
100/100, but a clean white upper-right corner. As the pink X is not on reference
rendering, seeing it is equivalent to a fail.

Comment 134 by dhw@chromium.org, Oct 18 2009
@dherbert30, joachim.e.rousseau, and Karen, the pink X problem is different, and is 
because web-fonts are currently disabled pending a security review and fix.  For 
details, see  Issue 17818 
Comment 135 by dhw@chromium.org, Oct 18 2009
This issue for Chrome to pass Acid3 test has been fixed and should remain closed.  Please submit new 
bug reports for any new issues.

@fabios4rulez, thanks for submitting the new regression bug in  Issue 25031 .  Further testing shows 
that this regression is related to existing profile data from past uses of Chrome.

If you start Chrome with a new profile (eg. "C:\path\to\chrome.exe" --user-data-dir=C:\new\directory)
the problem in  Issue 25031  does not occur.
Thanks for the update, dhw@chromium.org !...

Henri
reply:
Comment 115 by mahaniok, Jul 08, 2009
Sorry, this cannot be considered fixed.

2 tests are slow (test 26, test 69), the progress is not smooth, and the biggest thing 
- there is a huge white cross on a pink background in the top right corner, which 
doesn't exist in reference rendering.
Last note is still true - the white cross on magenta and red...
Comment 138 Deleted
Comment 139 by Deleted ...@, Nov 23 2009
In Latest Development Build 4.0.249.4 The Downloadable Fonts has been enabled and now 
chrome have issue with only tests 26 and 69, 
please if any one can show it in wikipedia acid 3 page and showing there the sources in 
Desktop browser progress for the Acid3 test Table.
Thanks.
Comment 140 Deleted
Labels: -Fixit bulkmove TaskForce-Fixit
Product Version      : 0.2.149.27
URLs (if applicable) : http://acid3.acidtests.org/
Other browsers tested:
Add OK or FAIL after other browsers where you have tested this issue:
     Safari 3: fail
    Firefox 3: fail
         IE 7: epic fail

What steps will reproduce the problem?
1. Visit: http://acid3.acidtests.org/
2. Wait for the test to run
3. Note that the result doesn't match the reference rendering: http://acid3.acidtests.org/reference.html

What is the expected result?
The test result should match the reference rendering here: http://acid3.acidtests.org/reference.html

What happens instead?
It fails with 63/100 test passed, when I tried it.

Acidtests.org appears to be very bust at the moment (odd that :-) so this might 
be affecting things a bit. I though that the webkit dev version passed acid3?

Yeah - here we are: http://webkit.org/blog/173/webkit-achieves-acid3-100100-in-
public-build/ - webkit blog entry from March 08 about how the webkit nightlies 
now pass acid3. What's happened since then?
Labels: Restrict-AddIssueComment-Commit
Project Member Comment 143 by bugdroid1@chromium.org, Mar 10 2013
Labels: -Area-WebKit -Mstone-3 Cr-Content M-3
Project Member Comment 144 by bugdroid1@chromium.org, Mar 13 2013
Labels: -Restrict-AddIssueComment-Commit Restrict-AddIssueComment-EditIssue
Project Member Comment 145 by bugdroid1@chromium.org, Apr 6 2013
Labels: -Cr-Content Cr-Blink
Sign in to add a comment