<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<!DOCTYPE bugzilla SYSTEM "https://bugs.webkit.org/page.cgi?id=bugzilla.dtd">

<bugzilla version="5.0.4.1"
          urlbase="https://bugs.webkit.org/"
          
          maintainer="admin@webkit.org"
>

    <bug>
          <bug_id>49523</bug_id>
          
          <creation_ts>2010-11-14 17:40:14 -0800</creation_ts>
          <short_desc>IME candidate characters inserted unexpectedly to textarea which listens compositionupdate event</short_desc>
          <delta_ts>2011-02-24 05:00:53 -0800</delta_ts>
          <reporter_accessible>1</reporter_accessible>
          <cclist_accessible>1</cclist_accessible>
          <classification_id>1</classification_id>
          <classification>Unclassified</classification>
          <product>WebKit</product>
          <component>HTML Editing</component>
          <version>528+ (Nightly build)</version>
          <rep_platform>PC</rep_platform>
          <op_sys>All</op_sys>
          <bug_status>NEW</bug_status>
          <resolution></resolution>
          
          
          <bug_file_loc></bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords></keywords>
          <priority>P2</priority>
          <bug_severity>Normal</bug_severity>
          <target_milestone>---</target_milestone>
          
          
          <everconfirmed>1</everconfirmed>
          <reporter name="Kenichi Ishibashi">bashi</reporter>
          <assigned_to name="Nobody">webkit-unassigned</assigned_to>
          <cc>ap</cc>
    
    <cc>bashi</cc>
    
    <cc>darin</cc>
    
    <cc>dbates</cc>
    
    <cc>ddavidso</cc>
    
    <cc>enrica</cc>
    
    <cc>hamaji</cc>
    
    <cc>hayato</cc>
    
    <cc>jiapu.mail</cc>
    
    <cc>justin.garcia</cc>
    
    <cc>mitz</cc>
    
    <cc>rniwa</cc>
    
    <cc>simon.fraser</cc>
    
    <cc>yuzo</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>309144</commentid>
    <comment_count>0</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-14 17:40:14 -0800</bug_when>
    <thetext>Steps to repro:
1. Open the attached HTML file.
2. In the textarea, type a Japanese character sequence like &quot;てすと&quot; [tesuto] with IME enabled.

Expected result:
An uncommitted IME composition text &quot;てすと&quot; should appear and any other committed characters should not be inserted in the textarea.

Actual result:
Some committed characters appears with IME composition text (See the attached image).</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309145</commentid>
    <comment_count>1</comment_count>
      <attachid>73866</attachid>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-14 17:41:24 -0800</bug_when>
    <thetext>Created attachment 73866
An HTML file to reproduce</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309146</commentid>
    <comment_count>2</comment_count>
      <attachid>73867</attachid>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-14 17:42:35 -0800</bug_when>
    <thetext>Created attachment 73867
Actual result</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309148</commentid>
    <comment_count>3</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-14 17:55:50 -0800</bug_when>
    <thetext>The cause of this bug is almost the same as the bug 46868. A similar text mismatch could occur when the textarea listens the compositonupdate event and changes its style on the event handler. This bug could occur not only on Mac but on other platforms. I&apos;ll post a trial patch to fix this, but I&apos;m not confident whether the fix is suitable. I&apos;d like to ask some advice.

Thanks,</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309149</commentid>
    <comment_count>4</comment_count>
      <attachid>73868</attachid>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-14 17:56:56 -0800</bug_when>
    <thetext>Created attachment 73868
Trial patch V0</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309337</commentid>
    <comment_count>5</comment_count>
    <who name="Alexey Proskuryakov">ap</who>
    <bug_when>2010-11-15 10:27:19 -0800</bug_when>
    <thetext>This seems to be a patch for a Cromium-specific problem in cross platform code. Why is that necessary? What does Chromium do differently to trigger this problem? Could it be a bug in Chromium code?

I guess I have the same concern about bug 46868, where this new call to updateStyleIfNeeded() has been added.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>309539</commentid>
    <comment_count>6</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-11-15 18:04:35 -0800</bug_when>
    <thetext>Hi,

(In reply to comment #5)
&gt; This seems to be a patch for a Cromium-specific problem in cross platform code. Why is that necessary? What does Chromium do differently to trigger this problem? Could it be a bug in Chromium code?

Whether this bug occurs depends on when the event handler is invoked. The primary cause of this bug is text mismatch of the textarea element and its render object. This mismatch will occur when the user agent invokes the event handler which changes style of the textarea before updating the composition text. Since the stable version of Safari invokes the event handler after updating composition text, this bug doesn&apos;t apper on it. However, IMHO, the text mismatch should not occur irrespective of the order of the process because there is no dependency between invoking event handlers that changes styles and updating composition text. So I think this bug is not only a Chromium specific issue but also a issue on other platforms. Actually, Webkit nightly builds (r71897 for Mac and r71499 for Windows) seem to have same problem. I tried to input some Japanese text with IME on the attached HTML page, the nightly builds inserted uncommitted composition text unexpectedly, even if the stable version of Safari doesn&apos;t do that.

Anyway, I&apos;m not really sure about the fix. There might be better place to call updateStyleIfNeeded(). So please let me know if you have another solution.

Regards,</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>317454</commentid>
    <comment_count>7</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2010-12-05 23:03:37 -0800</bug_when>
    <thetext>Could anyone take a look at this?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>317457</commentid>
    <comment_count>8</comment_count>
    <who name="Alexey Proskuryakov">ap</who>
    <bug_when>2010-12-05 23:25:02 -0800</bug_when>
    <thetext>I&apos;m not even sure if this is a bug (same about bug 46868). Changing style in a sync event handler is just an invitation for trouble.

The general direction has been to not dispatch synchronous event during DOM manipulation, and perhaps we just shouldn&apos;t dispatch compositionupdate synchronously (or at all).</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>317625</commentid>
    <comment_count>9</comment_count>
      <attachid>73868</attachid>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2010-12-06 07:57:20 -0800</bug_when>
    <thetext>Comment on attachment 73868
Trial patch V0

View in context: https://bugs.webkit.org/attachment.cgi?id=73868&amp;action=review

(In reply to comment #8)
&gt; I&apos;m not even sure if this is a bug (same about bug 46868). Changing style in a sync event handler is just an invitation for trouble.
&gt; The general direction has been to not dispatch synchronous event during DOM manipulation, and perhaps we just shouldn&apos;t dispatch compositionupdate synchronously (or at all).

Unlike DOM mutation events which could fire in the middle of executing editing commands, IME composition events are fired per user actions.  
So I don&apos;t think it&apos;ll cause too much trouble for us even it fired synchronously just like keydown or mousedown although we must be extremely careful when to fire them.

&gt; LayoutTests/platform/chromium/fast/text/chromium-duplicate-ime-composition.html:1
&gt; +&lt;html&gt;

Why is this test under chromium?  I can reproduce the same issue on WebKit r72487.  Shouldn&apos;t we test it on all platforms?

&gt; WebCore/editing/Editor.cpp:-1601
&gt; -    // Updates styles before setting selection for composition to prevent
&gt; -    // inserting the previous composition text into text nodes oddly.
&gt; -    // See https://bugs.webkit.org/show_bug.cgi?id=46868
&gt; -    m_frame-&gt;document()-&gt;updateStyleIfNeeded();
&gt; -

I don&apos;t quite get why this change is correct.  Could you at least explain why you&apos;re removing this?  Because as far as I checked, we don&apos;t always call deleteTextFromNode when we call setComposition.  I want to know why moving this layout update as supposed to adding another one in replaceTextInNode is correct.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>326217</commentid>
    <comment_count>10</comment_count>
      <attachid>73868</attachid>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2010-12-23 12:31:34 -0800</bug_when>
    <thetext>Comment on attachment 73868
Trial patch V0

r- because I want to see the test being ran on all platforms given that the bug reproduced on mac port.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>329388</commentid>
    <comment_count>11</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-01-04 19:50:59 -0800</bug_when>
    <thetext>Hi Niwa-san,

Thank you for reviewing and sorry for late reply. I&apos;ve checked this problem with ToT and I couldn&apos;t reproduce the problem on both Safari and Chromium. I cannot figure out which change fix the problem, but it might be good to withdraw the review for now.

Regards,

(In reply to comment #10)
&gt; (From update of attachment 73868 [details])
&gt; r- because I want to see the test being ran on all platforms given that the bug reproduced on mac port.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>329394</commentid>
    <comment_count>12</comment_count>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2011-01-04 20:27:11 -0800</bug_when>
    <thetext>(In reply to comment #11)
&gt; Hi Niwa-san,
&gt; 
&gt; Thank you for reviewing and sorry for late reply. I&apos;ve checked this problem with ToT and I couldn&apos;t reproduce the problem on both Safari and Chromium. I cannot figure out which change fix the problem, but it might be good to withdraw the review for now.

Could be https://bugs.webkit.org/show_bug.cgi?id=51693.  It&apos;ll be still nice to check in a test to make sure this doesn&apos;t regress in the future.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>329419</commentid>
    <comment_count>13</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-01-04 21:49:31 -0800</bug_when>
    <thetext>Hi Niwa-san,

(In reply to comment #12)
&gt; Could be https://bugs.webkit.org/show_bug.cgi?id=51693.  It&apos;ll be still nice to check in a test to make sure this doesn&apos;t regress in the future.

Thank you for the information and suggestion. I agree with you adding a test for this problem, but unfortunately, DRT doesn&apos;t provide functions for testing this problem except for Chromium port for now. Would it be worth adding functions for this problem in other ports?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>329423</commentid>
    <comment_count>14</comment_count>
    <who name="Darin Adler">darin</who>
    <bug_when>2011-01-04 22:20:26 -0800</bug_when>
    <thetext>(In reply to comment #13)
&gt; I agree with you adding a test for this problem, but unfortunately, DRT doesn&apos;t provide functions for testing this problem except for Chromium port for now. Would it be worth adding functions for this problem in other ports?

Sure.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>329429</commentid>
    <comment_count>15</comment_count>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2011-01-04 22:35:06 -0800</bug_when>
    <thetext>(In reply to comment #13)
&gt; Thank you for the information and suggestion. I agree with you adding a test for this problem, but unfortunately, DRT doesn&apos;t provide functions for testing this problem except for Chromium port for now. Would it be worth adding functions for this problem in other ports?

I agree with Darin that adding such features to DRT will be worthwhile.  Even if you could not implement it for all ports for now, implementing it for Mac &amp; Windows ports will give us much greater coverage.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>338904</commentid>
    <comment_count>16</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-01-24 02:19:44 -0800</bug_when>
    <thetext>Hi,

I&apos;ve tried to adding the feature for mac, but I can&apos;t come up with how to implement it. I tried to almost the same approach as chromium: introducing a method that sends keydown event with keycode = 0xE5 , then invokes editor()-&gt;setComposition() by calling [textInput setMarkedText:selectedRange:], into TextInputController. However, it doesn&apos;t work as expected. When I call the introduced method with a string, I expect that DRT replaces the text of the textarea to given string, but it just appends the string..

Is there any way to emulate IME input behavior on DRT mac (and Windows) port?

(In reply to comment #15)
&gt; (In reply to comment #13)
&gt; &gt; Thank you for the information and suggestion. I agree with you adding a test for this problem, but unfortunately, DRT doesn&apos;t provide functions for testing this problem except for Chromium port for now. Would it be worth adding functions for this problem in other ports?
&gt; 
&gt; I agree with Darin that adding such features to DRT will be worthwhile.  Even if you could not implement it for all ports for now, implementing it for Mac &amp; Windows ports will give us much greater coverage.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>352639</commentid>
    <comment_count>17</comment_count>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2011-02-17 00:15:34 -0800</bug_when>
    <thetext>(In reply to comment #16)
&gt; I&apos;ve tried to adding the feature for mac, but I can&apos;t come up with how to implement it. I tried to almost the same approach as chromium: introducing a method that sends keydown event with keycode = 0xE5 , then invokes editor()-&gt;setComposition() by calling [textInput setMarkedText:selectedRange:], into TextInputController. However, it doesn&apos;t work as expected. When I call the introduced method with a string, I expect that DRT replaces the text of the textarea to given string, but it just appends the string..
&gt; 
&gt; Is there any way to emulate IME input behavior on DRT mac (and Windows) port?

Enrica, Dan, Darin, or Justin: do you know the answer to this question?  If adding the said feature to DRT is too hard, it might make sense to land the current patch as is.  Although I still want explanation as to why moving updateLayout from one place to another fixes the bug.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>352676</commentid>
    <comment_count>18</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-02-17 01:27:23 -0800</bug_when>
    <thetext>Hi Niwa-san,

Thank you for taking care of this bug. 

(In reply to comment #17)
&gt; Although I still want explanation as to why moving updateLayout from one place to another fixes the bug.

During investigating another chromium bug (http://crbug.com/66848), I realized that the essential cause of this bug is not IME related stuff. Under some particular situation, modifying style attribute in event handlers causes an inconsistency between m_value of HTMLTextAreaElement and the return value of RenderTextControl::text() even if HTMLTextAreaElement::formControlValueMachesRenderer() returns true. It is very complicate to explain why such inconsistency could occur. In short, the reason is not calling HTMLTextAreaElement::setFormControlValueMachesRenderer(false) immediately after inserting text into the element.

As far as I investigated, the reason why Safari doesn&apos;t have such problems seems that the order of calling TypingCommand::markMisspellingsAfterTyping() and Editor::appliedEditing() in TypingCommand::typingAddedToOpenCommand().

I&apos;ll file a bug for http://crbug.com/66848 to bugs.webkit.org and try to explain a little more minutely.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>352701</commentid>
    <comment_count>19</comment_count>
    <who name="Ryosuke Niwa">rniwa</who>
    <bug_when>2011-02-17 02:39:13 -0800</bug_when>
    <thetext>(In reply to comment #18)
&gt; During investigating another chromium bug (http://crbug.com/66848), I realized that the essential cause of this bug is not IME related stuff. Under some particular situation, modifying style attribute in event handlers causes an inconsistency between m_value of HTMLTextAreaElement and the return value of RenderTextControl::text() even if HTMLTextAreaElement::formControlValueMachesRenderer() returns true.

Ah, I see.  But doesn&apos;t that mean your initial approach is incorrect?  Do you think fixing the root cause will help resolve this bug as well?

&gt; It is very complicate to explain why such inconsistency could occur. In short, the reason is not calling HTMLTextAreaElement::setFormControlValueMachesRenderer(false) immediately after inserting text into the element.

Oops! That sounds like a really bad bug.

&gt; As far as I investigated, the reason why Safari doesn&apos;t have such problems seems that the order of calling TypingCommand::markMisspellingsAfterTyping() and Editor::appliedEditing() in TypingCommand::typingAddedToOpenCommand().

Ok.

&gt; I&apos;ll file a bug for http://crbug.com/66848 to bugs.webkit.org and try to explain a little more minutely.

Looking forward to see it.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>352728</commentid>
    <comment_count>20</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-02-17 03:35:54 -0800</bug_when>
    <thetext>Hi All,

(In reply to comment #19)
&gt; But doesn&apos;t that mean your initial approach is incorrect?  Do you think fixing the root cause will help resolve this bug as well?

I agree that fixing the root cause will help even this bug is no longer occurred. However, I don&apos;t think my initial approach is appropriate way to fix the problem now. I think we can fix the cause more suitable way.

&gt; &gt; As far as I investigated, the reason why Safari doesn&apos;t have such problems seems that the order of calling TypingCommand::markMisspellingsAfterTyping() and Editor::appliedEditing() in TypingCommand::typingAddedToOpenCommand().
&gt; 
&gt; Ok.

As for this, I&apos;d like to ask Apple folks about the changeset 43419 (http://trac.webkit.org/changeset/43419). Could you give me the reason why this change is needed? As far as I tested, calling Editor::appliedEditing() before markMisspelingsAfterTyping() will fix the problem because appliedEditing() will promptly match the m_value of HTMLTextAreaElement and text value of its renderer. I think calling appliedEditing() before markMisspelingsAfterTyping() is simple and most suitable way to fix the problem.

Regards,</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>352961</commentid>
    <comment_count>21</comment_count>
    <who name="Alexey Proskuryakov">ap</who>
    <bug_when>2011-02-17 09:17:38 -0800</bug_when>
    <thetext>Sounds like Safari on 10.5 and on Windows should have the same issue as Chromium then.

Yes, I can tell you what the steps to reproduce were in more detail. By the way, Radar number is wrong in  r43419, it should be &lt;rdar://problem/6864072&gt;.

1. Open Mail or open Safari and login to GMail
2. Compose a new message
3. Type &quot;This doesn&apos;t work.&quot;

Results: &quot;doesn&quot; is underlined as a misspelling.

I don&apos;t know why there is a PLATFORM(MAC) check, it seems wrong. The second code path that was added in r43419 was needed to fix a bug on Mac OS X 10.5.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>355161</commentid>
    <comment_count>22</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-02-22 03:10:52 -0800</bug_when>
    <thetext>Hi Alexey,

Thank you for your prompt response and sorry for late reply since I was off couple of days.

&gt; Sounds like Safari on 10.5 and on Windows should have the same issue as Chromium then.

I tried reproduce the problem with Safari on Windows, but it didn&apos;t reproduce. It seems that there is another factor in the cause. I&apos;ll continue to investigate it.

&gt; Yes, I can tell you what the steps to reproduce were in more detail. By the way, Radar number is wrong in  r43419, it should be &lt;rdar://problem/6864072&gt;.
&gt; 
&gt; 1. Open Mail or open Safari and login to GMail
&gt; 2. Compose a new message
&gt; 3. Type &quot;This doesn&apos;t work.&quot;
&gt; 
&gt; Results: &quot;doesn&quot; is underlined as a misspelling.

I see.

&gt; I don&apos;t know why there is a PLATFORM(MAC) check, it seems wrong. The second code path that was added in r43419 was needed to fix a bug on Mac OS X 10.5.

So I think removing PLATFORM(MAC) is better to fix the bug described at http://crbug.com/66848. I&apos;d like to ask your opinion.

Thanks,</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>355670</commentid>
    <comment_count>23</comment_count>
    <who name="Alexey Proskuryakov">ap</who>
    <bug_when>2011-02-22 15:59:51 -0800</bug_when>
    <thetext>See also: bug 54975, which looks closely related to the history of this change.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>355699</commentid>
    <comment_count>24</comment_count>
    <who name="Jia Pu">jiapu.mail</who>
    <bug_when>2011-02-22 16:44:49 -0800</bug_when>
    <thetext>(In reply to comment #22)
&gt; &gt; I don&apos;t know why there is a PLATFORM(MAC) check, it seems wrong. The second code path that was added in r43419 was needed to fix a bug on Mac OS X 10.5.
&gt; 
&gt; So I think removing PLATFORM(MAC) is better to fix the bug described at http://crbug.com/66848. I&apos;d like to ask your opinion.
&gt; 
&gt; Thanks,

(In reply to comment #23)
&gt; See also: bug 54975, which looks closely related to the history of this change.

The change for bug 54975 is in a Mac specific function. I&apos;m not sure how relevant it is to this bug.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>355702</commentid>
    <comment_count>25</comment_count>
    <who name="Alexey Proskuryakov">ap</who>
    <bug_when>2011-02-22 16:49:31 -0800</bug_when>
    <thetext>Jia, this bug deals with effects of a fix for &lt;rdar://problem/6864072&gt;, &quot;Contraction base marked as misspelled even though contraction is a word.&quot; I don&apos;t know if there is any direct relation, but the symptoms are definitely similar.

Regardless of whether there is any relation, your comments about this bug would be appreciated. In particular, I&apos;m not sure about potential effects a fix could have on Safari for Windows, as well as on older Mac OS X versions.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>355893</commentid>
    <comment_count>26</comment_count>
    <who name="Jia Pu">jiapu.mail</who>
    <bug_when>2011-02-22 23:02:01 -0800</bug_when>
    <thetext>(In reply to comment #25)
&gt; Jia, this bug deals with effects of a fix for &lt;rdar://problem/6864072&gt;, &quot;Contraction base marked as misspelled even though contraction is a word.&quot; I don&apos;t know if there is any direct relation, but the symptoms are definitely similar.
&gt; 
&gt; Regardless of whether there is any relation, your comments about this bug would be appreciated. In particular, I&apos;m not sure about potential effects a fix could have on Safari for Windows, as well as on older Mac OS X versions.

I added Doug to the CC list. He&apos;s the author of changeset 43419, and may want to add my comment below.

We&apos;d recommend to use caution when removing &quot;PLATFORM(MAC)&quot;. Looking at changeset 43419 alone can be misleading. Prior to changeset 42911 (http://trac.webkit.org/changeset/42911/trunk/WebCore/editing/TypingCommand.cpp), markMisspellingsAfterTyping() is called before appliedEditing() on all platforms. During Snow Leopard development, the order of these two calls are switched in changeset 42911. Upon realizing that the switching has caused regression, the preprocessor was added for Snow Leopard. In other words, appliedEditing() is called before markMisspellingAfterTyping() only on Snow Leopard. So what changeset 43419 does is simply making sure the behavior on non-SL platforms is what is used to be.

Given this history, it seems removing &quot;PLATFORM(MAC)&quot; will not change the order of these two function calls on older Mac OS X versions. But it will change the behavior on other platforms, including Safari for Windows. And the new behavior hasn&apos;t been exercised on those platforms before.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>356895</commentid>
    <comment_count>27</comment_count>
    <who name="Kenichi Ishibashi">bashi</who>
    <bug_when>2011-02-24 05:00:53 -0800</bug_when>
    <thetext>Filed a bug for http://crbug.com/66848

https://bugs.webkit.org/show_bug.cgi?id=55133</thetext>
  </long_desc>
      
          <attachment
              isobsolete="0"
              ispatch="0"
              isprivate="0"
          >
            <attachid>73866</attachid>
            <date>2010-11-14 17:41:24 -0800</date>
            <delta_ts>2010-11-14 17:41:24 -0800</delta_ts>
            <desc>An HTML file to reproduce</desc>
            <filename>repro.html</filename>
            <type>text/html</type>
            <size>305</size>
            <attacher name="Kenichi Ishibashi">bashi</attacher>
            
              <data encoding="base64">PGh0bWw+CjxoZWFkPgo8c2NyaXB0PgpmdW5jdGlvbiBpbml0KCkgewogIHZhciB0YSA9IGRvY3Vt
ZW50LmdldEVsZW1lbnRCeUlkKCd0YScpOwogIHZhciBjaGFuZ2VTdHlsZSA9IGZ1bmN0aW9uKGUp
IHsKICAgIHRhLnN0eWxlWydvdmVyZmxvdyddID0gJ2hpZGRlbic7CiAgfTsKICB0YS5hZGRFdmVu
dExpc3RlbmVyKCdjb21wb3NpdGlvbnVwZGF0ZScsIGNoYW5nZVN0eWxlKTsKfQo8L3NjcmlwdD4K
PC9oZWFkPgo8Ym9keSBvbmxvYWQ9ImluaXQoKSI+Cjx0ZXh0YXJlYSBpZD0idGEiPjwvdGV4dGFy
ZWE+CjwvYm9keT4KPC9odG1sPgo=
</data>

          </attachment>
          <attachment
              isobsolete="0"
              ispatch="0"
              isprivate="0"
          >
            <attachid>73867</attachid>
            <date>2010-11-14 17:42:35 -0800</date>
            <delta_ts>2010-11-14 17:42:35 -0800</delta_ts>
            <desc>Actual result</desc>
            <filename>result.png</filename>
            <type>image/png</type>
            <size>4452</size>
            <attacher name="Kenichi Ishibashi">bashi</attacher>
            
              <data encoding="base64">iVBORw0KGgoAAAANSUhEUgAAAK8AAAAwCAIAAADowL20AAAKd2lDQ1BJQ0MgUHJvZmlsZQAAeAHV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</data>

          </attachment>
          <attachment
              isobsolete="0"
              ispatch="1"
              isprivate="0"
          >
            <attachid>73868</attachid>
            <date>2010-11-14 17:56:56 -0800</date>
            <delta_ts>2010-12-23 12:31:34 -0800</delta_ts>
            <desc>Trial patch V0</desc>
            <filename>bug-49523-20101115105654.patch</filename>
            <type>text/plain</type>
            <size>9559</size>
            <attacher name="Kenichi Ishibashi">bashi</attacher>
            
              <data encoding="base64">ZGlmZiAtLWdpdCBhL0xheW91dFRlc3RzL0NoYW5nZUxvZyBiL0xheW91dFRlc3RzL0NoYW5nZUxv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</data>
<flag name="review"
          id="64383"
          type_id="1"
          status="-"
          setter="rniwa"
    />
          </attachment>
      

    </bug>

</bugzilla>