<?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>44637</bug_id>
          
          <creation_ts>2010-08-25 14:46:34 -0700</creation_ts>
          <short_desc>REGRESSION(r64712): Safari fails to show most of the contents of http://www.milliyet.com.tr</short_desc>
          <delta_ts>2011-02-16 01:08:28 -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>WebCore Misc.</component>
          <version>528+ (Nightly build)</version>
          <rep_platform>PC</rep_platform>
          <op_sys>OS X 10.5</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>FIXED</resolution>
          
          
          <bug_file_loc>http://www.milliyet.com.tr/2010/08/25/index.html</bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords>HasReduction, InRadar, Regression</keywords>
          <priority>P2</priority>
          <bug_severity>Normal</bug_severity>
          <target_milestone>---</target_milestone>
          
          <blocked>41115</blocked>
          <everconfirmed>1</everconfirmed>
          <reporter name="Mark Rowe (bdash)">mrowe</reporter>
          <assigned_to name="Adam Barth">abarth</assigned_to>
          <cc>abarth</cc>
    
    <cc>aestes</cc>
    
    <cc>ap</cc>
    
    <cc>bdakin</cc>
    
    <cc>commit-queue</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>269519</commentid>
    <comment_count>0</comment_count>
    <who name="Mark Rowe (bdash)">mrowe</who>
    <bug_when>2010-08-25 14:46:34 -0700</bug_when>
    <thetext>After &lt;http://trac.webkit.org/changeset/64712&gt;, which enabled the HTML5 tree builder, pages on milliyet.com.tr are displayed as mostly blank.  See &lt;http://www.milliyet.com.tr/2010/08/25/index.html&gt; for an example of one such page.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>269521</commentid>
    <comment_count>1</comment_count>
    <who name="Mark Rowe (bdash)">mrowe</who>
    <bug_when>2010-08-25 14:47:23 -0700</bug_when>
    <thetext>&lt;rdar://problem/8354793&gt;</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>269524</commentid>
    <comment_count>2</comment_count>
    <who name="Adam Barth">abarth</who>
    <bug_when>2010-08-25 14:53:30 -0700</bug_when>
    <thetext>Will investigate.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>270646</commentid>
    <comment_count>3</comment_count>
    <who name="Adam Barth">abarth</who>
    <bug_when>2010-08-27 12:02:43 -0700</bug_when>
    <thetext>The site is doing a bad User Agent detect.  A few places in the page says:

 if(navigator.userAgent.indexOf(&quot;Firefox&quot;)==-1 &amp;&amp; navigator.userAgent.indexOf(&quot;Chrome&quot;)==-1) 

If you change that to

 if(navigator.userAgent.indexOf(&quot;Firefox&quot;)==-1 &amp;&amp; navigator.userAgent.indexOf(&quot;Safari&quot;)==-1) 

everything works great.  :(</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>341271</commentid>
    <comment_count>4</comment_count>
      <attachid>80407</attachid>
    <who name="Andy Estes">aestes</who>
    <bug_when>2011-01-27 20:07:56 -0800</bug_when>
    <thetext>Created attachment 80407
Reduction

Here is a reduction.

The site injects an unclosed object element towards the beginning of the page which causes the remainder of the document to be considered fallback content for the object. The legacy parser behavior was to close the object tag implicitly when its parent div was closed, but the HTML5 parser does not behave this way. Minefield shows the same behavior we do (if you trick it into following Safari&apos;s code path). In general, Firefox/Minefield don&apos;t suffer from this problem because the site injects an embed tag (which of course can&apos;t have fallback content).</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>341275</commentid>
    <comment_count>5</comment_count>
    <who name="Andy Estes">aestes</who>
    <bug_when>2011-01-27 20:15:02 -0800</bug_when>
    <thetext>and Chrome gets the embed tag too, of course :)

No love for Safari, unfortunately.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>341279</commentid>
    <comment_count>6</comment_count>
    <who name="Andy Estes">aestes</who>
    <bug_when>2011-01-27 20:21:29 -0800</bug_when>
    <thetext>milliyet.com.tr is actually quite popular in Turkey (ranked 7th by Alexa) and surprisingly popular globally (ranked #510). We should consider adding a evangelizing the site developer as well as perhaps adding a site-specific hack.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>351929</commentid>
    <comment_count>7</comment_count>
    <who name="Adam Barth">abarth</who>
    <bug_when>2011-02-16 01:08:28 -0800</bug_when>
    <thetext>The site appears to work fine now.</thetext>
  </long_desc>
      
          <attachment
              isobsolete="0"
              ispatch="0"
              isprivate="0"
          >
            <attachid>80407</attachid>
            <date>2011-01-27 20:07:56 -0800</date>
            <delta_ts>2011-01-27 20:07:56 -0800</delta_ts>
            <desc>Reduction</desc>
            <filename>reduction.html</filename>
            <type>text/html</type>
            <size>134</size>
            <attacher name="Andy Estes">aestes</attacher>
            
              <data encoding="base64">PGRpdj4KCTxvYmplY3QgaWQ9J2ZhaWwnIGRhdGE9J2h0dHA6Ly93d3cubGFmZmVydHkuY2Evd3At
Y29udGVudC91cGxvYWRzLzIwMDcvMDUvZmFpbC0yNC5qcGcnPgo8L2Rpdj4KPGgxPi4uLmJ1dCBh
Y3R1YWxseSBJIHBhc3MhPC9oMT4=
</data>

          </attachment>
      

    </bug>

</bugzilla>