<?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>162860</bug_id>
          
          <creation_ts>2016-10-02 10:29:46 -0700</creation_ts>
          <short_desc>[GStreamer] Audio glitches</short_desc>
          <delta_ts>2018-02-21 07:48:16 -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>Media</component>
          <version>Other</version>
          <rep_platform>PC</rep_platform>
          <op_sys>Linux</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>WORKSFORME</resolution>
          
          <see_also>https://bugzilla.gnome.org/show_bug.cgi?id=772294</see_also>
    
    <see_also>https://bugs.webkit.org/show_bug.cgi?id=182829</see_also>
          <bug_file_loc></bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords>InRadar</keywords>
          <priority>P2</priority>
          <bug_severity>Normal</bug_severity>
          <target_milestone>---</target_milestone>
          
          
          <everconfirmed>1</everconfirmed>
          <reporter>jeremy9856</reporter>
          <assigned_to name="Nobody">webkit-unassigned</assigned_to>
          <cc>bugs-noreply</cc>
    
    <cc>calvaris</cc>
    
    <cc>cturner</cc>
    
    <cc>mcatanzaro</cc>
    
    <cc>pnormand</cc>
    
    <cc>webkit-bug-importer</cc>
    
    <cc>webkit</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>1235725</commentid>
    <comment_count>0</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-02 10:29:46 -0700</bug_when>
    <thetext>Hello,

I just tried Epiphany and I noticed audio glitches when I played radio paradise (webradio - https://www.radioparadise.com). The stream work well in Firefox. I&apos;m on Fedora 24, WebKitGTK 2.12.5. 
I already reported the bug against Epiphany but they asked me to report here (https://bugzilla.gnome.org/show_bug.cgi?id=772294)

If you need some particular infos to fix this, I will be pleased to provide them :)

Thanks !</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1236960</commentid>
    <comment_count>1</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2016-10-06 01:38:42 -0700</bug_when>
    <thetext>Seems like this website uses Flash?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1236972</commentid>
    <comment_count>2</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-06 02:24:05 -0700</bug_when>
    <thetext>No it don&apos;t. I don&apos;t have flash installed.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237040</commentid>
    <comment_count>3</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2016-10-06 06:42:33 -0700</bug_when>
    <thetext>I see JavaScript errors:

[Error] TypeError: undefined is not an object (evaluating &apos;soundManager.getSoundById(&apos;rp&apos;).readyState&apos;)
	reStart (rp2_player_1.php:100)
	Global Code (Script Element 1:1)

Generally, I would say the presence of JS errors indicates the website is broken unless the website developers tell us they think otherwise.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237041</commentid>
    <comment_count>4</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2016-10-06 06:43:10 -0700</bug_when>
    <thetext>(In reply to comment #1)
&gt; Seems like this website uses Flash?

Why do you say that?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237043</commentid>
    <comment_count>5</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2016-10-06 06:44:05 -0700</bug_when>
    <thetext>Wait, you say &quot;audio glitches.&quot; Were you able to get sound at all? I hear nothing. Maybe they&apos;re using an encumbered codec like MP3?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237075</commentid>
    <comment_count>6</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-06 08:58:45 -0700</bug_when>
    <thetext>(In reply to comment #5)
&gt; Wait, you say &quot;audio glitches.&quot; Were you able to get sound at all? I hear
&gt; nothing. Maybe they&apos;re using an encumbered codec like MP3?

I think they use MP3. I tried some other stream in MP3 and the glitches are here too. I tried some Ogg vorbis stream and there was no glitch !

I use RPMfusion repo for MP3 support. Can this be a problem with the repo ?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237085</commentid>
    <comment_count>7</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2016-10-06 09:10:51 -0700</bug_when>
    <thetext>This plays without glitches here on Debian Testing (after having removed the Flash plugin, which is used by this website if installed).

Do you also hear glitches with this command?

gst-play-1.0 http://stream-uk1.radioparadise.com/mp3-192?1475769631


In this case WebKit doesn&apos;t do anything fancy with the audio rendering, so if the gst-play-1.0 command above also shows glitches, you should report the bug in GNOME&apos;s Bugzilla against the GStreamer product. Sorry for the various bugzilla redirections :)</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237092</commentid>
    <comment_count>8</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-06 09:21:12 -0700</bug_when>
    <thetext>No glitch with gst-play-1.0. On the other hand there is glitches even with the stream URL in Epiphany.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237094</commentid>
    <comment_count>9</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2016-10-06 09:27:47 -0700</bug_when>
    <thetext>(In reply to comment #8)
&gt; No glitch with gst-play-1.0. On the other hand there is glitches even with
&gt; the stream URL in Epiphany.

This doesn&apos;t make much sense.
Can you add this option to gst-play-1.0 ? --audiosink=&quot;scaletempo ! audioconvert ! audioresample ! autoaudiosink&quot;

This should replicate the same audio rendering chain as used in WebKit.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237160</commentid>
    <comment_count>10</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-06 11:07:45 -0700</bug_when>
    <thetext>The &quot;?&quot; seem to pose problem 

Lecture en cours de /home/jeremy/?
ERROR Ressource introuvable. for file:///home/jeremy/%3F
ERROR debug information: gstfilesrc.c(530): gst_file_src_start (): /GstPlayBin:playbin/GstURIDecodeBin:uridecodebin0/GstFileSrc:source:
No such file &quot;/home/jeremy/?&quot;

So I used this command

gst-play-1.0 --audiosink=&quot;scaletempo ! audioconvert ! audioresample ! autoaudiosink&quot; http://stream-uk1.radioparadise.com/mp3-192?1475769631

No glitch too</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1237162</commentid>
    <comment_count>11</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-10-06 11:09:24 -0700</bug_when>
    <thetext>Oh I re-read your comment and that was a question :D</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1258839</commentid>
    <comment_count>12</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2016-12-10 22:56:16 -0800</bug_when>
    <thetext>The problem is still here :(</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1289359</commentid>
    <comment_count>13</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2017-03-19 01:58:00 -0700</bug_when>
    <thetext>Same problem on Fedora 25</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1325085</commentid>
    <comment_count>14</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2017-07-03 04:00:24 -0700</bug_when>
    <thetext>This played fine using MiniBrowser from a nightly build.

I had the same issue as Michael using Epiphany from f25 linked again 2.16.3.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1328003</commentid>
    <comment_count>15</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2017-07-12 04:47:38 -0700</bug_when>
    <thetext>I don&apos;t hear glitches, and I suspect the issue heard was fixed by r218703, which could non-deterministically cause horrible audio glitches in Icecast streams.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1329143</commentid>
    <comment_count>16</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2017-07-14 22:48:01 -0700</bug_when>
    <thetext>Thanks. In which webkitgtk version I will be able to test that ?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1329169</commentid>
    <comment_count>17</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2017-07-15 05:23:33 -0700</bug_when>
    <thetext>Jeremy, 2.16.6</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1333663</commentid>
    <comment_count>18</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2017-07-31 22:36:29 -0700</bug_when>
    <thetext>Fedora 25 just had the 2.16.6 version and the problem is still here :(</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1335195</commentid>
    <comment_count>19</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2017-08-04 02:53:40 -0700</bug_when>
    <thetext>Sorry Jeremy, it looks like r218703 wasn&apos;t merged into the 2.16.6 branch. I&apos;m going to propose it for merging and I&apos;ll update you again.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1336422</commentid>
    <comment_count>20</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2017-08-08 03:03:50 -0700</bug_when>
    <thetext>This is now proposed for merging into 2.16.7: https://trac.webkit.org/wiki/WebKitGTK/2.16.x</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1354492</commentid>
    <comment_count>21</comment_count>
    <who name="">jeremy9856</who>
    <bug_when>2017-09-29 04:32:49 -0700</bug_when>
    <thetext>I can confirm it&apos;s fixed !
Thanks !</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1354493</commentid>
    <comment_count>22</comment_count>
    <who name="Radar WebKit Bug Importer">webkit-bug-importer</who>
    <bug_when>2017-09-29 04:34:05 -0700</bug_when>
    <thetext>&lt;rdar://problem/34736614&gt;</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1366988</commentid>
    <comment_count>23</comment_count>
    <who name="">webkit</who>
    <bug_when>2017-11-01 12:29:30 -0700</bug_when>
    <thetext>With webkitgtk4-2.18.2 (and epiphany-3.26.1), this seems to be mostly fixed, but not entirely.

With WebKit 2.16.x, I experienced regular glitches (slight audio discontinuity every half-second or so, which continued the entire time the stream played.

With 2.18.2, audio streams start with the same half-second glitches, but they reliably go away after 4-5 seconds of playback.

On e.g. Firefox or various stream-capable media players, there are no glitches at any point.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1367059</commentid>
    <comment_count>24</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2017-11-01 14:39:04 -0700</bug_when>
    <thetext>Is there a specific website you can reproduce this problem on?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1367067</commentid>
    <comment_count>25</comment_count>
    <who name="">webkit</who>
    <bug_when>2017-11-01 14:49:31 -0700</bug_when>
    <thetext>Several radio streams at least.

E.g.:

http://bbcmedia.ic.llnwd.net/stream/bbcmedia_radio1_mf_p
http://media-ice.musicradio.com/ChillMP3
http://icy-e-ba-07-boh.sharp-stream.com:8000/kissnational.mp3
https://ais.absoluteradio.co.uk/absoluteradiomed.aac (needs to be contained in an &lt;audio&gt; element to be played instead of downloaded, at least in Epiphany/Web)
http://radio.virginradio.co.uk/stream</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1367149</commentid>
    <comment_count>26</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2017-11-01 18:38:13 -0700</bug_when>
    <thetext>OK, I can hear it sometimes, but not always. It seems to happen more often on the first stream, at least for me.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1367319</commentid>
    <comment_count>27</comment_count>
    <who name="">webkit</who>
    <bug_when>2017-11-02 06:26:34 -0700</bug_when>
    <thetext>Interesting, happens every time on each one for me, though it&apos;s difficult to hear if it&apos;s spoken word (e.g. advertising, which seems to be sometimes pre-rolled on some of those streams).

The Absolute Radio AAC stream (https://ais.absoluteradio.co.uk/absoluteradiomed.aac) is especially obvious - plays back with errors at 150-200% speed for the first 4 or so seconds.

The Chill stream should also be useful to reproduce since it uses an identical pre-roll at each stream start.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1385775</commentid>
    <comment_count>28</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2018-01-04 05:08:25 -0800</bug_when>
    <thetext>(In reply to webkit from comment #27)
&gt; Interesting, happens every time on each one for me, though it&apos;s difficult to
&gt; hear if it&apos;s spoken word (e.g. advertising, which seems to be sometimes
&gt; pre-rolled on some of those streams).
&gt; 
&gt; The Absolute Radio AAC stream
&gt; (https://ais.absoluteradio.co.uk/absoluteradiomed.aac) is especially obvious
&gt; - plays back with errors at 150-200% speed for the first 4 or so seconds.
&gt; 
&gt; The Chill stream should also be useful to reproduce since it uses an
&gt; identical pre-roll at each stream start.

I can reproduce the issue on Absolute Radio AAC but it&apos;s not a playback speed issue. Seems more like the initial ICY metadata isn&apos;t correctly parsed.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1385777</commentid>
    <comment_count>29</comment_count>
    <who name="">webkit</who>
    <bug_when>2018-01-04 05:24:10 -0800</bug_when>
    <thetext>I don&apos;t think it&apos;s a playback speed issue with the Absolute stream it sounds like it&apos;s just not successfully getting ~ every second packet for the first few seconds, and then playing the remaining (buffered) ones without any gaps for the missing packets. So it&apos;s normal pitch, but sounds &quot;double-speed&quot; (i.e. like a &quot;time stretch&quot; high-speed playback, rather than a &quot;chipmunk&quot; high speed playback).

Are you hearing something different from that? I just re-tested and it&apos;s still doing the same thing on that stream.

I suspect that the behaviour (no silence inserted for missing packets) is the same on the other streams, and it&apos; just there is a lower ratio of missing packets for the glitchy few seconds on the other streams. So something like:

Absolute: 13579 etc. (10-packet section played back in 5-packet time)
Others: 12456790 etc. (10-packet section played back in 8-packet time)

Instead of:

Absolute: 1.3.5.7.9. etc.
Others: 12.4567.90 etc.

Where &quot;.&quot; represents a one packet-duration silence.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1385785</commentid>
    <comment_count>30</comment_count>
      <attachid>330461</attachid>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2018-01-04 06:07:04 -0800</bug_when>
    <thetext>Created attachment 330461
patch

Can you test this patch please?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1386133</commentid>
    <comment_count>31</comment_count>
    <who name="">webkit</who>
    <bug_when>2018-01-05 08:37:50 -0800</bug_when>
    <thetext>I tried to build from the latest Fedora 27 SRPM for webkitgtk4 (2.18.4-1), and though the patch applies fine, there&apos;s an eventual build error unfortunately. I haven&apos;t tried building without the patch to see if it still fails.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1386139</commentid>
    <comment_count>32</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2018-01-05 09:02:34 -0800</bug_when>
    <thetext>(In reply to webkit from comment #31)
&gt; I tried to build from the latest Fedora 27 SRPM for webkitgtk4 (2.18.4-1),
&gt; and though the patch applies fine, there&apos;s an eventual build error
&gt; unfortunately. I haven&apos;t tried building without the patch to see if it still
&gt; fails.

Ah well, don&apos;t bother. It seems to be yet another bug in our httpsrc element because if I disable it, there is no glitch.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1399499</commentid>
    <comment_count>33</comment_count>
    <who name="Charlie Turner">cturner</who>
    <bug_when>2018-02-15 10:08:42 -0800</bug_when>
    <thetext>Bug 182829 contains a patch that helps this issue quite a bit. Instead of ~4s of bad audio, it&apos;s about 1/2 s. That&apos;s still not good enough though.

I&apos;m testing the absolute radio stream.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1399808</commentid>
    <comment_count>34</comment_count>
      <attachid>330461</attachid>
    <who name="Xabier Rodríguez Calvar">calvaris</who>
    <bug_when>2018-02-15 23:25:18 -0800</bug_when>
    <thetext>Comment on attachment 330461
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=330461&amp;action=review

It would be awesome if you could explain why this improves the situation as I see several things but I fail to understand why they work together.

&gt; Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:873
&gt; -    // Emit a GST_EVENT_CUSTOM_DOWNSTREAM_STICKY event to let GStreamer know about the HTTP headers sent and received.
&gt; +    // Emit a GST_EVENT_CUSTOM_DOWNSTREAM_STICKY event and message to let
&gt; +    // GStreamer know about the HTTP headers sent and received.
&gt;      GstStructure* httpHeaders = gst_structure_new_empty(&quot;http-headers&quot;);
&gt; -    gst_structure_set(httpHeaders, &quot;uri&quot;, G_TYPE_STRING, priv-&gt;originalURI.data(), nullptr);
&gt; +    gst_structure_set(httpHeaders, &quot;uri&quot;, G_TYPE_STRING, priv-&gt;originalURI.data(),
&gt; +                      &quot;http-status-code&quot;, G_TYPE_UINT, response.httpStatusCode(), nullptr);

You can keep this in a single line.

&gt; Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:888
&gt; +    gst_element_post_message(GST_ELEMENT_CAST(src), gst_message_new_element(GST_OBJECT_CAST(src),
&gt; +        gst_structure_copy(httpHeaders)));

ditto</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1399840</commentid>
    <comment_count>35</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2018-02-16 01:09:46 -0800</bug_when>
    <thetext>(In reply to Xabier Rodríguez Calvar from comment #34)
&gt; Comment on attachment 330461 [details]
&gt; patch
&gt; 
&gt; View in context:
&gt; https://bugs.webkit.org/attachment.cgi?id=330461&amp;action=review
&gt; 
&gt; It would be awesome if you could explain why this improves the situation as
&gt; I see several things but I fail to understand why they work together.
&gt; 
&gt; &gt; Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:873
&gt; &gt; -    // Emit a GST_EVENT_CUSTOM_DOWNSTREAM_STICKY event to let GStreamer know about the HTTP headers sent and received.
&gt; &gt; +    // Emit a GST_EVENT_CUSTOM_DOWNSTREAM_STICKY event and message to let
&gt; &gt; +    // GStreamer know about the HTTP headers sent and received.
&gt; &gt;      GstStructure* httpHeaders = gst_structure_new_empty(&quot;http-headers&quot;);
&gt; &gt; -    gst_structure_set(httpHeaders, &quot;uri&quot;, G_TYPE_STRING, priv-&gt;originalURI.data(), nullptr);
&gt; &gt; +    gst_structure_set(httpHeaders, &quot;uri&quot;, G_TYPE_STRING, priv-&gt;originalURI.data(),
&gt; &gt; +                      &quot;http-status-code&quot;, G_TYPE_UINT, response.httpStatusCode(), nullptr);
&gt; 
&gt; You can keep this in a single line.
&gt; 
&gt; &gt; Source/WebCore/platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:888
&gt; &gt; +    gst_element_post_message(GST_ELEMENT_CAST(src), gst_message_new_element(GST_OBJECT_CAST(src),
&gt; &gt; +        gst_structure_copy(httpHeaders)));
&gt; 
&gt; ditto

I wasn&apos;t expecting a review of this patch. Some of its changes were merged some weeks ago though, as part of other bugs.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1399849</commentid>
    <comment_count>36</comment_count>
    <who name="Xabier Rodríguez Calvar">calvaris</who>
    <bug_when>2018-02-16 02:09:31 -0800</bug_when>
    <thetext>(In reply to Philippe Normand from comment #35)
&gt; I wasn&apos;t expecting a review of this patch. Some of its changes were merged
&gt; some weeks ago though, as part of other bugs.

I assumed it was WIP and you probably wanted feedback.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1401106</commentid>
    <comment_count>37</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2018-02-21 07:48:16 -0800</bug_when>
    <thetext>I think this can be closed, the behavior in gst-play and webkit are now the same, at least for the absolute AAC stream which had glitches in the past for us.

Please re-open again otherwise :)</thetext>
  </long_desc>
      
          <attachment
              isobsolete="0"
              ispatch="1"
              isprivate="0"
          >
            <attachid>330461</attachid>
            <date>2018-01-04 06:07:04 -0800</date>
            <delta_ts>2018-01-04 06:07:04 -0800</delta_ts>
            <desc>patch</desc>
            <filename>0001-Patch.patch</filename>
            <type>text/plain</type>
            <size>7162</size>
            <attacher name="Philippe Normand">pnormand</attacher>
            
              <data encoding="base64">RnJvbSAzMGQxYzY1OTYyNzIxNWIwNDVmNGMzNmI0OWM5NjY3OTZiZWU5YjE2IE1vbiBTZXAgMTcg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</data>

          </attachment>
      

    </bug>

</bugzilla>