<?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>201824</bug_id>
          
          <creation_ts>2019-09-16 09:53:13 -0700</creation_ts>
          <short_desc>[MSE][GStreamer] Allow MSE on builds with gst&gt;=1.14</short_desc>
          <delta_ts>2019-10-24 02:00:05 -0700</delta_ts>
          <reporter_accessible>1</reporter_accessible>
          <cclist_accessible>1</cclist_accessible>
          <classification_id>1</classification_id>
          <classification>Unclassified</classification>
          <product>WebKit</product>
          <component>WebKitGTK</component>
          <version>WebKit Nightly Build</version>
          <rep_platform>Unspecified</rep_platform>
          <op_sys>Unspecified</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>FIXED</resolution>
          
          <see_also>https://bugs.webkit.org/show_bug.cgi?id=201602</see_also>
          <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="Alicia Boya García">aboya</reporter>
          <assigned_to name="Alicia Boya García">aboya</assigned_to>
          <cc>annulen</cc>
    
    <cc>aperez</cc>
    
    <cc>berto</cc>
    
    <cc>bugs-noreply</cc>
    
    <cc>calvaris</cc>
    
    <cc>cgarcia</cc>
    
    <cc>clopez</cc>
    
    <cc>ews-watchlist</cc>
    
    <cc>gyuyoung.kim</cc>
    
    <cc>mcatanzaro</cc>
    
    <cc>pnormand</cc>
    
    <cc>rakuco</cc>
    
    <cc>ryuan.choi</cc>
    
    <cc>sergio</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>1571140</commentid>
    <comment_count>0</comment_count>
    <who name="Alicia Boya García">aboya</who>
    <bug_when>2019-09-16 09:53:13 -0700</bug_when>
    <thetext>The WebKitMediaSrc rework introduced usage of playbin3 for the
MSE player, which is not working as-is in 1.14, and therefore the
GStreamer requirement was bumped to 1.16.

playbin3 still has been found to be able to work in 1.14 with a simple
patch in gst-plugins-base, and being able to do this is important for
LTS distros, so we&apos;re lowering the requirement to 1.14 again.

This is the gst-plugins-base patch needed for playbin3 MSE to work in
WebKit: https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/issues/434</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571141</commentid>
    <comment_count>1</comment_count>
      <attachid>378866</attachid>
    <who name="Alicia Boya García">aboya</who>
    <bug_when>2019-09-16 09:54:44 -0700</bug_when>
    <thetext>Created attachment 378866
Patch</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571458</commentid>
    <comment_count>2</comment_count>
    <who name="Carlos Garcia Campos">cgarcia</who>
    <bug_when>2019-09-17 00:49:35 -0700</bug_when>
    <thetext>So, do we need to backport that patch to gst 1.14 branch? or should we notify distros to use that patch to make MSE work in webkit-2.26 with gst 1.14?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571479</commentid>
    <comment_count>3</comment_count>
    <who name="Alberto Garcia">berto</who>
    <bug_when>2019-09-17 01:32:12 -0700</bug_when>
    <thetext>(In reply to Carlos Garcia Campos from comment #2)
&gt; So, do we need to backport that patch to gst 1.14 branch? or should we
&gt; notify distros to use that patch to make MSE work in webkit-2.26 with gst
&gt; 1.14?

I&apos;m not sure if having gst patched in order to be able to use webkit 2.26 is an option...

Or did I understand it incorrectly?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571480</commentid>
    <comment_count>4</comment_count>
    <who name="Carlos Garcia Campos">cgarcia</who>
    <bug_when>2019-09-17 01:36:31 -0700</bug_when>
    <thetext>(In reply to Alberto Garcia from comment #3)
&gt; (In reply to Carlos Garcia Campos from comment #2)
&gt; &gt; So, do we need to backport that patch to gst 1.14 branch? or should we
&gt; &gt; notify distros to use that patch to make MSE work in webkit-2.26 with gst
&gt; &gt; 1.14?
&gt; 
&gt; I&apos;m not sure if having gst patched in order to be able to use webkit 2.26 is
&gt; an option...
&gt; 
&gt; Or did I understand it incorrectly?

Doesn&apos;t debian update gst in stable version? The patch is already in gst, we only need to backport it to 1.14 branch upstream</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571487</commentid>
    <comment_count>5</comment_count>
    <who name="Alberto Garcia">berto</who>
    <bug_when>2019-09-17 02:26:07 -0700</bug_when>
    <thetext>(In reply to Carlos Garcia Campos from comment #4)
&gt; (In reply to Alberto Garcia from comment #3)
&gt; Doesn&apos;t debian update gst in stable version? The patch is already in gst, we
&gt; only need to backport it to 1.14 branch upstream

Yes, we can ask the patch to be backported, but I&apos;m not sure if
requiring changes in another package for a new update to work is
something acceptable, e.g. someone could get webkitgtk 2.26.x via
security updates and not have a patched gst installed.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571494</commentid>
    <comment_count>6</comment_count>
    <who name="Adrian Perez">aperez</who>
    <bug_when>2019-09-17 03:07:27 -0700</bug_when>
    <thetext>(In reply to Alberto Garcia from comment #5)
&gt; (In reply to Carlos Garcia Campos from comment #4)
&gt; &gt; (In reply to Alberto Garcia from comment #3)
&gt; &gt; Doesn&apos;t debian update gst in stable version? The patch is already in gst, we
&gt; &gt; only need to backport it to 1.14 branch upstream
&gt; 
&gt; Yes, we can ask the patch to be backported, but I&apos;m not sure if
&gt; requiring changes in another package for a new update to work is
&gt; something acceptable, e.g. someone could get webkitgtk 2.26.x via
&gt; security updates and not have a patched gst installed.

If we know that the distribution is willing to update the GStreamer
package, cannot the update to the WebKitGTK package depend on a
version equal or newer of the GStreamer package than the one that
included the fix?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571525</commentid>
    <comment_count>7</comment_count>
      <attachid>378866</attachid>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-17 07:24:19 -0700</bug_when>
    <thetext>Comment on attachment 378866
Patch

This will break distros. You can&apos;t expect distros to know that a special patch has to be applied. You should require the three-digit version number that&apos;s needed for this to work. E.g. require 1.14.6. Sadly, GStreamer developers say 1.14 is dead and there will be no 1.14.6, so can&apos;t do that.

I guess stable distros will just have to disable MSE. :/ This is not only important for LTS distros. Fedora 29 is still supported for another two months, and it will have to disable MSE to update to 2.26.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571526</commentid>
    <comment_count>8</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-17 07:28:22 -0700</bug_when>
    <thetext>One option would be to revert the MSE rewrite for 2.26 and leave it only in trunk, that will avoid the problem except for LTS distros. LTS distros don&apos;t expect all features to be enabled anyway. And if we defer to 2.28, the timing will work out nicely such that 18.04 users will be able to upgrade to 20.04 around the same time that 18.04 loses support for MSE.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571531</commentid>
    <comment_count>9</comment_count>
    <who name="Alberto Garcia">berto</who>
    <bug_when>2019-09-17 07:32:57 -0700</bug_when>
    <thetext>(In reply to Michael Catanzaro from comment #8)
&gt; One option would be to revert the MSE rewrite for 2.26 and leave it
&gt; only in trunk, that will avoid the problem except for LTS
&gt; distros. LTS distros don&apos;t expect all features to be enabled anyway.

But isn&apos;t MSE a requirement for YouTube? Surely LTE users don&apos;t expect
that YouTube starts to break after they install the latest security
upgrade.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571534</commentid>
    <comment_count>10</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-17 07:47:04 -0700</bug_when>
    <thetext>(In reply to Alberto Garcia from comment #9)
&gt; But isn&apos;t MSE a requirement for YouTube? Surely LTE users don&apos;t expect
&gt; that YouTube starts to break after they install the latest security
&gt; upgrade.

Without H.264, yes. With H.264: not to my knowledge, so Ubuntu should be OK... hopefully.

At any rate, that&apos;s why I suggested the delay, so the new LTS will be out by the time the older LTS loses the MSE support.

Really, the best solution by far is to just do a GStreamer 1.14.6. That would make this a lot easier.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571535</commentid>
    <comment_count>11</comment_count>
    <who name="Alberto Garcia">berto</who>
    <bug_when>2019-09-17 08:01:05 -0700</bug_when>
    <thetext>(In reply to Michael Catanzaro from comment #10)
&gt; Really, the best solution by far is to just do a GStreamer 1.14.6. That
&gt; would make this a lot easier.

It this the only patch that needs to be backported?

https://gitlab.freedesktop.org/gstreamer/gst-plugins-base/commit/89d0e9cc92a86aa0227ee87406737b6d31670aea</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571540</commentid>
    <comment_count>12</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-17 08:20:23 -0700</bug_when>
    <thetext>That&apos;s what Alicia says. So if you patch that downstream, you can also patch WebKit downstream to remove the requirement for gst 1.16 and all should be good.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571558</commentid>
    <comment_count>13</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2019-09-17 09:20:36 -0700</bug_when>
    <thetext>(In reply to Alberto Garcia from comment #9)
&gt; (In reply to Michael Catanzaro from comment #8)
&gt; &gt; One option would be to revert the MSE rewrite for 2.26 and leave it
&gt; &gt; only in trunk, that will avoid the problem except for LTS
&gt; &gt; distros. LTS distros don&apos;t expect all features to be enabled anyway.
&gt; 
&gt; But isn&apos;t MSE a requirement for YouTube? 

It is.

Michael suggests to revert the MSE backend rewrite. Which means the old, slightly broken backend would be back in 2.26.

(In reply to Michael Catanzaro from comment #10)
&gt; (In reply to Alberto Garcia from comment #9)
&gt; &gt; But isn&apos;t MSE a requirement for YouTube? Surely LTE users don&apos;t expect
&gt; &gt; that YouTube starts to break after they install the latest security
&gt; &gt; upgrade.
&gt; 
&gt; Without H.264, yes. With H.264: not to my knowledge, so Ubuntu should be
&gt; OK... hopefully.
&gt; 

Youtube desktop now relies on MSE, this has nothing to do with the codec, AFAIK.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571562</commentid>
    <comment_count>14</comment_count>
    <who name="Konstantin Tokarev">annulen</who>
    <bug_when>2019-09-17 09:24:03 -0700</bug_when>
    <thetext>(In reply to Philippe Normand from comment #13)
&gt; Youtube desktop now relies on MSE, this has nothing to do with the codec,
&gt; AFAIK.

FWIW, it works fine here with H.264 and disabled MSE</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571834</commentid>
    <comment_count>15</comment_count>
    <who name="Carlos Garcia Campos">cgarcia</who>
    <bug_when>2019-09-18 00:41:04 -0700</bug_when>
    <thetext>We can still go with the approach of supporting both. It&apos;s a pain to maintain, but it&apos;s the only way qwe can fix it without disabling MSE and without requiring any additional action from distros.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571837</commentid>
    <comment_count>16</comment_count>
    <who name="Alberto Garcia">berto</who>
    <bug_when>2019-09-18 00:47:11 -0700</bug_when>
    <thetext>(In reply to Michael Catanzaro from comment #12)
&gt; That&apos;s what Alicia says. So if you patch that downstream, you can also patch
&gt; WebKit downstream to remove the requirement for gst 1.16 and all should be
&gt; good.

Note that e.g. Debian has gst 1.14.4 (not 1.14.5). Would it still be enough to backport just that patch?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571876</commentid>
    <comment_count>17</comment_count>
    <who name="Alicia Boya García">aboya</who>
    <bug_when>2019-09-18 03:44:06 -0700</bug_when>
    <thetext>(In reply to Carlos Garcia Campos from comment #15)
&gt; We can still go with the approach of supporting both. It&apos;s a pain to
&gt; maintain, but it&apos;s the only way qwe can fix it without disabling MSE and
&gt; without requiring any additional action from distros.

Please, no. I think it&apos;s unreasonable to keep two MSE backends in WebKit trunk just because some downstream distros refuse to update or patch broken GStreamer versions.

If you want to keep running old versions some patches will be needed here and there, and in the case of Debian stable they have already A LOT of GStreamer patches, I don&apos;t think we&apos;re such a special case.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1571877</commentid>
    <comment_count>18</comment_count>
    <who name="Alicia Boya García">aboya</who>
    <bug_when>2019-09-18 03:45:40 -0700</bug_when>
    <thetext>(In reply to Alberto Garcia from comment #16)
&gt; (In reply to Michael Catanzaro from comment #12)
&gt; &gt; That&apos;s what Alicia says. So if you patch that downstream, you can also patch
&gt; &gt; WebKit downstream to remove the requirement for gst 1.16 and all should be
&gt; &gt; good.
&gt; 
&gt; Note that e.g. Debian has gst 1.14.4 (not 1.14.5). Would it still be enough
&gt; to backport just that patch?

Unless there is an unforeseen issue, yes.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1572284</commentid>
    <comment_count>19</comment_count>
    <who name="Carlos Garcia Campos">cgarcia</who>
    <bug_when>2019-09-19 00:48:00 -0700</bug_when>
    <thetext>(In reply to Alicia Boya García from comment #17)
&gt; (In reply to Carlos Garcia Campos from comment #15)
&gt; &gt; We can still go with the approach of supporting both. It&apos;s a pain to
&gt; &gt; maintain, but it&apos;s the only way qwe can fix it without disabling MSE and
&gt; &gt; without requiring any additional action from distros.
&gt; 
&gt; Please, no. I think it&apos;s unreasonable to keep two MSE backends in WebKit
&gt; trunk just because some downstream distros refuse to update or patch broken
&gt; GStreamer versions.

It&apos;s a huge pain, but it&apos;s reasonable to keep backwards compatibility. We could start by applying the patch in the stable branch only, to ensure everybody can update to 2.26.1 without having to do anything else. Then we notify distros that a patch in gst 1.14 is needed, and once the patch si applied, we revert the patch in stable, land this one in trunk and backport it to stable.

&gt; If you want to keep running old versions some patches will be needed here
&gt; and there, and in the case of Debian stable they have already A LOT of
&gt; GStreamer patches, I don&apos;t think we&apos;re such a special case.

WebKitGTK provides backwards compatibility in API and ABI, so in theory, you shouldn&apos;t need any patch or anything else to use new versions of the lib in old apps/distros. I agree that sometimes that&apos;s not so easy and patches here and there are required, but that&apos;s the exception. We need to ensure that all distros that update to WebKitGTK 2.26 include the gst patch, not just Debian.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1572326</commentid>
    <comment_count>20</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-19 05:48:36 -0700</bug_when>
    <thetext>We need to tell distros to patch https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/merge_requests/427 anyway so we can do both at the same time. I&apos;m still seeing regular web process deadlocks in Ephy Tech Preview, despairing, and surprise! it&apos;s still this old bug because the fix was never released.

Would be much better to get GStreamer upstream to make new releases, of course.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1572354</commentid>
    <comment_count>21</comment_count>
    <who name="Michael Catanzaro">mcatanzaro</who>
    <bug_when>2019-09-19 07:46:30 -0700</bug_when>
    <thetext>(In reply to Carlos Garcia Campos from comment #19)
&gt; It&apos;s a huge pain, but it&apos;s reasonable to keep backwards compatibility. We
&gt; could start by applying the patch in the stable branch only, to ensure
&gt; everybody can update to 2.26.1 without having to do anything else. Then we
&gt; notify distros that a patch in gst 1.14 is needed, and once the patch si
&gt; applied, we revert the patch in stable, land this one in trunk and backport
&gt; it to stable.

We agreed to revert Alicia&apos;s work in 2.26 branch, so no changes will be needed for stable.

For trunk, I think the simplest solution is to just close this bug. We tell distros they should apply the necessary gstreamer patch and also give them a patch to drop the build requirement back to 1.14, which they can apply if they also use the gstreamer patch. But upstream we need to require 1.16, because GStreamer developers have EOLed 1.14 and will not accept the patch. And I don&apos;t think it&apos;s reasonable to ask Alicia to keep the old codepath around in trunk. This way we can save Alicia from doing a bunch of extra work.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>1583282</commentid>
    <comment_count>22</comment_count>
    <who name="Philippe Normand">pnormand</who>
    <bug_when>2019-10-24 02:00:05 -0700</bug_when>
    <thetext>This isn&apos;t an issue anymore because the old MSE backend is back in trunk and in 2.26.</thetext>
  </long_desc>
      
          <attachment
              isobsolete="0"
              ispatch="1"
              isprivate="0"
          >
            <attachid>378866</attachid>
            <date>2019-09-16 09:54:44 -0700</date>
            <delta_ts>2019-09-17 07:24:19 -0700</delta_ts>
            <desc>Patch</desc>
            <filename>bug-201824-20190916185443.patch</filename>
            <type>text/plain</type>
            <size>1929</size>
            <attacher name="Alicia Boya García">aboya</attacher>
            
              <data encoding="base64">U3VidmVyc2lvbiBSZXZpc2lvbjogMjQ5MjA1CmRpZmYgLS1naXQgYS9Tb3VyY2UvY21ha2UvR1N0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</data>
<flag name="review"
          id="394530"
          type_id="1"
          status="-"
          setter="mcatanzaro"
    />
          </attachment>
      

    </bug>

</bugzilla>