bugGnash - The GNU Flash player - Bugs: bug #31295, youtube: ~15-25" to properly...

 
 

bug #31295: youtube: ~15-25" to properly start

Submitter:  Gabriele Giacone <gg0>
Submitted:  Sun 10 Oct 2010 07:23:19 PM UTC
   
 
Category:  plugin Severity:  4 - Important
Release:  trunk Status:  Fixed
Privacy:  Public Assigned to:  strk
Open/Closed:  Closed
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

Thu 04 Nov 2010 08:15:16 PM UTC, comment #10: 

Pushed commit fe4faf3 to master..

Sandro Santilli <strk>
Group Member
Wed 03 Nov 2010 08:45:25 PM UTC, comment #9: 

I tested the patch on Debian/Lenny using the git trunk version, and the youtube video I tested, http://youtube.com/watch?v=QAUyaELfwBo , started a lot quicker with the patch in place.

Petter Reinholdtsen <pere>
Group Member
Wed 03 Nov 2010 07:52:36 PM UTC, comment #8: 

Tried and videos start immediately, great!
Well done, Sandro.

IMHO yt was the main improvement in 0.8.8, wouldn't this fix be enough to release 0.8.8.1 ?
Will users have to wait for 0.8.9 to have yt video played well?


Gabriele Giacone <gg0>
Group Member
Wed 03 Nov 2010 03:49:48 PM UTC, comment #7: 

The attached patch drops the select call from ExternalInterface::readBrowser, likewise it is done already in ExternalInterface::ExternalEventCheck.

Seems to fix the blockage here.

Please test and review, as I've no real tests for that spot
(readBrowser if used for host-to-standalone communications)


(file #21914)

Sandro Santilli <strk>
Group Member
Wed 03 Nov 2010 03:37:18 PM UTC, comment #6: 

This is where gdb finds gnash standalone blocked:

#0  0x00007f313ed24fb3 in select () at ../sysdeps/unix/syscall-template.S:82
#1  0x00007f3141ad2076 in gnash::ExternalInterface::readBrowser (fd=70)
    at /usr/src/gnash/gnash-head/libcore/ExternalInterface.cpp:490
#2  0x00007f3141ae275e in gnash::movie_root::callExternalJavascript (
    this=0x7fffde034760, name=<value optimized out>,
    fnargs=<value optimized out>)
    at /usr/src/gnash/gnash-head/libcore/movie_root.cpp:1837
#3  0x00007f3141c16579 in externalinterface_call (fn=...)
    at /usr/src/gnash/gnash-head/libcore/asobj/flash/external/ExternalInterface_as.cpp:291
#4  0x00007f3141b7d7c5 in gnash::builtin_function::call(gnash::fn_call const&)
    ()
   from /usr/src/gnash/build/head/gtk+kde4+sdl+fb+dump_ffmpeg+gst/libcore/.libs/libgnashcore-master.so

Sandro Santilli <strk>
Group Member
Tue 02 Nov 2010 09:47:25 PM UTC, comment #5: 

Blocking is due to standalone-plugin communication.
Dropping the -F argument fixes the blockage.
It should be carefully reviewed that both standalone and plugin use the communication channel in non-blocking mode.

Marking as important as it affects one of the last working video sites :P

Sandro Santilli <strk>
Group Member
Tue 02 Nov 2010 03:25:42 PM UTC, comment #4: 

I've noticed this too. It happens. Didn't research on why.
Would be interesting to know if anyone was able to reproduce with just the launcher  (no plugin involved)

Sandro Santilli <strk>
Group Member
Fri 29 Oct 2010 06:37:45 PM UTC, comment #3: 

Hello, Gnash team. I can confirm strange delays before playing random videos on youtube.com.

In attachment there is output of browser plugin when this happened. First video (http://www.youtube.com/watch?v=JTJhCpN0dhM) begins playing very good and without any problems, but the second (I click on near video "Bandit 1250F" wich URL is http://www.youtube.com/watch?v=qYFJbyUBVeU&feature=channel) begins playing only with delay 25-35 sec.

The behaviour is similar how described by Gabriele Giacone.

In addition, there is significant delay (~20 sec) between right click on video and appearance of the menu.

I don't use any proxy, my connection speed is 128KByte/sec. With Adobe Flash player such delays newer happen.

Gnash was compiled from source (today git snapshot)

user@lenovo-laptop:~$ gnash --version
Gnash master

Copyright (C) 2005, 2006, 2007, 2008, 2009, 2010 Free Software Foundation, Inc.
Gnash comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of Gnash under the terms of the GNU General
Public License.  For more information, see the file named COPYING.

Build options master
   Renderers: OpenGL, Cairo, AGG
   Hardware Acceleration:
   GUI: GTK
   Media handlers: gst ffmpeg
   Configured with: --host=i686-linux-gnu --build=i686-linux-gnu --prefix=/usr --disable-testsuite --without-swfdec-testsuite --without-ming --disable-dependency-tracking --disable-rpath --disable-cygnal --disable-python --enable-jemalloc --enable-hwaccel=none --enable-media=gst,ffmpeg --enable-gui=gtk --enable-renderer=all --with-plugins-install=system --enable-shared=yes
   CXXFLAGS: -g -O2          -W     -Wall     -Wcast-align     -Wcast-qual     -Wpointer-arith     -Wreturn-type     -Wnon-virtual-dtor     -Wunused      -fvisibility-inlines-hidden
   Version: master:20101029


utilitytrack <utilitytrack>
Thu 14 Oct 2010 11:46:05 PM UTC, comment #2: 


> I can't reproduce, altough I'm pretty sure I've seen this sometimes.


Sometimes it happens, sometimes not. I suggest you to randomly choose at least ~10 clips to reproduce it.

> Are you getting this behavior consistently ?


Most times it happens.

> Are you using a proxy ? Is media handler gst or ffmpeg ?


I've also reproduced it on slower machine behind a proxy: same phases with different duration (what happens after 10", happens at ~20").

> Does it happen again if you restart (MovieControl->Restart)


Yes, it happens again.
And surprise, when it restarts, we can heard two overlapped audio tracks, two original unsynced audio tracks.

Configure options:
                --host=$(DEB_HOST_GNU_TYPE) \
                --build=$(DEB_BUILD_GNU_TYPE) \
                --prefix=/usr \
                --sysconfdir=/etc \
                --mandir=\$${prefix}/share/man \
                --infodir=\$${prefix}/share/info \
                --disable-testsuite \
                --without-swfdec-testsuite \
                --without-ming \
                --disable-dependency-tracking \
                --disable-rpath \
                --enable-cygnal \
                --enable-python \
                --enable-jemalloc \
                --enable-hwaccel=none \
                --enable-media='gst,ffmpeg' \
                --enable-gui='gtk,kde4' \
                --enable-renderer=all \
                --enable-docbook \
                --enable-extensions=all \
                --with-npapi-plugindir=\$${prefix}/lib/gnash \
                --with-plugins-install=system \
                --enable-shared=yes \
                --without-included-ltdl

Gabriele Giacone <gg0>
Group Member
Tue 12 Oct 2010 05:05:18 PM UTC, comment #1: 

I can't reproduce, altough I'm pretty sure I've seen this sometimes.

Are you getting this behavior consistently ?
Are you using a proxy ? Is media handler gst or ffmpeg ?
Does it happen again if you restart (MovieControl->Restart) ?

Sandro Santilli <strk>
Group Member
Sun 10 Oct 2010 07:23:19 PM UTC, original submission:  


Too lazy to rewrite? ;)

<gg0> my gnash plugin behaves in this way. there are 3 phases:
<gg0> 1 - first 5-10 seconds: first frame frozen, no audio
<gg0> 2 - frame at 10'' frozen, progress bar moved to 10'' and frozen, audio starts (from 0'')
<gg0> 3 - after 10 secs, video and progress bar start to follow audio
<gg0> I'm talking about youtube only
[...]
<gg0> interesting, if I replace page url with video url (I mean replacing "watch?v=" with "v/") and I run it from command line (so no plugin), it works immediately fine

To reproduce, start from yt homepage and randomly choose any videos, after the first one, the following from the right pane.
Some videos might start immediately.
Reproduced with both 0.8.8 and trunk.

Attached debuglog. As filename says, phase2 (above) starts at 17:15:34.


Gabriele Giacone <gg0>
Group Member

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:
   
   
Comment:
   

Attached Files
file #21914:  readBrowserNoSelect.patch added by strk (736B - text/x-diff)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -email is unavailable- added by bwy (Updated the item)
  • -email is unavailable- added by pere (Posted a comment)
  • -email is unavailable- added by utilitytrack (Posted a comment)
  • -email is unavailable- added by strk (Posted a comment)
  • -email is unavailable- added by gg0 (Submitted the item)
  •  

    There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

    Only logged-in users can vote.

     

    Follow 9 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2010-12-10 bwy StatusReady For Test Fixed
        Open/ClosedOpen Closed
    2010-11-04 strk StatusConfirmed Ready For Test
        Assigned toNone strk
    2010-11-03 strk Attached File- Added readBrowserNoSelect.patch, #21914
    2010-11-02 strk Severity3 - Normal 4 - Important
    2010-11-02 strk StatusNone Confirmed
    2010-10-29 utilitytrack Attached File- Added gnash-git_29.10.2010_delay_before_playing.log, #21864
    2010-10-10 gg0 Attached File- Added gnash-dbg_17:15:34_0.10_frozen_full.log, #21653

    Back to the top

    Powered by Savane 3.13-4b48.
    Corresponding source code