This wiki was updated to MediaWiki 1.37. If you notice any issues, please report them to admin[at]

Archive:Conference video 2012

Jump to: navigation, search

The SUSE Video team is happy to announce, that we have everything ready to make live streams and recordings available. We cover all talks in the rooms named Kirk, McCoy, Spock, Scotty, Picard, Riker, and Data.

Live video streams (auto-recorded)

  • plays on smartphones and everywhere with a flash-player:

Video recordings

Questions? Please contact or <>


The video team took responsibility for planning, offering, and executing video recordings. Executive Summay: Success. We achieved more than we initially offered, 4 rooms (not 3) recordings and live streams. But we had to master several planned and unplanned issues, that gave us some learing curve.


First plans were made at FOSMDE 2012, when the dvswitch setup was evaluated (jw,cwh). Cost estimate: ca. 2000 EUR/room. Operators needed: 2 people/room. With dvwitch postproduction effort is substantially lower than with a conventional taping, editing, uploading cycle that we had done several years ago. This year we wanted to provide a recording service again, but based on a completely new setup.


One set of equipment was purchased in March, and inhouse tests were successful. An extensive 'test' was done at LinuxTag Berlin [1]: 25 recordings. Key challenges were: video signal quality, firewire reliability, confirmed volunteers. Further tests were done during XDC2012 using our setup also for live-streaming, and operated exclusivly by XDC-Staff after a short coaching session.


For the openSUSE Conference, we offered to cover 3 rooms during 4 days (Sat/Sun/Mon/Tue). Recordings available on next day. Availability of live-streams not asserted. Audio relying on existing systems.

While synchronizing our plans with the campus, their capable Audio/Video-Center team volunteered to cover two days in one of the three main lecture halls. This freed resources, and we offered recording in a fourth lecture room: Gentoo. Also with the help of the local AVC we could take care of recording interviews in the Press Room.

An extraordinary challenge was put on the team due to the relocation of the conference. Monday and Tuesday was held at a different venue, where the rooms were not well suited. Only one room was equipped with an audio system and a video projector. The video team coordinated with the Services Team to set up network, video projectors, power in 5 rooms, Speakers, Microphone, Amplifiers as required for the recording in three rooms (two rooms asserted, one optional).


Network setup was discussed and prepared ahead of time as a joint venture of the video team, our IT Services team and the Campus IT team. We agreed on bringing our own video server (16 CPU, 2TB disk) and a VLAN spanning all rooms. This allowed remote controlling of all machines, but had the acceptable disadvantage of fixed port assignment with no options to plug additional machines where needed.

The rooms were not available on the days before the conference, until Friday late afternoon. We

  • formed a 'video setup squad team' of well trained specialists.
  • planned the exact placement and cabling of all equipment in each room.
  • pre-packed the equipment 'one box per room'

A second round of setup took place on Sunday (afternoon until 22h). We moved all our equipment the moment talks ended in each room to the new venue and recreated the setup. By that time we had more volunteers than we had expected. There was actually a slight delay due to not having all helpers properly instructed. Ontop of recreating the setup, we also took take care of the complete audio systems for 3 rooms. Including microphones, mixers, loudspeakers.

In all rooms, we connected a VGA grabber between the speakers laptop and videoprojector. The grabber and a required mini-pc were placed at the speakers desk. For venue2, we also took care of all video projection facilities. For the large hall this required convincing local staff to return the projector that was there before; for all other rooms we brought mobile projectors from Prague and Nuremberg office and placed them on extra tables. One additional white screen was needed as well as proper cabling in all rooms.


Shift plans with volunteers were reduced to 2 rooms for Sat/Sun, as Gentoo and the main lecture hall where handled by Gentoo and AVC volunteers. Additionally we provided a camera operator for the Press room, where ca on dozend interviews where recorded.

Instead of the planned 3 rooms at venue 1 and 2 rooms at venue 2, we covered 4 and 3 rooms respectively. Also we had added an experimental second camera pointing back to the audience in each room. Although sometimes unreliable, the mixer operators appreciated the additional source.

For venue2, we rented additional audio equipment (wireless microphones, amplifiers, cables, loudspeakers) and used the audio mixer at the recording desk to control both the recording and the Speakers in the room. One avialable pair of Sennheiser Microphones was left behind in Nuremberg, but the AVC team helped us here again.

Machine setup partly automated. The capture machines for slides and front camera were created as appliances booting diskless from a usb-memorystick.

A complete set of spare equipment was available at all times (except for slides capture). This allowed us start setup in venue2 earlier and allowed us to identify and swap faulty hardware during production. 1 Audio Mixer failed, 2 used DV-Cameras, and FireWirePort of one mixer-laptop died. An additional camera was on hot standby and was used to compensate temporary issues. The total hardware budget was 5000 EUR. This required bringing low budget equipment, and 'renting' several items free of charge.

We had prepared two setups for live streaming (providing both flash and ogg), but used mainly bambuser (with flash), for its additional feature of automatic live recordings. Although the live streams had to be restarted a few times during each day, we had several hundred views on the automatic recordings and live feeds during the conference.

Post production efforts were

  • copy all recordings from the mixer machines to the video server in a timely manner, freeing enough disk space for the next day.
  • keep track of room names and talk titles, so that this info would be available during upload.
  • check all recordings for excessive or missing cuts. This required much more man-hours than expected, as the system introduced unexpected cuts, when cpu or network overload conditions appeared.
  • encode all recordings to webm format and push to and youtube via their web interface.

The list of over 100 recordings below demonstrates the overall success of the video production. Live streams were available at all times, with short interrupts of a few minutes only.


  • Finding enough volunteers was critical. We were not able to have 6 operators (1 mixer + 1 camera per room) at all times during the 4 days. Luckily Gentoo could take care of themselves after a short coaching session.
  • Audio recording was problematic. Quality issues surfaced that were not seen inhouse, or at Linuxtag where the room was serviced with a profesional sound engineer.
  • Linux FireWire drivers are easily confused when two signals are captured by a single machine. dvsource-firewire-next was created and systemd scripts were adapted.
  • Twinpact DV-Signals and FireWire cabling in general exposed reliability issues. Sporadic failures were efficiently handled by the team.
  • Having fixed IP-Addresses in the appliances complicated the network setup in the first venue.
  • The video server encounterd disk errors once. Later the internal 2TB disk space filled, forcing us to rely on external USB-Disks again.
  • The decision to move or not move the video server between venues was revised twice. In the end we moved it Sunday 21:00h on short notice to the local IT team.
  • The network link of the second venue was 100MB only, not the initially advertised 1Gbit.
  • Upload of the keynote speech was delayed 7 days, as this room was operated with a conventional setup by the AVC team. Their setup did not allow the short turnaround times we achived with dvswitch.

Future Outlook

  • We should invest some development effort in post production automation. That saves manpower during the event. E.g. meta-data (Speaker names, topics, time, room) should be associated with file names during the recording, not reverse engineered afterwards. A standardized event calender format is required. Ical? Gnome Planner?
  • rate limited file transfer and proper monitoring aids need to be scripted.
  • DI-Boxes needed to improve audio quality.
  • We now exceed the storage capacity of the openSUSE storage room. Video needs its own room.
  • Replace/repair broken hardware. Some components were used items from ebay; this needs better funding next time.

Video Team Members

Credits for their work at mixer, camera and in the 'Kancelar' go to: (reverse sorted by length of email)

  • Marco Fleckinger
  • jantomanek <at> jantom****.cz Jan Tomanek
  • Wolfgang Rosenauer
  • Theodoros Chatzimichos
  • Matthias Griessmeier
  • Dimitris Papapoulios
  • Petr Pulc (+ his team)
  • Christian Boltz
  • Gerhard Schlotter
  • Bernhard Wiedemann
  • Thomas Schmidt
  • Vit Pelcak
  • Robert Wawrig
  • Kilian Petsch
  • Harald Müller-Ney
  • Martin Caj
  • Christopher Hoffmann
  • Jürgen Weigert

Playlist by name and title

Saturday Kirk sat-09-10 Opening and Keynote - Agustin Benito Bethencourt: SME as target for GNU/Linux distributions Kirk sat-10-10:30 Petr Černý: Don't stand in the queue, let the queue stand in for you. Kirk sat-10:30-11 Bernhard Wiedemann: openQA - fully automated OS testing Kirk sat-11-12 Klaus Kämpf: Google, grep, and usbmon: Reverse-engineering a USB scanner protocol Kirk sat-13-14 Egbert Eich: Wayland Kirk sat-14-15 Klaas Freitag: ownCloud under the Hood Kirk sat-15-16 Michael Meeks: LibreOffice - what's new Kirk sat-16-17 Tor Lillqvist: Attitude adjustment needed: Experiences from LibreOffice porting to Android Kirk sat-17-18 Stefan Seyfried: Crashing the kernel for fun and profit McCoy sat-10-11 Thijs de Vries: Gamification - using game elements and tactics in a non-game context McCoy sat-11-12 Shane Coughlan: OpenRelief - Using Open Source Software and Open Hardware For Frontline Disaster Relief McCoy sat-12-12:30 Bas van Abel: If you can't open it, you don't own it McCoy sat-12:30-13:30 Round Table: Why 'open' matters? McCoy sat-14-15 Armijn Hemel : Defensive Publications: ensuring freedom of use in the Open Source Community McCoy sat-15-15:30 Ramon Roca: Extending the freedom to network infrastructures with the Bottom-up Broadband and the Commons McCoy sat-15:30-16 Lydia Pintscher: Wikidata - Wikimedia going structured data McCoy sat-16-16:30 Georg Greve: What you don't understand will still control you McCoy sat-16:30-17:30 Round Table: How do you DO 'open'? McCoy sat-17:30-18 Closing session (Saturday) Scotty sat-10-10:30 Fabian Groffen & Robin H. Johnson: Council and Trustees: Managing Gentoo Scotty sat-10:30-11 Sebastien Fabbro: gentoo@home Scotty sat-11-11:30 Tomáš Chvátal: Gentoo KDE: stable, fresh, and bleeding edge! Scotty sat-11:30-12 Alex Legler : Keeping Gentoo secure Scotty sat-12-12:30 Andrea Arteaga: Benchmarking suite for Gentoo Scotty sat-13:30-14:30 Jorge Manuel B.S. Vicetto: Using Catalyst to create a custom stage and ISO (part 1) Scotty sat-14:30-15 Jorge Manuel B.S. Vicetto: Catalyst features, limitations and feature requests Scotty sat-15-16 Fabian Groffen: Gentoo Prefix: The World beyond / Scotty sat-16-17 Robin H. Johnson & Gentoo Infra Team: State of Gentoo Infrastructure Scotty sat-17-18 Alex Legler : Working on Gentoo's PR Spock sat-10-10:30 Tomáš Chvátal: Pokud se to hýbe, zkompiluj to! Spock sat-10:30-11:30 Jakub Hrozek, Martin Košek: FreeIPA, SSSD Spock sat-11:30-12 Michal Čihař: Online translation using Weblate Spock sat-12-13 Petr Baudiš: Lowlevel Userspace Programming Spock sat-13-13:30 Elvir Kuric : Cgroups usage Spock sat-14:30-15 Thomas Renninger: Performance monitoring of modern X86 platforms using the CPUpower monitoring tool Spock sat-15-16 Michal Zima: WrapSix -- how to build it and how to run it Spock sat-16-17 Tomáš Vondra: Proč používat PostgreSQL Spock sat-17-18 Jiří Eischmann & Jaroslav Řezník: Jak přispívat do Fedory

Sunday Kirk sun-11-11:30 Karsten Keil: Use the openSUSE buildservice for embedded projects Kirk sun-11:30-12 James Tan: DevOps, Puppet & Chef Kirk sun-12-13 Sam Aaron: Zen and the art of Live Programming with Clojure Kirk sun-13-13:30 Baltasar Ortega: Teaching Free Software for everybody Kirk sun-14:30-15 Libor Pecháček: How software gets from the community to commercial enterprise Kirk sun-15-15:30 Torsten Rahn: Marble Virtual Globe - a Swiss Army Knife for Maps Kirk sun-15:30-16 Jos Poortvliet: openSUSE Around the World Kirk sun-16-17 (Miguel Angel Barajas:) Watson: Winning Jeopardy with SUSE, Power and Hadoop Kirk sun-17-18 Lightning Talks McCoy sun-11-11:30 Vojtěch Trefný: Launchpad McCoy sun-11:30-12 Widen-Damian Ivanov : Unity on openSUSE McCoy sun-12-13 Helmuth Peer, Markus Fleck, Matthias Praunegger, Stefan Reisinger: desktop4education and server4education - a free client and server solution for education McCoy sun-14-14:30 Pavel Šimerda: News from the world of Linux networking McCoy sun-14:30-15 Olaf Kirch, Marius Tomaschewski: Apocalypse Now - News from the Network Jungle McCoy sun-15-15:30 Networking round table (Sunday) McCoy sun-15:30-16:30 Radek Neužil: (Ne)bezpečnost internetové sítě McCoy sun-16:30-17 Petr Krčmář: Co je nového v Debian Wheezy McCoy sun-17:15-17:45 Tadeáš Pařík: Nasazení Ubuntu v praxi McCoy sun-17:45-18:15 Petr Krčmář, Tomáš Matějíček: Budoucnost Slaxu Scotty sun-11-11:30 Robin H. Johnson: Gentoo @ IsoHunt Scotty sun-11:30-12 David Heidelberger: 3D, games and everything about Graphic performance under Linux/Gentoo
W: not recorded Scotty sun-12-12:30 Christian Aistleitner: SHA1 and OpenPGP/GnuPG
W: not recorded Scotty sun-13:30-14:30 OpenPGP/GnuPG key signing party
W: not held Scotty sun-14:30-15:00 Jorge Manuel B.S. Vicetto: Using Catalyst to create a custom stage and ISO (part 2) Scotty sun-15-16 Theo Chatzimichos & Alec Warner: The Puppet Show Scotty sun-16-17 Jorge Manuel B.S. Vicetto & Hans de Graaff & Fabian Groffen: Gentoo testing, testing and automated testing Scotty sun-17:17-18:15 Tomáš Chvátal & Markos Chandras: Contributing to Gentoo: Getting Involved more Scotty sun-18:15-18:30 Closing Session (Sunday, Scotty) Spock sun-11-11:30 Jan Engelhardt: openSUSE on SPARC Spock sun-11:30-12 Dinar Valeev: openSUSE on Power Spock sun-13-13:30 Jozef Mlich: Announcement of the Openmobility summer contest results Spock sun-13:30-14 Alexander Graf, Dirk Müller, Andrew Wafaa: openSUSE ARMs Spock sun-14-14:30 Vladimír Čunát: Nix(OS) purely functional package management Spock sun-14:30-15:30 Peter Czanik: Why proper logging is important even in the bootstrapping phase of a software project? Spock sun-15:30-16 David Majda: Scanny — Ruby on Rails Security Scanner Spock sun-16-17 Davidlohr Bueso, Petr Uzel: fdisk: a 21st century disk partitioning tool Spock sun-17:15-17:45 Karsten Keil: POTS, FAX,ISDN in the era of voice over IP
W: Presentation_not_done Spock sun-17:45-18:15 Arul Selvan Rama Samy: Prospective Printing Solutions

Monday Picard mon-10-11 Prof. Joe Doupnik: A complete server to assist charities Picard mon-11-12 Coly Li: Kernel development/maintenance in Taobao Picard mon-12-13 Stephan Kulow: openSUSE release schedule NG Picard mon-14:30-15:30 openSUSE Project meeting Picard mon-15:30-16:30 (150views blip on broken clip) Marcus Meissner: openSUSE Maintenance - overview of technology and processes Picard mon-16:30-17:30 Robert Schweikert: Something is busted -- Food for thought Picard mon-17:30-18 Michal Hrušecký: Whats new in openSUSE Connect Riker mon-10-10:30 Daniel Lovasko: Packaging Tools Riker mon-11-11:30 Hannes Reinecke: megasas on steroids: qemu device-passthrough with VFIO Riker mon-11:30-12 Hannes Reinecke: dev_printk revisited: structured kernel message logging and the SCSI mess Riker mon-15:30-16 Gerhard Rigo: BYOD – Device Onboarding for LINUX based Clients (scheduled in Data) Riker mon-16:30-17:30 Kostas Koudaras: Ambassadors 2.0 Riker mon-17:30-18 Efstathios Iosifidis: Nonverbal communication at the booth Data mon-10-10:30 Dirk Müller: Alternative architectures: ARM Data mon-10:30-11 Dalibor Kurek: Linux on System z - consolidation makes cents Data mon-11-11:30 Thorsten Behrens: SUSE is leading the interoperability with Microsoft Word Data mon-11:30-12 Ismail Doenmez: About openSUSE Data mon-12-12:30 Vladimír Botka: KIWI and SUSE Linux Enterprise Point of Service Data mon-12:30-13 Michal Švec: SUSE Studio: Behind-the-scenes: agile system management Data mon-14:30-15 David Disseldorp, Arvin Schnell: Make it Snappery - Remote Snapshots with Samba Data mon-15-15:30 Jiaju Zhang: Booth - Geo Clustering for SUSE Linux Enterprise High Availability Extension Data mon-15:30-16 Izabel Valverde: The openSUSE Travel Support Program (scheduled in Riker) Data mon-16-18 UEFI Tutorial

Tuesday Picard tue-14-15 Kostas Koudaras: oSC13 The Spirit and the City Riker tue-10-10:30 Max Lin: Introduction to openSUSE KDE maintenance model Riker tue-10:30-11:30 Robert Schweikert: House cleaning is necessary - Package Maintainers Step up Riker tue-11:30-12:30 Frederic Crozat: systemd, dracut and openSUSE: where are we, what is missing, what do we plan for the future? Riker tue-14-15 Michal Vyskočil: Optimizing boot time - aka 2 second boot