Ducknapping4 min read

Following up on Seif’s blog post, I would like to provide you with an update on our efforts to recover the missing duck.
We’ve been in a very delicate situation lately. It turns out that a group of hacktivists kidnapped Duckie! They sent us this picture:

As you can imagine, until now we refrained from publicly releasing this information to avoid causing a panic.
In addition to requiring the port of everything to GStreamer 1.0, the ducknappers also sent us the following list of demands, in exchange for the safe return of the hostage:

• World timezones in the GNOME Shell clock
• Voting in GNOME Bugzilla (see bug 629161)
• The immediate termination of Tracker
• Abandon GNOME 3 and bring back the state-of-the-art GNOME 2 UI
• The sacking of the design team
• Make a new video editor, one that doesn’t suck
• Release E17

We’ve had some pretty intense negotiations with them. It seems they already killed Bonobo! These guys don’t mess around.
Therefore, we assembled a special task force to meet some of their demands during the GStreamer conference. René Stadler, Thibault Saunier and I spent countless hours of our spare time finishing the port of Pitivi to GTK3 and GStreamer 1.0. We also spent a lot of time discussing a major refactoring of the timeline UI code, with the ultimate goal of porting it to Clutter.
For the time being, PiTiVi looks like this:

As you can see, the migration is done. We’re here, we’re alive, and we friggin’ deliver. If you were in San Diego, you might have felt a few tremors (buildings shaking, a rumbling sound and all that); that was not an earthquake, it was the aftershocks of the massive merge of the pygi branch into PiTiVi “master”:

The diff of that pygi migration is pretty big: 73 files and more than five thousand lines of code touched throughout 122 commits.
We’ve got the UI pretty much nailed down (please see the list of bugs in GTK that affect us). Still, we always welcome your help. GStreamer 1.0 needs testing and fixing, especially GNonLin. If you’re looking for something to get involved in GStreamer development, this is a wonderful occasion. There’s also plenty of low-hanging fruit available to work on GES if you’re interested, so come talk to us!
Obviously, we can’t meet every single demand of the ducknappers by ourselves, so we need the help of the Open Source Community™ to do things like fixing GNonLin and chasing down the GNOME Design Team with pitchforks and so on.
Stay tuned for updates on the Clutter-based timeline and plans of a general UI overhaul. Meanwhile, here’s a very quick retrospective on the PiTiVi GSoC projects this year:

• Alongside other features, a manual layers interface was implemented by Paul Lange. He’s currently working on the required improvements in GES to unleash the potential of that new UI.
• Matas Brazdeikis added easy automated UI testing with Dogtail (I should definitely write a blog post dedicated to that) and implemented the title/text editor (finally!). Due to the bugs with GNonLin at the moment, this has not yet been fully tested, and I would like to improve the UI as it has rough edges, but the meat is there.
• Pēteris Krišjānis worked on providing a generic sound waveform analysis and display library for GStreamer. This is a joint effort between us, Buzztard and Jokosher. Pēteris wrote up on his experience. This is not yet integrated into PiTiVi and other apps, but we’re confident he will help us do that when the library is ready 🙂
• Volodymyr Rudyj worked on the design and implementation of “materials” in GES. Thibault is currently putting the finishing touches on that implementation work.

I’m very happy with how things are going. Although the road ahead of us is still quite long to reach feature parity with PiTiVi 0.15.x (in terms of stabilizing GStreamer, GES and GNonLin), we’ve already got a ton of measurable improvements to the quality of our codebase, our user experience and our community processes. I’ve made many friends in recent months, and at LinuxCon I even met a new partner to assist me personally: