FOSSwire

Back in January 2009, I wrote a post on opening video on the web. At the time, the Mozilla Foundation had just invested $100,000 in the Wikimedia Foundation to use Theora for videos on their sites.

In the last year and a half, the situation with video on the web has changed quite a lot. So what do those changes mean for free software and open source — where are we now with video on the web, are we any closer to open source and patent free web video?

HTML5, audio and video tags

Something really important has happened in the last 18 months. Adobe Flash, long the dominant platform for delivering web video, is in trouble. HTML5, a standard which will soon be supported by the major browsers, supports audio and video tags which are designed to allow video and audio playback to be supported natively by the browser — no plugins required.

This all sounds great in theory — the standardisation of this way of embedding content should make interoperability with web video even better, but unfortunately these tags alone don’t tell the whole story. audio and video themselves are just ways to embed audio and video content; you still need audio and video codecs that the browser will support. This is where things still remain very messy and riddled with legal issues.

Codec Soup: H.264, Theora, VP8 (WebM) ...

I discussed Theora and H.264 in my previous post on this subject and many of the issues that were present then still are there now.

H.264 seems to be one of the best codecs from a technical point of view and has wide support from commercial vendors. Most notably, there are a significant body of consumer devices that now have H.264 encoding and decoding hardware. While covered by patents and having a possibly uncertain future (it is free of charge only for personal use until the end of 2015), there is only a single body that holds all the patents and rights, the MPEGLA, making it more attractive for businesses who otherwise may fear being sued.

Theora has traditionally faced criticism for being technically inferior to competing codecs like H.264, but it is supposed to be an open standard which can be distributed without licensing fees and used without issues in completely open source systems, which should make it ideal. Unfortunately, it isn’t as simple as that. Owners of video compression patents feel threatened by competing, free technologies like Theora, and recent rumours suggest that Theora could be the target of legal threats because of some of these patents which it supposedly infringes upon.

There’s also a new kid on the block, WebM. With the backing of Google, WebM is a separate attempt from Theora to create a royalty free video compression standard which works great on the web. The codec used for the video is VP8, originally developed by On2 Technologies (who also developed VP3, which is now Theora). It is supposed to be technically superior to Theora, the implementation code is open source under the BSD licence and Google have also freed the patents covering the technology. So could WebM be the ideal open source solution if Theora is stuck in the legal doghouse? Unfortunately, it still remains unclear — the patents granted to various technology companies for video technology are so numerous and fundamental that it is likely VP8 could be legally challenged in the same way. To be clear: less direct threats have been made against WebM, and the backing from Google, Mozilla and Adobe could help defend against any future problems, but it still remains a concern. It is exactly that uncertainty that makes the open solution less attractive to businesses who fear legal action.

Browser Support

One of the big ideas of this new web video technology, as I explained earlier, is not needing any plugins to play web video. Of course, that will only work in practice if each browser supports both the HTML5 video tag and the right codec. There’s a useful table on codec support on Wikipedia, but I’ll summarise here.

Mozilla supports Theora in today’s Firefox and by version 4 will also support WebM. They have decided not to support H.264.

Apple’s Safari will support only H.264.

Internet Explorer, from version 9, will support H.264 natively, and WebM will work if the user has separately installed the WebM codec.

Google Chrome supports both Theora and H.264, will soon support WebM in the regular release.

In fact, recent builds of Firefox, Opera and Flock are now WebM-compatible.

My guess is that H.264 is likely to become the most popular format for HTML5 web video, particularly given the momentum it already has due to the Apple iPad, but if Mozilla stick to their word about not supporting it, website owners will probably be forced to encode in two formats – one for 'most’ browsers, H.264 and either WebM or Theora for Firefox users, along with other browsers such as Opera which are unlikely to support H.264.

Conclusion

So, where are we now with open video on the web?

Sadly, the answer is — still in a mess. There is, I think, still hope for truly open video on the web. There are significant players in this space, including Google, who seem to have the will to get behind open technologies and to challenge the dominant position of H.264 and the MPEGLA. The momentum behind WebM, in particular, and the impressively wide browser support already gained by Theora, means that the free software and patent-free approach still has influence in what will happen in the future.

I think one of the pivotal issues right now is what decision Mozilla will make about supporting H.264. It’s a tough decision, but if they stand their ground on only supporting Theora and WebM, there will either be a backlash against Firefox (a risk it seems they must take), or website owners will have to also support either of those formats in order to serve their Firefox users, meaning H.264 will not become the new de facto standard, or at least not on its own.

This is a really complicated and controversial issue, but it’s one that will be key to the future of the web. As challenging as this is, I believe that if those with influence stick to their principles, we will at least have choice about which technologies we want to use.

Our sister site YouMakeMedia has a tutorial for converting older FLV Flash video to H.264 if you do need to support those clients for the videos on your site!

VHS tape image is from http://www.sxc.hu/photo/992499. The Xiph Fish Logo and its theora.org variant are trademarks of Xiph.Org. The WebM logo is owned by The WebM Project.

Ohio LinuxFest is definitely one of my favorite Linux and open-source conferences. If you’ve never been to a Linux conference, now is a great time to begin. If you have, it’s also a great time to experience this conference.

The main reason you should go? It’s free. OLF is funded by sponsors, so you aren’t required to pay to get in (though you certainly can upgrade your conference pass if you choose). The only thing you have to do is register. Registration closes on September 19th, so be sure you’re not left out!

From Ohio LinuxFest: Back to the Future of Linux!:

bq.
Don’t miss your last chance to register for Ohio Linux Fest! A mere two weeks away on September 25-27, it’s coming faster than a Free Software freight train!
...
People come to OLF to meet with others that have helped create things like the backbone of the Internet, FireFox, Linux and the Apache web server. If you come to Ohio Linux Fest you will gain a unique perspective of this community. You will see the technology and some of the faces that have made possible such a wonderful change to the digital world.

There are some great speakers at the event, including Doug McIlroy, the inventor of the UNIX pipe, as a keynote speaker:

The featured keynote speaker this year is Doug McIlroy, a major contributor to the great grandfather of Linux and a direct contributor to code that powers many of the Unix products of today. Doug invented the concept of pipes inter-process data streams that no true Unix / Linux geek can live without. This concept made its way into less stable products as well from other proprietary vendors.

It’s definitely an event you don’t want to miss. Visit the Ubuntu booth and there’s a good chance you’ll see me there.

OLFU training classes will be offered the day before the event on the 25th. This is, literally, a bargain. This years training will be guided by LOPSA Where else can you go and get a days worth of training for only $350.00?
What other crazy things will you miss out on if you miss OLF? You’ll miss cool Tee Shirts, the yearly After-Party, the Free-Geek Columbus Linux Basics and more. Don’t be sorry that you didn’t register for the event.

There’s a rumor about that some people go to OLF for the after-party.

Regardless of your reason, make it a point to register. You’ll have a great time, trust me.

See you there!

This.

The File menu has been around since the dawn of user interfaces. But do you know what? It has overstayed its welcome. The File menu has been abused in far too many applications.

A File menu should deal with files. That’s it. Open a file, save a file, print a file, quit editing the file. Some, no.. most applications have been using the File menu to do everything from opening new windows to changing preferences. They don’t belong there: opening a new window should go in a Window menu, and changing preferences should be under Edit or Tools.

Here are some specific instances I’m talking about. I’m not trying to pick on any application or developer, these are just applications I have installed at the moment. But they should be fixed.

• Firefox, Epiphany, Opera, and many other browsers are abusers of the File menu. A browser generally does not have a whole lot to do with files at all. Two, maybe three items on their File menus might actually fit. But things like Open Tab, Location, and Send Link? No. It would make more sense to rename the File menu to “Browser” or “Web”.
• GNOME System applets like system-config-printer-applet that only have one option in their File menu: Close. Really?
• Seahorse has a File menu that will create, import, or export keys and passwords. It kind of makes sense, but a something like “Key” or “Keyring” sounds more in-context.
• GNOME Terminal has a File menu that doesn’t have a thing to do with files.
• Evolution has a huge File menu that has little do with files. “Mail” or even “Evolution” would be more appropriate.

• Synaptic's File menu could be “Package,” but considering there is already a Package menu other things would need to be reorganized.
• Olive's File menu has more to do with branches, though some items would fit as File.
• GConf Editor's File menu has five items for managing keys, two of which are disabled unless you know what you’re doing.
• D-Feet has a File menu that allows you to connect to different D-Bus sessions. I can’t quite put my finger on what this should really be labeled.
• Devhelp, a documentation browser, has a File menu that opens new windows and prints.
• Many, many more.

Again, if you’re a developer of something on this list, don’t feel too offended. Glade (a GTK+ designer program) sticks a File menu on new menubars by default, so it may just be an oversight. That doesn’t mean that it shouldn’t be fixed.

There are programs where it does make sense to have a File menu: Gedit, OpenOffice.org, GIMP, PiTiVi, and other applications that edit files as their main function.

Instead of dropping the File menu, it should be renamed to something more useful. Even though many applications abuse the File menu, there are plenty that do not. Banshee has a Media menu; Rhythmbox has Music. Most GNOME games have a Game menu. Baobab has Analyzer. Totem: Movie; Vinagre: Machine; Empathy: Chat; Gwibber: Gwibber.
Yes, even renaming your File menu to the name of your application might even be a good option.

So the next time you’re designing a GUI, keep in mind that your menus should make sense. Do that and we’ll have world peace. Well, maybe not, but it will shut me up.