Slashdot Mirror
Wikimedia Community Debates H.264 Support On Wikipedia Sites.
bigmammoth writes "Wikimedia has been a long time supporter of royalty free formats, but is now considering a shift in their position. From the RfC: 'To support the MP4 standard as a complement to the open formats now used on our sites, it has been proposed that videos be automatically transcoded and stored in both open and MP4 formats on our sites, as soon as they are uploaded or viewed by users. The unencumbered WebM and Ogg versions would remain our primary reference for platforms that support them. But the MP4 versions 'would enable many mobile and desktop users who cannot view these unencumbered video files to watch them in MP4 format.'
This has stirred a heated debate within the Wikimedia community as to whether the mp4 / h.264 format should be supported. Many Wikimedia regulars have weighed in, resulting in currently an even split between adding the H.264 support or not. The request for comment is open to all users of Wikimedia, including the broader community of readers. What do you think about supporting H.264 on Wikimedia sites?"
Wikimedia should stand their ground to provide a good reason for device manufacturers to add support for open video formats.
I am becoming gerund, destroyer of verbs.
And for these people the cost of paying for a H.264 encoder license is trivial compared to royalties they have to pay for images, video, and music.
And what about the developing world that is slowly coming online via shared community hubs? Won't they have the right to publish content too without paying exuberant rents compared to their income? The cost is trivial for everyone. I am sorry but open formats are the only way forward for a level playing field. All we are seeing with this WWF/H.264 debacle is a small amount of vested interests trying to justify extracting rents from the world population, when non are really required.
That these closed proprietary formats/DRM are clawing their way back into our "open" standards, services like Wikipedia and browsers is a testament to how committees, foundations (and once democratic institutions serving the public interest) can be infiltrated by vested interest and their purpose corrupted slowly from the inside out. It is a slippery slope, read todays news to see how absolutely low you can slide.
It's an encyclopedia
Exactly, it should just support formats that users have and not play politics.
Wrong. I think it should "play politics" in this case. Wikipedia is one of the very few sites which, because of its popularity, uniqueness, and non-commercial nature, has some leverage over browser vendors, and has more freedom than others to make use of it.
Almost everywhere else on the web it's the other way round: The browser vendors can force the site owners into compliance. If you have a smallish website and you want to provide video content on it, you often have no choice than to use an encoding like H.264 that all browsers support -- thereby furthering the agenda of consortiums like MPEG LA to steer the market towards a universal adoption of a patent-encumbered "hands off" format, and also lessening the incentive for browser vendors to support open royalty-free encoding formats. But if you run the like 4th most popular site in the world, the only one of its kind, AND you're not commercially bound to maximize your number of visitors no matter what, then you have some power to drive the web (and the whole industry) in the direction of truly open, royalty-free, "free to tinker with" video encoding formats, which would help lower costs and market entry barriers for new companies and individuals. Wikipedia shouldn't throw this leverage away.
In every meaningful sense, MP4 is the most 'open' useful video CODEC every made available.
Only for a very narrowly defined sense of the word "meaningful", and a particular meaning of "open".
The H.264 video standard is patent-encumbered. In some countries, the government doesn't grant or enforce patents on software, so this may not matter to you. But the USA is one country with software patents, and I live in the USA, so it matters to me.
And it matters to anyone in the USA who would like to use Wikimedia, even if they don't understand the issues yet.
The world's BEST video encoder, x264, is open-source and free.
But still patent-encumbered. Thus, the nice folks who wrote x264 and gave away their work do not charge you to use it; but in the USA, if you use it, you must obey the demands of the MPEG-LA and pay the royalties they require.
Thus, x264 is free and open-source software for a non-free and non-open standard.
Meanwhile, the dreadful CODEC that Google bought was created illegally by using close-source development as a method of hiding the fact that it ripped off (badly) patented MPEG standards.
Are you a lawyer? Is this legal advice?
I'm not a lawyer, and this is not legal advice, but my layman's opinion is that you are just completely wrong on all points here. VP8, as I understand it, was created by people who studied H.264 and made sure that VP8 did not infringe on any patents. Many things VP8 does are similar to things H.264 does, but that's not illegal.
Here's an example that may shed some light, from the Linux kernel. The Linux kernel had a driver for "VFAT" file systems, which can have "long filenames" (also known as "non-broken filenames" or just "filenames"; compare with older FAT file systems that can only have "short filenames" of 8 letters followed by a 3-letter extension). It turns out that Microsoft had a patent that covers VFAT, so for a while VFAT support was ripped out of the Linux kernel. But someone studied the patent and saw that it was a patent on a particular method of storing two filenames for each file: a long filename and a short filename. Thus, the Linux VFAT driver was re-written, such that when writing a file, it wrote a nice legal "long filename" and put garbage bytes in the "short filename" field. Since the garbage bytes were chosen to not be a valid filename, the Linux VFAT driver was not infringing on a patent that covers writing two filenames.
The above hack figured out what was patented, figured out a workaround, and implemented the workaround. The "long filenames" are written exactly as described in the patent, but the patent was not on "a method of storing long filenames" it was on a method of storing two filenames for each file.
Returning to VP8, my understanding is they did this sort of thing for video coding. They avoided patents but found similar things that would work.
Did they succeed? Well, there was a delay of many months after Google bought On2 and before Google released their free version of VP8, and I believe during that period Google had their lawyers reviewing all the patent issues. They thought they succeeded. And then, MPEG-LA announced that they were forming a patent pool on VP8, but over a year later there were no patents in that patent pool. That is the best possible evidence that On2 did succeed: even with the source code to study, no patent owner was able to find infringing code.
On2 also claimed that VP8 was "better" than H.264, but we know that is definitely not true. But it's the next best thing, and it's way better than older standards like H.263.
After Google released the source, and the truth became obvious, Google simply used its billions to pay off the various IP owners whose patents the code infringed on.
Nope. Factually untrue. After the patent owners failed to find any patents that infringed, Google was able to strike a deal with MPEG-LA where Google admitted no wrongdoing, gave MPEG-LA some money, and MP
The only reason the fight was given up was because Google gave up. Apparently they had no faith in WebM, despite having both formats on Youtube, so they renegged on their "promise" to drop support for H264 in Chrome.
If you ask me, it's got EVERYTHING to do with politics. Every time they do something like that, it causes trouble for competitors. Mozilla had to implement H264 when they did that. It had to also support VP9 when they introduced that. And they're having to scramble to support Media-streaming Extensions now as well. Opera just gave up entirely and became a Chrome clone. All because Google desired it, not because they had to do so.
You see, Google's about making money. It's not about winning or losing in a codec war, it's about who controls the future of web video. And if you can keep everyone scrambling to catch up, you can dictate what happens next. They're doing this in many areas of the web; SPDY (which basically became HTTP2), Pepper, NaCl, WebP, etc. They won't win them all, but they're all power plays to make sure Google's the one ahead of the curve while everyone else plays catch-up.
If someone can take a principled stance against this, they should. Right now the only entities with a spine are the ones who aren't for-profit, and none of them can stand up to it, and end users should really be supporting them. Do we really want Google and the MPEG-LA to dominate like that, knowing what happens every time companies dominate something like that?
Yes - it's called "most people don't care what their computer or mobile phone runs - they just want things to work when they click/tap them."
When a kid in middle school, working on a Windows XP computer that the district can't afford to replace, and can barely afford to (under)pay an IT staff to maintain, accesses Wikipedia to do research for a report, and can't view the video because IE doesn't support Ogg, that kid gives up on Wikipedia.
When a grandma, working on her iPhone, tries to watch a short video about a topic she's interested in, and can't, she gives up on Wikipedia.
You're absolutely right - it is wrong. And Wikimedia stubbornly sticking to "free only!" doesn't fix it. Even a giant "YOU NEED AN OPEN PLATFORM TO VIEW THIS - CLICK HERE TO FIND OUT MORE!" when you try to view a video will only scare people away, not get them to move to open platforms.
Another non-functioning site was "uncertainty.microsoft.com."
The purpose of that site was not known.
Wow. Troll? No - truth.
Go ahead and play politics - but if your mission is "...to empower a global volunteer community to collect and develop the world's knowledge and to make it available to everyone for free, for any purpose." then to me "make it available to everyone" is primary take away.
"Make available to everyone" means *MAKE AVAILABLE*. They're not the Free Software Foundation. They're not GNU, they're not even Creative Commons. Their mission is to make the information available to as many people as possible. To me, this means that supporting closed FORMATS for open INFORMATION gets to the maximum number of people.
They also specifically call out that they are about "free content" - notably SEPARATING it from "open content". The part of the content they care about is the freedom of the CONTENT itself. Public Domain, CC-licensed, etc. The mission of Wikimedia doesn't mention supporting OPEN content as a priority. And that is as it should be!
Another non-functioning site was "uncertainty.microsoft.com."
The purpose of that site was not known.