I didn't even know this existed, and now it's gone! I've been looking for something like this for awhile and would gladly pay for it. The few similar things out there are all clunky, slow, and just not up to the task. I'd throw in $50 or so just for this if it worked :)
Honestly, we need to start changing the laws to prevent patent abuse. If your patents become part of a must-implement standard (with your permission, of course, mpeg-la members have obviously given theirs), then your patents become null for anyone implementing that standards, ie, you are forced to license them for free.
Using patents like this is pretty much the patent equivalent of entrapment.
Also, why isn't Vysor just using the HMTL5 facilities to use hardware h264 enc/decs?
Someone could link the sea of posts anticipating this exact scenario when the standard was discussed, but it's a waste of time. After all, browser makers surely have the users' best interest in mind, right? Apple, Google and Microsoft cannot all be wrong (and evil) at the same time, right? Right?
AV codecs are a patent minefield, and sitting on your hands waiting for a guaranteed free standard is a pretty shitty option for users. Even Firefox realized that.
Based on his comments, it seems that the MediaSource APIs are not sufficient for low latency, and most Android devices don't have HW VP[89] encoder support.
there is actually no way to get a h264 stream(!) decoded with onboard html5.
using pepper plugins would limit your users to chrome.
another alternative might be flash but i dont blame anyone for not taking this alternative serious.
Standards-setting bodies typically address this today through the use of standard-essential-patent policies. IEEE, for example, requires assurance from such patent holders that they will either not enforce the patent against those implementing the standard, will grant a license free of charge, or will grant a license "under reasonable rates."
Not sure why improving this state of affairs would require a change in underlying patent laws. We just need standards bodies to adopt more general patent policies--which could be accomplished by preferring to use standards from orgs with better patent policies.
> Not sure why improving this state of affairs would require a change in underlying patent laws.
The problem with this is that there will always be someone who will happily say "Sure, I'll call your patent a 'standard' if you give me a little bit of money to 'certify' it."
Actually the problem these days is that standard-essential patents are considered less valuable because they tend to require FRAND licensing. This is the problem companies like Motorola and Samsung ran into during the smartphone patent wars. They could not ask for injunctions or high royalties for many of their patents as they were standards-essential, and they lost a lot of leverage.
Secondly, standards setting bodies are usually a conglomerate of members from competing companies and hence a huge political bureaucratic mess, so it's not as easy to get something into the standard just for some money.
Maybe. But then there's the little matter of adoption. The fact that some body or other called a patent technology a "standard" is meaningless, and not especially problematic. The problem comes when this standard actually comes into widespread use.
Keep an eye on Alliance of Open Media ( http://aomedia.org ). Basically Amazon, Cisco, Google, Intel, Microsoft, Mozilla and Netflix got fed up with the licensing model of H.265/HEVC which is even worse than that of H.264 (dual patent pools which different licensing terms where you have to pay both).
Intent is for it to be royalty free, open and supported in major browsers and devices.
If we want people to pay us for software, we have to contribute toward a system where value is added and money is extracted as goods and/or services of increasing value are developed and transition between entities. This radical concept is known as "industry."
The alternative is to be mindless coder drones that produce "content" for people who don't value it but profit from it. Like musicians producing music for music companies but somehow not getting paid for it, you too can be a doofus with a Google+ account adding value to the Android or Apple ecosystem with your free app. Get "wrote an app" on your resume, make a snappy logo, then flounce out with a sad Google+ post to extract maximum brogrammer value!
I increasingly see these issues as self-serving ways for some dude to get out of the hassles of maintaining some bullshit free app with 8 users.
Somewhere there's a real company who'll cut the $50k check. Somewhere there's a hungrier, smarter developer who'll work around the issue. Somewhere there's a dirtier, angrier developer who'll fight for what's right against inane licensing terms. This developer is none of the above, nor is he contributing to furthering the industry. Begone.
He included an unlicensed H.264 decoder in his app. He could have avoided this by either using Chrome's built-in APIs[1], utilizing Cisco's free H.264 implementation[2], or simply paying the license fees himself. Google and Cisco are already paying the fees to make developers lives easier.
As much as I hate patent trolls, this isn't really patent trolling.
To make matters worse, I informed Koushik Dutta about this possible issue back in 2015[3] and he closed my GitHub issue and ignored it. I'm not surprised at the outcome.
What difference does it make? It even seems backwards in some sense --- that writing code based on a freely available spec means you have to pay a license fee, yet someone else can give away their implementation freely? IANAL, so perhaps what we think of as the crazy and contorted bureaucratic mess that is the patent system would make sense in some strange and twisted way to those who designed it.
Not really, it's just compliance with the license that a piece of software is offered with. Things are no less silly in the world of open source. You are bound to honor the license of the open source software that you use. For instance, if your app links to a GPL library, you are bound to offer your code under the GPL too. However if it links to a BSD-licensed library that does the exact same thing, you don't have to.
Of course, their IP doesn't have value because it's particularly clever, it has value because it's in a specification that people have to implement in order to inter-operate with everything else that's based on the same specification. Their patent could be on a trivial tweak to something that's unpatented with no benefit other than making them money in licensing fees, and as long as they can get their version into the standard...
As a consumer I can watch videos for free on multiple services. I can create them for free on multiple devices with free software packages. And, as someone who battled multiple inane video and audio formats since the late 80s, now's the only time in history where it's fast, easy, cheap and trivial to distribute media that just works on any device.
I'm sure somebody will try to claim "chilling effect" based on one of those "free" not being Free. But things are better than they've ever been from the point of view that matters most: getting stuff done, getting stuff published, and getting yourself heard without forcing your audience to install stuff first.
Some tiny royalty payment pre-paid by the platform doesn't destroy the experience. But forcing my mom to install some sketchy open source software with the wrong volume control UI then telling her to drag and drop my Happy Mother's Day video onto a traffic cone sure does.
You don't have to pay for writing code based on the spec, you have to pay for users using it. Just as Cisco does for users of their free binaries. (They probably don't pay all that much extra since they already pay tons of fees for their own products, and it has to be some kind of flat-rate deal since they can't track precise user numbers)
If you take their code and compile it yourself, you are on the hook for fees of users, just as if you wrote the code yourself.
I don't like it, but it is a logical option that if someone pays it is the one responsible for making the "product" (=binary including the codec)
He closed it five days ago, which seems especially interesting.
fwiw, speaking as the maintainer of a reasonably popular OSS project, it may be worthwhile to include actionable next steps in your issue. Including the comment you just wrote here might have made it easier for him to take action on this back in 2015.
Yeah, I took a break from the project since last August. Hadn't read any of the issues. When I got back, I bulk closed everything that did not look relevant to the functionality of Vysor (didn't even read the issue). Was just trying to get through the bug backlog effeciently. Three days later, I received a letter from the MPEG-LA.
Btw, Chrome's built in APIs (Pepper VideoDecoder) were not available until around 10 months after my first implementation of Vysor.
Your issue was opened in November. I closed it on May 11th, and received the letter on May 13th. Timing here is definitely suspect.
Couldn't he have just added a disclaimer that the app may contain patented technology, but software patents are not enforceable everywhere? I doubt FFmpeg has paid a single cent of license fees to anyone, and yet it's survived and still available for anyone to download and use:
Downstream projects of FFmpeg such as VLC certainly have been affected [1]. The interesting thing is that most of the problems have been with audio formats - this is the first I've seen for an open source project using H.264.
The Wikipedia article is a bit misleading because although it cites the EU as an example, most software patents are allowed in the EU, including many H.264 patents. I would encourage you to fight software patents not by infringing, but by using royalty-free alternatives.
You got it! Thanks for the tip. I just found my way over there and the whole thing is 100x better than I expected to find. Not an ad in sight and not even a message asking for donations. I'm thirsty for more now!
Just wanted to provide some bits about the MPEG-LA AVC patent portfolio licensing (which includes H.264 enc/dec). I went through this process for work and it's pretty painless. The licensors are very responsive will help you find a solution. They're not trying to shake you down.
(I'm not familiar with Vysor but whatever you do, stay away from libvlc if you're looking for a decoder/encoder. That stuff is not kosher in most countries. ffmpeg is a gray area too.)
Some highlights from a few years ago (may differ slightly now but unlikely):
It's my understanding that a "shake down" is an act built on /unreasonable/ demands. This is hardly unreasonable though do recognize that's subjective.
54 comments
[ 2.8 ms ] story [ 111 ms ] threadUsing patents like this is pretty much the patent equivalent of entrapment.
Also, why isn't Vysor just using the HMTL5 facilities to use hardware h264 enc/decs?
http://wiki.webmproject.org/hardware/socs
Or the built-in Chrome API for that, PPB_VideoEncoder: https://developer.chrome.com/native-client/pepper_dev/c/stru...
Well actually, it might be okay on desktop.
But yes, I've switched.
https://standards.ieee.org/develop/policies/bylaws/approved-...
Not sure why improving this state of affairs would require a change in underlying patent laws. We just need standards bodies to adopt more general patent policies--which could be accomplished by preferring to use standards from orgs with better patent policies.
The problem with this is that there will always be someone who will happily say "Sure, I'll call your patent a 'standard' if you give me a little bit of money to 'certify' it."
Secondly, standards setting bodies are usually a conglomerate of members from competing companies and hence a huge political bureaucratic mess, so it's not as easy to get something into the standard just for some money.
Intent is for it to be royalty free, open and supported in major browsers and devices.
The alternative is to be mindless coder drones that produce "content" for people who don't value it but profit from it. Like musicians producing music for music companies but somehow not getting paid for it, you too can be a doofus with a Google+ account adding value to the Android or Apple ecosystem with your free app. Get "wrote an app" on your resume, make a snappy logo, then flounce out with a sad Google+ post to extract maximum brogrammer value!
I increasingly see these issues as self-serving ways for some dude to get out of the hassles of maintaining some bullshit free app with 8 users.
Somewhere there's a real company who'll cut the $50k check. Somewhere there's a hungrier, smarter developer who'll work around the issue. Somewhere there's a dirtier, angrier developer who'll fight for what's right against inane licensing terms. This developer is none of the above, nor is he contributing to furthering the industry. Begone.
As much as I hate patent trolls, this isn't really patent trolling.
To make matters worse, I informed Koushik Dutta about this possible issue back in 2015[3] and he closed my GitHub issue and ignored it. I'm not surprised at the outcome.
[1]: https://developer.chrome.com/native-client/pepper_dev/c/stru...
[2]: http://www.openh264.org/
[3]: https://github.com/koush/vysor.io/issues/84
What difference does it make? It even seems backwards in some sense --- that writing code based on a freely available spec means you have to pay a license fee, yet someone else can give away their implementation freely? IANAL, so perhaps what we think of as the crazy and contorted bureaucratic mess that is the patent system would make sense in some strange and twisted way to those who designed it.
That's so fucking stupid.
Why should anyone be able to nullify the entire value of another company just by implementing the spec they published?
Therefore, the patent system.
I'm sure somebody will try to claim "chilling effect" based on one of those "free" not being Free. But things are better than they've ever been from the point of view that matters most: getting stuff done, getting stuff published, and getting yourself heard without forcing your audience to install stuff first.
Some tiny royalty payment pre-paid by the platform doesn't destroy the experience. But forcing my mom to install some sketchy open source software with the wrong volume control UI then telling her to drag and drop my Happy Mother's Day video onto a traffic cone sure does.
If you take their code and compile it yourself, you are on the hook for fees of users, just as if you wrote the code yourself.
I don't like it, but it is a logical option that if someone pays it is the one responsible for making the "product" (=binary including the codec)
I'm sorry, how is using the source which needs compilation different from using the binary provided?
He closed it five days ago, which seems especially interesting.
fwiw, speaking as the maintainer of a reasonably popular OSS project, it may be worthwhile to include actionable next steps in your issue. Including the comment you just wrote here might have made it easier for him to take action on this back in 2015.
Btw, Chrome's built in APIs (Pepper VideoDecoder) were not available until around 10 months after my first implementation of Vysor.
Your issue was opened in November. I closed it on May 11th, and received the letter on May 13th. Timing here is definitely suspect.
https://en.wikipedia.org/wiki/FFmpeg#Legal_aspects
I believe the way to fight software patents is civil disobedience.
The Wikipedia article is a bit misleading because although it cites the EU as an example, most software patents are allowed in the EU, including many H.264 patents. I would encourage you to fight software patents not by infringing, but by using royalty-free alternatives.
[1] https://forum.videolan.org/viewtopic.php?t=125032
(I'm not familiar with Vysor but whatever you do, stay away from libvlc if you're looking for a decoder/encoder. That stuff is not kosher in most countries. ffmpeg is a gray area too.)
Some highlights from a few years ago (may differ slightly now but unlikely):
* No licensing fee
* Free for private/personal, non-commercial usage (e.g. non-monetized YouTube, cat videos)
* < 100,000 units = no royalties! (Unit defined as encoder or decoder)
* > 100,000 units = 0.20c/unit
The process is simple, you ping them, they FedEx you a stack of easy to read papers, sign it, and mail it back.
[IANAL so obvious use-common-sense boilerplate goes here.]
... yes, that's exactly what they are doing?