Just had a demand from Thomson that I pay royalties on the mp3 decoder and encoder in dBpowerAMP, two scenarios:
they accept a negotiated royalty rate - $2.50 per encoder and $.75 per decoder with minimum of $15,000 y/r is more than is affordable ( dBpowerAMP makes about as much money as working at McDonalds would net you, it is not much but has never been pushed as a money making wagon ),
or the Lame codec + mp3 decoder are removed and links are provided to Russian servers - 'that old trick'.
Just host the offending components on an offshore server, and have the program download them off the internet automatically on first launch.
Sorry to hear that spoon.
You have to wonder what other apps will be approached by Thomson too.
Hmmm, interesting. I thought I remember reading on Thomson's site that they would not pursue royalties on freeware applications. Maybe its the fact that dBpowerAMP Music Converter has a PowerPack add-on you pay for that has Thomson wanting money even though the PowerPack does not add MP3 encoding. I guess they figure that Spoon is making money off a program that has MP3 functionality so they want some of that money.
they accept a negotiated royalty rate - $2.50 per encoder and $.75 per decoder with minimum of $15,000 y/r
This negociated rate seems to be identical to the usual rate.
they accept a negotiated royalty rate - $2.50 per encoder and $.75 per decoder with minimum of $15,000 y/r
This negociated rate seems to be identical to the usual rate.
[a href="index.php?act=findpost&pid=227784"][{POST_SNAPBACK}][/a]
Those were the rates which were unacceptable what I do not mind it paying a certain rate (no minimum) on the parts that are charged for, it is silly to assume (for them) they can extract $2.50 per encoder on a free converter - I would owe them millions!
I will also make a point of geographically splitting up where people are from - I will only pay for those in Germany and USA (where their society has accepted that a mathamatical formula can be patented, ie E=mc^2 - really this has existed, nothing has been invented, it was discovered that is mathamatics).
This is one thing I will not bend over and take it ** *** ****, I will fight as much as possible, I have a strong view of what is fair and what is not.
Does anyone know when Mp3 patents expire? is it 2008?
Cobra ban evasion posts + replies moved here (http://www.hydrogenaudio.org/forums/index.php?showtopic=24084).
or the Lame codec + mp3 decoder are removed and links are provided to Russian servers - 'that old trick'.
If you're talking about mitiok, it's hosted on free.fr...
Or you have other options in mind ?
(BTW www.matroska.org is now hosted in Russia...)
I'm confused. I thought recent versions of LAME had completely removed any patent-infringing code. Or are all MP3 encoders subject to patents, i.e. anything that produces MPEG Layer III compliant files? If so, I guess this would also mean that FAAC infringes on the AAC patent.
We removed all ISO code.
This means that ISO does not have anymore any copyright regarding Lame.
But it does not change the patent situation.
In an mp3 encoder implementation you could avoid SOME of the FhG/Thomson patents, but not ALL of them.
So anyway, an encoder is always using those patents.
Just had a demand from Thomson that I pay royalties on the mp3 decoder and encoder in dBpowerAMP, two scenarios:
they accept a negotiated royalty rate - $2.50 per encoder and $.75 per decoder with minimum of $15,000 y/r is more than is affordable ( dBpowerAMP makes about as much money as working at McDonalds would net you, it is not much but has never been pushed as a money making wagon ),
or the Lame codec + mp3 decoder are removed and links are provided to Russian servers - 'that old trick'.
[a href="index.php?act=findpost&pid=227763"][{POST_SNAPBACK}][/a]
I'm a patent attorney located in NYC. I'd love to take a look at the patents. Does anyone have the patent number(s) handy?
Maybe here:
http://www.mp3licensing.com/patents/index.html (http://www.mp3licensing.com/patents/index.html)
I'm a patent attorney located in NYC. I'd love to take a look at the patents. Does anyone have the patent number(s) handy?
[a href="index.php?act=findpost&pid=227826"][{POST_SNAPBACK}][/a]
It's nice to have someone with the legal experience to take a look at this stuff. I'd love, just for once, to see the little guys smack a large corporation.
It's nice to have someone with the legal experience to take a look at this stuff. I'd love, just for once, to see the little guys smack a large corporation.
[a href="index.php?act=findpost&pid=227947"][{POST_SNAPBACK}][/a]
Why? Don't you invest in "large corporations" everyday?
Just had a demand from Thomson that I pay royalties on the mp3 decoder and encoder in dBpowerAMP, two scenarios:
[a href="index.php?act=findpost&pid=227763"][{POST_SNAPBACK}][/a]
Can we see the letter or e-mail?
I think it might be prudent to see exactly what they had to say before everyone starts speculating and making a bunch of assumptions about this situation.
It's a little hard to offer anything more than "that sucks" for something as vague as what you've told us so far
Maybe here:
http://www.mp3licensing.com/patents/index.html (http://www.mp3licensing.com/patents/index.html)
[a href="index.php?act=findpost&pid=227830"][{POST_SNAPBACK}][/a]
That's a long list of patents. Anyone know which of these they've asserted in the present instance? I wouldn't be surprised, of course, if it were all of them.
That's a long list of patents. Anyone know which of these they've asserted in the present instance? I wouldn't be surprised, of course, if it were all of them.
It is a portfolio. Licenses are granted for the whole portfolio or nothing.
edit: Lame is probably infringing all of them except
*mp3pro ones
*multichannel ones
But beware, there are other portfolios...
I cannot disclose the exact letter they sent as it had confidential all over it, but it is VERY similar to:
http://divineintervention.biz/pdf/divine-c...ng-10-24-02.pdf (http://divineintervention.biz/pdf/divine-chilling-10-24-02.pdf)
A group of companies banded together to get that impressive collection of patents, really (as I have said to them) it is a monolopy and is very anti-competitive for small guys wanting to license - part of the requirements of becoming an ISO standard was that licenses were offered on fair terms, I will drop a letter of complaint to ISO standards, the guy that sells 100 'codecs' should not have to pay the same as the guy that sells 7000, but (see below) that does not seem the problem.
Negotiations are not going so great, but these things take weeks if not months, on further discussion it is not the fact that I am making a little money and they want a cut of that money, it is the fact that there is a free encoder in my free music converter (that is the one the other patent holders have complained about), this is weakening their patent, so any agreement will have to be central around that, it is looking very doubtful I can supply any free mp3 encoder. What gets my back up is that they want to charge me $2.50 per encoder, but they will not charge apple that - apple must have 40,000,000 iTunes out there for which they paid one sum - otherwise apple would have a bill for 100 million dollars, so if you are big then you are ok, if you have limited finances you get pissed on.
Spoon, what is the status of this MP3 licensing issue with Thomson? Are they still breathing down your neck, or did you reach an agreement with them? Obviously, you're still including MP3-encoding capabilities and have a licensing model set-up, but I'm curious to know how it all went with them.
Also, if one purchases the MP3 encoding license that comes with dbpoweramp, a) is there any limit as to how many MP3 files can be encoded using the software, and b) is there any restriction on what can later be done with those files? Can the MP3s be sold without the need to worry about additional royalty problems with Thomson?
I am curious for the sake of my own projects.
Thanks in advance,
Ben
What gets my back up is that they want to charge me $2.50 per encoder, but they will not charge apple that - apple must have 40,000,000 iTunes out there for which they paid one sum - otherwise apple would have a bill for 100 million dollars, so if you are big then you are ok, if you have limited finances you get pissed on.
Unfortunately the world, and particularly the part of it known as patent licensing, is unfair.
Apple will have paid a truckload of money for a royalty buyout which allows them to ship a unlimited number of encoders/decoders. They've done something very similar with H.264 encoders and decoders, although even then they probably haven't bought out their users from the "Internet Broadcast" bits. Interestingly they didn't buy out the MPEG2 patents and to this day it costs $19.99 to license the "QuickTime MPEG2 Playback components".
VC-1 is even more vicious, for example the following is taken from the MS Silverlight EULA
"THIS PRODUCT IS LICENSED UNDER THE VC-1 PATENT PORTFOLIO LICENSES FOR THE PERSONAL AND NON-COMMERCIAL USE OF A CONSUMER TO (A) ENCODE VIDEO IN COMPLIANCE WITH THE VC-1 STANDARD ("VC-1 VIDEO") OR (B) DECODE VC-1 VIDEO THAT WAS ENCODED BY A CONSUMER ENGAGED IN A PERSONAL AND NON-COMMERCIAL ACTIVITY AND/OR WAS OBTAINED FROM A VIDEO PROVIDER LICENSED TO PROVIDE VC-1 VIDEO. NO LICENSE IS GRANTED OR SHALL BE IMPLIED FOR ANY OTHER USE."
This basically says you are in violation of the license agreement if *you* decode something which has been encoded without a valid license from MPEG-LA. This clause caused my company to ban the installation of Silverlight.
A thought does occur though, you could write a script which downloads the lame source and a C compiler (VC Toolkit 2003 can be got for free from MS) then compiles the lame dll on the user's machine after a big click through warning (which everyone will ignore) saying the user needs a patent license from MPEG-LA if they wish to use the binary. As a particularly nasty note you could include a link in the click through which sent a proforma email to MPEG-LA asking for a license for a single encoder/decoder. If you user says no to the click through you fail back onto the MS licensed directshow codec that Windows comes with with a warning saying the quality sucks.
it costs $19.99 to license the "QuickTime MPEG2 Playback components".
Mac OS X, unlike Windows XP (and maybe Vista) comes with DVD decoders pre-installed meaning you can play the dvd from the moment you pop it in the drive.
This, of course, does not apply to QuickTime for Windows.
Speaking of sleeping dragons waking up.... this thread is 4 years old.
VC-1 is even more vicious, for example the following is taken from the MS Silverlight EULA
"THIS PRODUCT IS LICENSED UNDER THE VC-1 PATENT PORTFOLIO LICENSES FOR THE PERSONAL AND NON-COMMERCIAL USE OF A CONSUMER TO (A) ENCODE VIDEO IN COMPLIANCE WITH THE VC-1 STANDARD ("VC-1 VIDEO") OR (B) DECODE VC-1 VIDEO THAT WAS ENCODED BY A CONSUMER ENGAGED IN A PERSONAL AND NON-COMMERCIAL ACTIVITY AND/OR WAS OBTAINED FROM A VIDEO PROVIDER LICENSED TO PROVIDE VC-1 VIDEO. NO LICENSE IS GRANTED OR SHALL BE IMPLIED FOR ANY OTHER USE."
This basically says you are in violation of the license agreement if *you* decode something which has been encoded without a valid license from MPEG-LA. This clause caused my company to ban the installation of Silverlight.
I'm not really sure how 'free for noncommercial uses' is more viscous then 'not free for noncommercial uses'. Actually, compared to the usual MPEG terms, that sounds surprisingly reasonable.
Actually, I'm wondering since I don't have the rest of the license in front of me, do they allow those terms on any VC-1 codec, or just the precompiled one in Silverlight? Because it would be really nice if those terms applied to ffmpeg.
Mac OS X, unlike Windows XP (and maybe Vista) comes with DVD decoders pre-installed meaning you can play the dvd from the moment you pop it in the drive.
It comes with DVDPlayer and decoders specifically licensed for use with DVDPlayer for playing DVDs. To my knowledge the only version of Windows which comes with the required codecs so that WMP can play DVDs is Vista Ultimate. AFAIK OS X does not come with QuickTime components allowing MPEG2 video to be played a normal QuickTime player, if it did there would be no reason for Apple to continue to sell the MPEG2 playback component http://www.apple.com/quicktime/mpeg2/ (http://www.apple.com/quicktime/mpeg2/).
Of course if you buy Final Cut Studio it comes with a "free" copy of the required components...
I'm not really sure how 'free for noncommercial uses' is more viscous then 'not free for noncommercial uses'. Actually, compared to the usual MPEG terms, that sounds surprisingly reasonable.
Actually, I'm wondering since I don't have the rest of the license in front of me, do they allow those terms on any VC-1 codec, or just the precompiled one in Silverlight? Because it would be really nice if those terms applied to ffmpeg.
The license just applies to the binary codec in the MS Silverlight distribution, if it applied to anything then MS wouldn't have the Windows lock in they desire for example it would make Moonlight actually useful, rather than being pointless, if it applied there. You can find the EULA on MSes website if you dig hard enough.
Regarding the viciousness of the license I don't think you understood what I was pointing out. I've not seen a license before which attempts to say that the user *playing* the video is in violation of license agreement if the video was originally *encoded* in without a valid patent license for the codec. How on earth can you know if a video hosted on a random website has been encoded with in the confines of either the Silverlight license or a separate commercial license agreement with MPEG-LA for another implementation of VC-1?
That's a long list of patents. Anyone know which of these they've asserted in the present instance? I wouldn't be surprised, of course, if it were all of them.
It is a portfolio. Licenses are granted for the whole portfolio or nothing.
edit: Lame is probably infringing all of them except
*mp3pro ones
*multichannel ones
But beware, there are other portfolios...
Gabriel - I'm new to this subject, but work in a semi-legal environment and I'm curious about LAME. Has Fraunhofer ever come after LAME? What is the legal status of all the rips I've made with it? Thanks!
LAME does not infringe the patents as long as it is just source code. It is whoever compiles it that is infringing.
LAME does not infringe the patents as long as it is just source code. It is whoever compiles it that is infringing.
So users are not in violation, only those who are developing it?
Funny that this thread came up again.
This is one thing I will not bend over and take it ** *** ****, I will fight as much as possible, I have a strong view of what is fair and what is not.
Yet, stealing from open source projects doesn't keep you awake at night?
Spoon, what is the status of this MP3 licensing issue with Thomson? Are they still breathing down your neck, or did you reach an agreement with them? Obviously, you're still including MP3-encoding capabilities and have a licensing model set-up, but I'm curious to know how it all went with them.
Also, if one purchases the MP3 encoding license that comes with dbpoweramp, a) is there any limit as to how many MP3 files can be encoded using the software, and b) is there any restriction on what can later be done with those files? Can the MP3s be sold without the need to worry about additional royalty problems with Thomson?
I am curious for the sake of my own projects.
It turned out ok, we had a large settlement with them for all the free versions of dBpoweramp we gave out, so it was quite a few lean years when we switched over to licensing. I always have a small smile when I see someone post along the lines 'Lame is free, they should not be charing for it, stick with dBpoweramp Rxx which was free', I smile because I had in the end to pay even for their free version...
>Can the MP3s be sold without the need to worry about
>additional royalty problems with Thomson?
If you were Amazon and sold mp3s you had better get everything legal tickty-boo, because mp3 licensing want 2% of your revenue, allthough I am not sure on what legal grounds this is based, ie I am not convinced that 2% has to be paid (if a patent license is paid for in a program), as there is no legal agreement between yourself and Thompson and patent law does decend into licensed usage. Realistically if you are not in the top 50 of mp3 sellers you will never hear from them. Although I am no lawyer.
Has Fraunhofer ever come after LAME? What is the legal status of all the rips I've made with it? Thanks!
It is whoever compiles it that is infringing.
Patents extend to distribution, I am quite sure you would be perfectly in the law to take lame, compile it yourself and use it yourself without infringing. If you put lame as part of a program, even for free then you would fall under the requirements for a patent.
Yet, stealing from open source projects doesn't keep you awake at night?
Please elaborate. You seem to imply we have stole something from an OS project.
Using something in a way that is not allowed is considered stealing, I would say. Maybe the term stealing is not completely correct. But anyway, you have no right to use FAAD2 (and I think also MAD) in the way you do.
Spoon, what is the status of this MP3 licensing issue with Thomson? Are they still breathing down your neck, or did you reach an agreement with them? Obviously, you're still including MP3-encoding capabilities and have a licensing model set-up, but I'm curious to know how it all went with them.
Also, if one purchases the MP3 encoding license that comes with dbpoweramp, a) is there any limit as to how many MP3 files can be encoded using the software, and b) is there any restriction on what can later be done with those files? Can the MP3s be sold without the need to worry about additional royalty problems with Thomson?
I am curious for the sake of my own projects.
It turned out ok, we had a large settlement with them for all the free versions of dBpoweramp we gave out, so it was quite a few lean years when we switched over to licensing. I always have a small smile when I see someone post along the lines 'Lame is free, they should not be charing for it, stick with dBpoweramp Rxx which was free', I smile because I had in the end to pay even for their free version...
>Can the MP3s be sold without the need to worry about
>additional royalty problems with Thomson?
If you were Amazon and sold mp3s you had better get everything legal tickty-boo, because mp3 licensing want 2% of your revenue, allthough I am not sure on what legal grounds this is based, ie I am not convinced that 2% has to be paid (if a patent license is paid for in a program), as there is no legal agreement between yourself and Thompson and patent law does decend into licensed usage. Realistically if you are not in the top 50 of mp3 sellers you will never hear from them. Although I am no lawyer.
Has Fraunhofer ever come after LAME? What is the legal status of all the rips I've made with it? Thanks!
It is whoever compiles it that is infringing.
Patents extend to distribution, I am quite sure you would be perfectly in the law to take lame, compile it yourself and use it yourself without infringing. If you put lame as part of a program, even for free then you would fall under the requirements for a patent.
Yet, stealing from open source projects doesn't keep you awake at night?
Please elaborate. You seem to imply we have stole something from an OS project.
So as long as I use LAME for nothing more than creating my own MP3 files (with EAC) that are not to be sold afterward, as long as I don't use LAME as a part of another program that I develop, I'm OK?
I, too, have a strong sense of right/wrong and don't want to break the law, even if there's no consequence.
>Using something in a way that is not allowed is considered stealing,
Stealing is a strong word, kind of like the RIIA accusing those who download mp3s as stealing (your referral is a GPL violation, RIIAs is a copyright violation).
Funny you should keep mentioning this, I have an email from 2003 from Jim Corbett (Exeutive Director @ Nero) from 2003 specifically talking about FAAD2 which allows our usage of it, sorry not allows, but actively encourages it....and even if we did not I believe our usage of FAAD2 falls 100% under the sprit of GPL, that is we use it as a library, unmodified. Legally (in our opinion) there is very little difference from a component which is say faad2.exe which could stream out audio on STDIO and a faad2.lib / .dll, they are components, but again this has never been tested in court.
I am a person of principles, we will not use something which an orgainsation or individual does not want us to use, if I get a letter from Nero's management which request that we stop using faad2, then I will remove it (in a commercially reasonable time, given our previous ok to use it from Nero) and fund development of a LGPL m4a (aac) decoder.
I also think the correct method for this kind of communication is by private mail (in the first instance), not to do the dirty laundry in public so to speak.
Of course we encourage usage of the library... under the terms of the GPL.
What is your view on derivative works then and GPL? which is classed as a derrived work in your opinion:
FAAD2 complied as exe and uses stdio, does the program calling FAAD2.exe fall under gpl?
FAAD2 complied as dll, does the program calling FAAD2.dll fall under gpl?
FAAD2 complied as lib, statically linked, does the program calling FAAD2.lib fall under gpl?
In my opinion none are derrived works.
I don't care too much about what is a derived work, but I think GNU considers the last 2 as derived work. For me the 2nd option is sort of a gray area. I would have less problems with that if you wouldn't distribute the binary with your product (so at least offering the part that uses FAAD2 for free) but separate it and giving credit where credit is due. Nowhere on your site do I read that you are using GPLed software in your product (why not? maybe you're not so sure that there is no issue?) nor are you providing any downloads of the code or directions where to obtain it.
I understand you got permission at one point, but that is not our current policy and it would not be hard to retract that, also before it was not our policy to allow usage in any kind of commercial software (I guess you got lucky). Also, I don't see how that gives you any right to not credit us at all.
Did you get similar permission for MAD?
In Germany there is no court decision regarding derivative work yet. But in my company the guideline is IIRC (can look it up again if needed):
- LGPL lib dynamically linked is no derivative work
- GPL lib dynamically linked and the program being tailored for that specific lib is derivative work
- GPL lib and program designed not specificially for that lib (easily exchangable for another lib that does similar things) is no derivative work
- linking statically LGPL or GPL is derivative work
- don't remember about stdio
Perhaps some reading at GNU.org would shed some light on the matter?
http://www.gnu.org/licenses/gpl-faq.html#NFUseGPLPlugins (http://www.gnu.org/licenses/gpl-faq.html#NFUseGPLPlugins)
http://www.gnu.org/licenses/gpl-faq.html#Prelinking (http://www.gnu.org/licenses/gpl-faq.html#Prelinking)
http://www.gnu.org/licenses/gpl-faq.html#GPLCompatInstaller (http://www.gnu.org/licenses/gpl-faq.html#GPLCompatInstaller)
http://www.gnu.org/licenses/gpl-faq.html#G...oprietarySystem (http://www.gnu.org/licenses/gpl-faq.html#GPLInProprietarySystem)
EDIT: The top link sort of ruins the rest ... it would seem that 'borderline' best describes this until more concrete guidance can be issued.
hmmm... if Vorbis were ever the dominant format....... [ in bliss ]
I also think the correct method for this kind of communication is by private mail (in the first instance), not to do the dirty laundry in public so to speak.
Yes.
Well, if the dBpoweramp violates GPL in some way then perhaps people should be aware of this issue before they distribute the program to their friends/etc. and possibly violate the license as well. It would be good to know which versions of the program are illegal at least.
Back to MP3 licensing... Here's what I don't understand... Why can't I buy a licensed copy of LAME (or another stand-alone MP3 encoder)? Is there someone who sells a licensed version?
Lots of programs, especially free programs such as EAC, allow me to use the LAME encoder. But, as far as I can tell, there is no way to use it legally (with these free programs). The royalty is only $2.50 and I wouldn't mind paying $5 or $10 for a fully-legal encoder!
I guess that has to do with the annual minimum amount ($15000).
I would have less problems with that if you wouldn't distribute the binary with your product (so at least offering the part that uses FAAD2 for free) but separate it and giving credit where credit is due. Nowhere on your site do I read that you are using GPLed software in your product (why not? maybe you're not so sure that there is no issue?) nor are you providing any downloads of the code or directions where to obtain it.
I understand you got permission at one point, but that is not our current policy and it would not be hard to retract that, also before it was not our policy to allow usage in any kind of commercial software (I guess you got lucky). Also, I don't see how that gives you any right to not credit us at all.
It was always my understanding our relative permissions to use FAAD2 were never under the GPL, that is not to say this situation does not need resolving. Our m4a codec was always a separate download, no dBpoweramp install includes faad2 (until the current beta release of uPlayer), so we will remove it from uPlayer and include on the web site that faad2 is used under the GPL, then everyone is happy (or not, who knows ).
Did you get similar permission for MAD?
We are under an active NDA for MAD, but even if we were not I would never discuss details with Nero who is potentially a rival.
but I think GNU considers the last 2 as derived work. For me the 2nd option is sort of a gray area.
There is no real difference between:
Program << FAAD2 dll
Program << Generic Decoder DLL << using FAAD2 dll
Program << FAAD2 (stdio)
If your statement is correct 'program' above has to be GPL and therefore nothing can use FAAD2 without being GPL. And what if the FAAD2 is a separate install from the main program (like we do), further muddies the waters. Like I said we are amicable, if Nero do not want us to use FAAD2 we will not.
licensed copy of LAME (or another stand-alone MP3 encoder)? Is there someone who sells a licensed version?
We supply a 'licensed' (as in we pay a patent fee for each version we supply), but as I said before patent laws do not affect the individual.
I think we can discuss GPL for a long time. I agree that we can find a decent solution. Send through PM.
FAAD2 complied as exe and uses stdio, does the program calling FAAD2.exe fall under gpl?
No
FAAD2 complied as dll, does the program calling FAAD2.dll fall under gpl?
Yes. (No for LGPL)
FAAD2 complied as lib, statically linked, does the program calling FAAD2.lib fall under gpl?
Yes. (Yes for LGPL)
It should be pointed out that "fall under gpl" is not right "must be under a GPL compatible license" is correct.
It's vaguely possible to weasel out of case 2 if you have a generic plugin API which where your GPL incompatible host loads a GPLed plugin however it's all very borderline. It helps if the API wasn't invented for the purposes of evading the license and is a standard implemented by many programs. If you are writing such a plugin for a recognised API it's always better to add a specific exception to allow hosts that conform to the plugin API. Obviously in order to add such an exception you must either have the copyright on all the files that make the plugin or get the rights holders to agree to such a license change on their work.
licensed copy of LAME (or another stand-alone MP3 encoder)? Is there someone who sells a licensed version?
We supply a 'licensed' (as in we pay a patent fee for each version we supply), but as I said before patent laws do not affect the individual.
I understand that the patents & royalties don't apply directly to the individual, and I'm not worried about Thomson "coming after me". But, Thomson is
supposed to get a royalty (from the supplier/distributor) for each individual using the software.
I have a couple of applications that use
lame_enc.dll, which (like most people) I downloaded for free. I'm just saying that I wouldn't mind paying (a small amount) for a "legal" copy of that same DLL. (i.e. If the distributor was paying the proper royalty to Thomson.) I also have at least one application with a fully-licensed MP3 encoder.
We are under an active NDA for MAD, but even if we were not I would never discuss details with Nero who is potentially a rival.
Presumably illustrate have some sort of commercial license for MAD which Underbit are perfectly entitled to do being copyright holders on libmad. It is up to the copyright holder to take action if they believe their copyright is being violated, everyone else just has to assume an appropriate agreement is in place. However it is a bit strange to trot out the "I'm under an NDA" as a response to a this question since Underbit advertise on their website that MAD is available under a commercial license.
There is no real difference between:
Program << FAAD2 dll
Program << Generic Decoder DLL << using FAAD2 dll
Program << FAAD2 (stdio)
There is a very real difference between these methods as the FSF (in its license FAQs) and any lawyer versed in the GPL will tell you.
If your statement is correct 'program' above has to be GPL and therefore nothing can use FAAD2 without being GPL. And what if the FAAD2 is a separate install from the main program (like we do), further muddies the waters.
It is correct to say that nothing can link to FAAD2 unless it is licensed under a GPL compatible license. Which licenses are compatible with the GPL depend on the version of the GPL. For example the Apache 2 license is compatible with GPLv3 but not GPLv2 due to additional restrictions (mostly to do with patents) which are present in the Apache 2 license which fall foul of the "no additional restrictions" clause in the GPL. The GPLv3 has its own equivalent patent grant clauses so the additional restrictions clauses do not apply and the licenses are compatible.
Going back to your case having it as a "separate install" is irrelevant. It is how your program uses it which matters.
I presume that the FAAD2 decoder is one of the dlls in this http://codecs.dbpoweramp.com/codecs-new/dB...p-Codec-m4a.exe (http://codecs.dbpoweramp.com/codecs-new/dBpoweramp-Codec-m4a.exe) package. If so then leaving aside the question of linking it into dBpoweramp you are still don't have rights to distribute under the GPL due to:
- Not discharging your obligations under sections 3a or 3b to either accompany the binary with source or make an offer of the source. Section 3c doesn't apply because it counts as commercial distribution.
- Adding additional restrictions in your click wrap license agreement which violates section 6.
Obviously none of this matters if there is a separate license agreement in place with FAAD2 copyright holders.
It is correct to say that nothing can link to FAAD2 unless it is licensed under a GPL compatible license
Lets think some more about this...so if the hard line is taken nothing can go near a GPL lib without being GPL. Lets take Winamp as an example (and faad2, for a moment lets suppose Winamp had no m4a decoding). Now if Nullsoft were to implement a decoder using FAAD2 you are saying that Winamp would have to fall under GPL, even if said codec was a seperate download. What if a 3rd party makes the codec? what if Winamp linked to that codec on their site, or included it on their site? Not so clear cut is it?
I take a different view on these libraries they created to perform a defined function, by including that function in a larger system you are not creating a new derrived work of the library, as the library is still doing what it was designed to do, unchanged, interacting through the interfaces as designated under the GPL (be them functions in a DLL or command line switches of an EXE, no real difference). Now as far as the GPL is concerned what matters legally is the text in the GPL and how a court would rule on that in specific cases not what is written after the fact on gnu.org, notice how the line taken there is even a seperate exe can pass GPL to a controlling program if shared memory is used to communicate, to me it smacks of people subverting ideals and going for the control factor above the ideal of the GPL. Lets not forget why the GPL exists and keeping it specific to FAAD2, the idea of GPL is if anyone improved FAAD2, or fixed a bug in it then the code has to be made available for all, that is it, pure and simple. Including FAAD2 in a larger program does not change it one bit.
... Lets not forget why the GPL exists and keeping it specific to FAAD2, the idea of GPL is if anyone improved FAAD2, or fixed a bug in it then the code has to be made available for all, that is it, pure and simple. Including FAAD2 in a larger program does not change it one bit.
Well, if you link statically FAAD2 with a larger program, you are extending FAAD2's functionality.
I cannot disclose the exact letter they sent as it had confidential all over it, but it is VERY similar to:
Just because it says "confidential" in the email, it doesn't necessarily mean it is so.
I don't wish to drag spoon's specific situation out in public, but I am interested in how "free" free software really is.
This is what got me thinking...
http://www.fsf.org/licensing/essays/free-sw.html (http://www.fsf.org/licensing/essays/free-sw.html)
In my limited, and probably wrong understanding, isn't the present situation the kind of mess that LGPL is supposed to clear up - LGPL says go ahead - use it as a library - GPL doesn't. ?
I wish there was a "do what the **** you like with it" licence!
Of course, in reality, if the software author can't afford a lawyer or court case, then that is the licence that people can use the software under. Licenses, like patents, are a waste of time if you can't defend them; better to spend your time/money elsewhere if this is the case. (not in the current case, just a general observation for truly non-commercial activities).
Cheers,
David.
P.S. These are my questions / opinions to provoke an interesting discussion, where I hope I might learn something!
In my limited, and probably wrong understanding, isn't the present situation the kind of mess that LGPL is supposed to clear up - LGPL says go ahead - use it as a library - GPL doesn't. ?
Yes, but MAD and FAAD2 are not under LGPL.
I wish there was a "do what the **** you like with it" licence!
Wasn't its name BSD licence (http://en.wikipedia.org/wiki/BSD_license) ?
but I am interested in how "free" free software really is
I have seen things change, over the years, it seems the main focus these days is to ring fence free software against the big commercial players like Microsoft, Apple. Gone are the ideals....Take this quote:
``Free software'' is a matter of liberty, not price
This storm in a teacup, has resulted in FAAD2 being removed from the install of uPlayer, uPlayer is a free program, I am not sure that removing FAAD2 from the install has increased peoples liberties, rather the reverse, they now might try it and pronouce it as 'crap' as it does not decode m4a (regardless if they figured out to install m4a as seperate). Then again who are we to grumble, FAAD2 is not ours, in many ways we are like a dog which has been thrown a bone, the dog should not complain about the bone.
Lots of programs, especially free programs such as EAC, allow me to use the LAME encoder. But, as far as I can tell, there is no way to use it legally (with these free programs). The royalty is only $2.50 and I wouldn't mind paying $5 or $10 for a fully-legal encoder!
IIRC, a long time ago I read a suggestion that since most people (ie Windows and Mac users) had an OS that came with an MP3 player, a payment had already been made for the use of the MP3 patents on that machine. Linux, of course, is very important, but is not a practical concern for a lot of us.
Having seen what a legal tangle "Free" software can create (I have no clue about the rights and wrongs of the specific case in this thread, but for the first time I see why M$ have such a big hate on FOSS), I don't know whether having a copy of WMP or iTunes legally covers you for the use of LAME, but it seems OK to me ethically, and I don't think the patent holders are going to come after me.
Lets think some more about this...so if the hard line is taken nothing can go near a GPL lib without being GPL.
I'm fed up arguing with you since you clearly don't understand the intent or the meaning of the GPL and you take a view on what you can do with a GPL library which is not shared by the FSF or various legal judgements which have actually gone to court (mainly in Germany). The GPLv2 is very widely discussed and its implications for C programs are very well understood.
You persist in misunderstanding your obligations under the GPL, I suspect that only thing that will educate you is a lawsuit which based on your current posts you *will* lose if you claiming rights to under the GPL to distribute any of the codecs on your "codec central" page.
You make it sound so clear cut, even Novell disagree with you:
http://www.novell.com/coolsolutions/feature/1532.html (http://www.novell.com/coolsolutions/feature/1532.html)
generally consider a work a derivative if it contains a substantial amount of material from a preexisting work.....The courts have not, however, done much to relate this concept of derivatives to software law, where software routinely "links" to other software in order to create one "meta-program.".....so, too, it seems reasonable to believe that software that dynamically links to GPL code (including Linux) should be safe from the GPL's so-called "contamination."
I guess Novell do not understand it your way either.
> (mainly in Germany).
You might find Germany rule one way, USA another, France another...I know my jurisdiction is outside of the EU so can safely ignore any ruling in Germany.
This has never been about how we use FAAD2 (dll, lib or exe), rather about attribution (an admitted oversight on ourselves, even with the relevent permissions we had from Nero), I believe there is no longer an issue between Nero and ourselves.
I will end with these points: it would take 30 minutes work to move any codec from .lib to .dll, or .exe using stdio, I am 100% sure:
Main Program << Stdio << Generic Stdio Helper.exe << GPL lib
that the main program would not fall under the GPL, where as the generic stdio helper would.
In the end, assuming Nero is the sole copyright owner of the GPL'd code...whatever mutually agreed-upon arrangement is made between spoon and Nero is fine, since they can license it within the GPL or in any other custom way they want.
If that happens, then the discussion is moot.
-brendan
I wish there was a "do what the **** you like with it" licence!
There is
exactly that: WTFPL (http://sam.zoy.org/wtfpl/)
although as already pointed out, BSD is perhaps more popular, if slightly more restrictive.
I demand a bumfight to resolve the issue. Right now.
It is correct to say that nothing can link to FAAD2 unless it is licensed under a GPL compatible license
Lets think some more about this...so if the hard line is taken nothing can go near a GPL lib without being GPL. Lets take Winamp as an example (and faad2, for a moment lets suppose Winamp had no m4a decoding). Now if Nullsoft were to implement a decoder using FAAD2 you are saying that Winamp would have to fall under GPL, even if said codec was a seperate download.
I'm really surprised to hear you ask this. Obviously yes, you cannot make a GLPed Winamp plugin. This is the whole point of the GPL.
What if a 3rd party makes the codec?
Anyone who uses it is violating the GPL.
what if Winamp linked to that codec on their site, or included it on their site? Not so clear cut is it?
No I'd say this is pretty clear. You can't do any of that.
I take a different view on these libraries they created to perform a defined function, by including that function in a larger system you are not creating a new derrived work of the library, as the library is still doing what it was designed to do, unchanged, interacting through the interfaces as designated under the GPL (be them functions in a DLL or command line switches of an EXE, no real difference).
Well your different view differs primarily in that it is incorrect. Otherwise the
Library GPL would be rather redundant, eh?
Anyway, the GNU people even have a FAQ entry:
If a library is released under the GPL (not the LGPL), does that mean that any program which uses it has to be under the GPL or a GPL-compatible license?
Yes, because the program as it is actually run includes the library.
Now as far as the GPL is concerned what matters legally is the text in the GPL and how a court would rule on that in specific cases not what is written after the fact on gnu.org, notice how the line taken there is even a seperate exe can pass GPL to a controlling program if shared memory is used to communicate, to me it smacks of people subverting ideals and going for the control factor above the ideal of the GPL. Lets not forget why the GPL exists and keeping it specific to FAAD2, the idea of GPL is if anyone improved FAAD2, or fixed a bug in it then the code has to be made available for all, that is it, pure and simple. Including FAAD2 in a larger program does not change it one bit.
Wait, if you don't think you're violating the GPL, why be concerned with how you can best cheat it in court?
And then if you're going to try and cheat your way out of it, why are you concerned with "why the GPL exists"? After all, you've already said that all that matters is "how a court would rule"?
None of the plugins that Nullsoft creates use any GPL code nor are licensed under the GPL. There is some LGPL which we fulfill our obligations for either by LGPL'ing the plugin (like the old FLAKE encoder plugin) or by using DLLs for the appropriate libraries (such as the libsndfile-based in_wave which uses libsndfile.dll). I realize no one is making any accusations about us in this thread, just stating that we've adopted a "No GPL" policy to avoid the problems being discussed in the 2008-half of this thread.
That said, there is certainly the possibility Winamp will load a third party GPL plugin, such as in_mad. My understanding is that a GPL Winamp plugin is illegal "on it's face" without specific exemptions. It is certainly an interesting conundrum as Winamp did not purposefully load some GPL plugin, rather it was written by a third party and placed in a certain location by the user.
In response to spoon's query, this is why we use libmp4v2 (which uses the Mozilla Public License which is similar in spirit to LGPL. Most of the changes in libmp4v2 in the past few years have been from contributions given back to MPEG4IP) and license AAC decoding commercially from Coding Technologies. Menno was very helpful to us when we first implemented MP4 support (the 5.02 release notes even thank him) so we were certainly very aware of FAAD but choose not to use it.
It is correct to say that nothing can link to FAAD2 unless it is licensed under a GPL compatible license
Lets think some more about this...so if the hard line is taken nothing can go near a GPL lib without being GPL. Lets take Winamp as an example (and faad2, for a moment lets suppose Winamp had no m4a decoding). Now if Nullsoft were to implement a decoder using FAAD2 you are saying that Winamp would have to fall under GPL, even if said codec was a seperate download.
I'm really surprised to hear you ask this. Obviously yes, you cannot make a GLPed Winamp plugin. This is the whole point of the GPL.
Just to clarify this point, you can make a GPL plugin for Winamp. But only the FAAD copyright owners can make a FAAD GPL plugin for Winamp, because an exception must be made to the license.
Quote: (http://www.gnu.org/licenses/gpl-faq.html#GPLAndPlugins)
Can I apply the GPL when writing a plug-in for a non-free program?
If the program uses fork and exec to invoke plug-ins, then the plug-ins are separate programs, so the license for the main program makes no requirements for them. So you can use the GPL for a plug-in, and there are no special requirements.
If the program dynamically links plug-ins, and they make function calls to each other and share data structures, we believe they form a single program, which must be treated as an extension of both the main program and the plug-ins. This means that combination of the GPL-covered plug-in with the non-free main program would violate the GPL. However, you can resolve that legal problem by adding an exception to your plug-in's license, giving permission to link it with the non-free main program.
Can I apply the GPL when writing a plug-in for a non-free program?
If the program uses fork and exec to invoke plug-ins , then the plug-ins are separate programs, so the license for the main program makes no requirements for them. So you can use the GPL for a plug-in
Exactly the point I have been trying to make, any GPL codec can go into an .exe of its own and be controlled by the main program, without shared memory and by that definition any GPL code could be used...but like 99.999% of software developers, we steer clear of the GPL (if you had of asked me 5 days ago I would have replied nothing we use is under the GPL, Menno said otherwise, we agreed).
People tend to take the GPL not as a legal agreement between two parties (to be sorted out amicably between those 2 parties), instead view it as a religion and deleve in all guns blazing, common sense goes out of the window. Often the double standards of those GPL fanatics is unbelievable, if a Harry Potter book was shared around in a .txt file, it might be viewed as fair game, when in reality both Harry Potter and GPL code are exactly the same, a copyrighted item.
For us now, the matter is closed.
People tend to take the GPL not as a legal agreement between two parties (to be sorted out amicably between those 2 parties), instead view it as a religion and deleve in all guns blazing, common sense goes out of the window. Often the double standards of those GPL fanatics is unbelievable, if a Harry Potter book was shared around in a .txt file, it might be viewed as fair game, when in reality both Harry Potter and GPL code are exactly the same, a copyrighted item.
I'm not really sure how expecting people to read or at least understand the gist of licenses on their code is "a religion", but in general I can see why you'd upset a lot people if you keep depending on dumb luck to be compliant with their licenses.
Why not opensourcing uPlayer/dbPoweramp? It's not like that going to change much for 99,999% of it's users.
Why not opensourcing uPlayer/dbPoweramp? It's not like that going to change much for 99,999% of it's users.
Do you think it's ok to pay for software? dbPoweramp isn't free.
Possibly someone ist trying to make (part of) a living from it...
For me that's perfectly ok. And dbPoweramp surely isn't a mainstream software with many (possible) purchasers but something that is very useful for some people with special requirements. I don't think, anyone has the right to expect the author to give it away for free. If you want an open source solution with the same functionality, write it by yourself and release it...
Opensourcing software doesn't mean you can't keep on selling a binary version. Most people will not know or want to know how to compile the source. Also, I don't see where I expect someone to give 'it' away for free (there is a free version of dbPoweramp, right?), I'm just enquiring if spoon has considered another way out of this (namely, opensourcing his software).
Like I said, for everyone but the few people who actually care it changes nothing. But the upside is you can link to just about any other opensource software without worrying bout this license business.
You're comment about how I should write my own software if I want it opensourced so badly conveys an unbased fear of opensource in general to me. Really, no need. Especially in this software niche where we're in it is extremely unlikely someone will actually fork and maintain you software, because you know best of all that is a tedious job. And the people willing to download the source and compile themselves are doing it for sports, because a free binary is provided already. And those few that download the source to get the paid pro features, we're not going to buy it anyway. And on top of that, there are way around that too (make the pro features closed source for example, you're the author so you can decide if you allow yourself to link against it ).
(make the pro features closed source for example, you're the author so you can decide if you allow yourself to link against it ).
This is a good suggestion, but I think if the "pro" features linked into any GPL software that Spoon did not own the copyright to, then you would be back to the original issue of linking a closed application with GPL code, right?
I am in agreement with Spoon on the fact that the different semantics used to access a GPL library are trivial differentiators. It seems that allowing non-GPL code to access GPL code through an EXE is a loophole that anyone can exploit to get around the license. Yes, in some cases it is not practical to design software in such a way, but in the majority of cases it would have little or no impact.
He could make his main program GPL (or LGPL, I don't know anything of this license business ) and link all the things (from other) against this, which is allowed. The, the pro features are a seperate package, which links against this main program too (but not, directly, to the other parts). Seems to me this is perfectly possible and allowed.
My point was, there are other ways to get out of this while still using FAAD (which is the big problem from what I can gather). In the end I think it's pretty dumb not to have acces to AAC music over some license issue, while I think I just demonstrated it wouldn't have to be like this, and still play by the rules.
Opensourcing software doesn't mean you can't keep on selling a binary version. Most people will not know or want to know how to compile the source. Also, I don't see where I expect someone to give 'it' away for free (there is a free version of dbPoweramp, right?), I'm just enquiring if spoon has considered another way out of this (namely, opensourcing his software).
Opensourcing something isn't always the answer, in the case of dbPoweramp if he put the core under a GPL compatible license he'd be able to link to FAAD2 but he may well lose the right to link to other commercial software. I don't know if dbPoweramp does it but for example it's impossible to add GraceNote cddb support using most opensource licenses.
Linking considerations beside opensourcing something also generally only makese sense when lots of people have an interest in maintaining it as an enabling technology. For example GCC is maintained by lots of companies to enable people to compile stuff for their processor, operating system etc.
Out of interest what does Foobar2000 use for AAC decoding? It says it supports AAC internally but it also says on the license page
"Note that a separate less restrictive license applies to open source parts, downloadable separately.
FAAD2 - Freeware Advanced Audio (AAC) Decoder including SBR decoding
Copyright © 2003-2005 M. Bakker, Nero Software AG, http://www.nero.com" (http://www.nero.com")
I am in agreement with Spoon on the fact that the different semantics used to access a GPL library are trivial differentiators. It seems that allowing non-GPL code to access GPL code through an EXE is a loophole that anyone can exploit to get around the license. Yes, in some cases it is not practical to design software in such a way, but in the majority of cases it would have little or no impact.
It's a barrier of entry thing, in order to "access it through an exe" the GPLed bit of software needs to either be something that can be driven from the command line or via a interprocess communication method. Feeding data via stdin, stdout, sockets etc. In reality there are only a small subset of code you might want to link to which is usable in this way due to speed or interface design reasons. Therefore the GPL achieves the majority of what it sets out to achieve and only fails in a subset of cases which if the FSF tried to fix it would cause a huge amount of collateral damage. It would be extremely difficult, if not impossible, to craft something which allowed a closed source program to be loaded by a GPLed OS but prevented a closed source program exec(3) ing a GPLed program and then talking to it via an interprocess communication method.
...Also, I don't see where I expect someone to give 'it' away for free...
You are right, you never said so.
...You're comment about how I should write my own software if I want it opensourced...
Was absolutely inappropriate.
Please accept my apologies!
I am maintaining my opinion but the way i expressed it wasn't fair.
Thomas
(make the pro features closed source for example, you're the author so you can decide if you allow yourself to link against it ).
This is a good suggestion, but I think if the "pro" features linked into any GPL software that Spoon did not own the copyright to, then you would be back to the original issue of linking a closed application with GPL code, right?
Exactly.
I find it ironic that the best ripping programs are not open source, yet open source code does exist for this function. Instead of demanding that someone change the way he conducts business, maybe spend your time on improving the code that's currently available to you. If open sourcing is so superior then you guys should be able to surpass dBpoweramp and EAC in no time.
A propos MP3 patents... Do you actually only pay 0.75 USD per unit or do you have to pay 15.000 USD per year even if you don't sell 20.000 units? How do you proove how many units you sold?
A propos MP3 patents...
You mean steering away from the open source zealotry and back on topic? Well played!
0.75 USD is only per decoder, not encoder, which is from 2.50 to 5.00, with 5.00 per codec you can have a lessor minimum per year (if I remember correctly).
I find it ironic that the best ripping programs are not open source, yet open source code does exist for this function. Instead of demanding that someone change the way he conducts business, maybe spend your time on improving the code that's currently available to you. If open sourcing is so superior then you guys should be able to surpass dBpoweramp and EAC in no time.
Who's demanding opensourcing code here? Who's saying opensourcing is superior (as opposed to what)? I've seen no-one do that. What I've seen is people floating the idea around to do that, because from some points of view it makes sense. Also, from an endusers perspective, I really can't think of any reason against it.
Also, on the topic of ripping CD's, I think we all agree that cdparanoia is the best way to rip CD's (EAC), and cdparanoia is opensource.
So please, stop insinuating that opensource zealots are trying to 'force' their GPL around, because that's just not true. What apparently is true, that some people have misconceptions about the point of OS, the reason for OS and the way it can work in a scheme that makes money for the softwarewriter.
Correct me if I'm wrong but I thought EAC didn't use the cdparanoia method any more? The newer secure method is different than the cdparanoia method, but the cdparanoia method is still available if users want to use it.
EAC has a "paranoid mode" but it doesn't use cdparanoia.
Continuing off-topic discussions about open-sourcing dbpoweramp will be moved to the recycle bin.
I wish there was a "do what the **** you like with it" licence!
MIT License?