Author has written the code without permission from original manufacturer of silicon .
You cannot have GPL on code that is basically pirated.
Then you write code based on this and claim GPL since the first code is "GPL"
Whats even funnier is that link above claims that license is Apache . Rpi-ES9018K2M-DAC/LICENSE at master * luoyi/Rpi-ES9018K2M-DAC * GitHub
Apache and GPL and incompatible , but its besides the point since the first code in fact has no license , no permission to be distributed or be made public.
For me the license is the text in the file itself.
That code is a regular alsa asoc driver communicating over standard i2c with the chip. The proprietary parts are only the constant values in the header file (regs, values). Is that what the manufacturer forbids to reveal in its NDA when asking for documentation?
I have worked on a few drivers with documentation provided by the manufacturer under NDA (linux/juli.c at master * torvalds/linux * GitHub , linux/quartet.c at master * torvalds/linux * GitHub ). But the resultant driver is GPL.
That code is a regular alsa asoc driver communicating over standard i2c with the chip. The proprietary parts are only the constant values in the header file (regs, values). Is that what the manufacturer forbids to reveal in its NDA when asking for documentation?
I have worked on a few drivers with documentation provided by the manufacturer under NDA (linux/juli.c at master * torvalds/linux * GitHub , linux/quartet.c at master * torvalds/linux * GitHub ). But the resultant driver is GPL.
If the manufacturer permits to release the information (under the NDA) its perfectly fine. Your NDA probably permitted the release of info.
Ess sabre is not allowing releasing the info. Any work based on pirated code will not be able to claim GPL .
This is why after so many years of DACs using Ess Sabre IC , no one has ever used a linux driver (legally) and Clive backed off from his driver. He developed it because he expected that Ess will allow it. Ess main office stopped the driver development and release.
I am not a lawyer. I dont know the answers to hypothetical question , reverse engineering in Japan and if the code can be GPLed after that.
To answer your question someone from Software Freedom Law Center will be better qualified
What I do know is that Gondolf work based on Clive and Luoy cannot be GPLed since (at least Clive) was forbidden to release it. A derivative work (and Gondolf is clear when he says what his sources are) of that cannot be GPLed. Makes sense ?
Is gandolf supposed to know the history of a GPL-marked code he found on github? Actually not history of the code itself (no reason to doubt its author), but of information leading to this code... which is quite easy to get via reverse engineering of slow unencrypted I2C communication...
I understand you have done a lot of work which is "incompatible" with this code...
For me the result is simple - I will always try hard to avoid ESS chips/products based on them. Their attitude is plain stupid, there is no real-value know-how in a list of registers and their expected values - basically the very simpliest user guide for their product.
Gondolf signed a NDA with Ess Sabre . He is aware of what can and cannot be published . Taking what amounts to pirated code and publishing , modify it and distributing it (like he does on diyaudio) ..?
Frankly , would you do that ? Be frank also, do you think this is code that can be distributed under GPL ?
On reverse engineering the i2c and reading registers, I think its a question beyond this conversation . A good article is here Coders’ Rights Project Reverse Engineering FAQ | Electronic Frontier Foundation
I agree with your ess sabre is not being easy. However they have a fallowing and a good product.
Since the beginning I disclosed that I am also making a Ess Sbare 9038 DAC. Its bothering that we are fallowing the rules and complying with NDA and licensing while others try to find an easy shortcut and undermine (ours) and many others hard work.
I understand the appeal of Gondolf code especially for people with ess9028 boards from the far east . Use it in the privacy of your home for testing purposes . Do not link to it , do not distribute it. Do not pretend its GPL
OK, signing an NDA is a binding act. I would not sign in this situation. But I am a hobby developer, I am not making my living by producing audio items...
I believe ESS will reconsider their attitude/decision. Maybe bad in the short term for your business (or maybe not) but certainly good for the users. There is no technical reason to control an MCU to control the DAC if it can be done directly with similar commands. Of course unless the MCU does other things too (LCD, buttons, etc.) but that is a different story.
I believe ESS will reconsider their attitude/decision. Maybe bad in the short term for your business (or maybe not) but certainly good for the users. There is no technical reason to control an MCU to control the DAC if it can be done directly with similar commands. Of course unless the MCU does other things too (LCD, buttons, etc.) but that is a different story.
The NDA does not protect publicly available information, I have never be disclosed about anything regarding ES9018K2M.
This code looks like it was written to control an ES9028/38Pro, not an 9018K2M. Something weird is going on.
ESS should really reconsider all of this NDA nonsense. It's really not helping their brand. ******* off your most enthusiastic customers is never good for business. And yes, we, the DIY audio community are definitely their most enthusiastic customers.
"ESS should really reconsider all of this NDA nonsense. It's really not helping their brand."
Trying to protect your business. Nonsense?
Target customers of ESS are large scale manufactures. This target group can afford some extra hurdles.
ESS couldn't care less about some DIY-audio nerds.
@cdsgames. Thx for sharing your experience and getting all that down
to the bottom.
And I do hope that people over here reading all that don't consider this discussion a "destructive" effort.
I do think that for guys like Ian it's good to know that he has to walk
the extra (MCU) mile.
And for most other fellows around here it's important to know to better stay away from HW that's lacking proper software support.
And not to forget Gandolf. He doesn't want to get trapped in any kind of legal issues. He wouldn't be the first!
Thx. Great discucssion.
Trying to protect your business. Nonsense?
Target customers of ESS are large scale manufactures. This target group can afford some extra hurdles.
ESS couldn't care less about some DIY-audio nerds.
@cdsgames. Thx for sharing your experience and getting all that down
to the bottom.
And I do hope that people over here reading all that don't consider this discussion a "destructive" effort.
I do think that for guys like Ian it's good to know that he has to walk
the extra (MCU) mile.
And for most other fellows around here it's important to know to better stay away from HW that's lacking proper software support.
And not to forget Gandolf. He doesn't want to get trapped in any kind of legal issues. He wouldn't be the first!
Thx. Great discucssion.
Last edited:
The NDA states any publicaly available information is not considered confidential. If you want to accuse the original author is at fault you will need proof that he or she has signed NDA and was not disclosed by anyone except ESS. Otherwise the code can be used without ESS’s permission. As for the license issues, I don’t claim any credit, I simply discovered the GitHub repo and pasted it here. If you have further concerns please contact the original author. If you feel the content is not suitable for the community, feel free to report it to the forum admin.
Last edited:
Well this has all got very tetchy. The only thing an NDA does is establish a civil contract between two parties. It does not protect information once it has become public (contracts between two parties are not binding on third parties who haven't signed them). If a third party uses information that was formerly protected by an NDA but has now become public, there's no legal recourse for the originator. That doesn't necessarily stop them paying lawyers to write threatening letters, but there's no substance to them.
I'm actually not sure how Gondolf and Clive and Luoy stand: if they breached the terms of their NDA (I don't know the terms so I can't comment) then they may be in trouble for breach of contract. Gondolf says the information subject to NDA was already public and therefore not covered, and he is therefore okay. However third parties who receive the information would not be in any kind of trobule, as they have no contract with ESS to breach.
Also don't confuse breach of contract with piracy, which is breach of copyright. The fact that information in this software was once protected by an NDA has no bearing on the copyright licence that it is under.
I'm actually not sure how Gondolf and Clive and Luoy stand: if they breached the terms of their NDA (I don't know the terms so I can't comment) then they may be in trouble for breach of contract. Gondolf says the information subject to NDA was already public and therefore not covered, and he is therefore okay. However third parties who receive the information would not be in any kind of trobule, as they have no contract with ESS to breach.
Also don't confuse breach of contract with piracy, which is breach of copyright. The fact that information in this software was once protected by an NDA has no bearing on the copyright licence that it is under.
Last edited:
Considering that pretty much every other dac chip manufacturer out there is giving out their full datasheets without being afraid that they'll be "ripped off", I'd definitely say 'nonsense'.
Target customers of ESS are large scale manufactures. This target group can afford some extra hurdles.
ESS couldn't care less about some DIY-audio nerds.
That's what I'm saying.. But we are indeed their most enthusiastic customers..
I have no intention to do so.
Gandolf you seem to have the talent to help this community , I hope to see your contributions in the future .
Open source package for LG phone V30 (lgmv300s) contains kernel drivers for two ESS codecs - ES9218 and ES9218P (second one copyrighted by ESS).
I believe they might be used as a base for the ES9028Q2M codec.
Check the register list excerpts from these drivers:
I believe they might be used as a base for the ES9028Q2M codec.
Check the register list excerpts from these drivers:
You can find LG sources at http://opensource.lge.com/
Interesting find!
Are they undermining their own policy!?!?
If you can sell several millions of DACs at once, they might consider it worth it.
They deliver the DACs since G5 - I think - for the G and V series.
Now the question is if this will enable Gandolf and others to go ahead...
...and most important, just apply - if possible at all - the stuff to a different type of DAC!?!?
You'll figure it out.
Are they undermining their own policy!?!?
If you can sell several millions of DACs at once, they might consider it worth it.
They deliver the DACs since G5 - I think - for the G and V series.
Now the question is if this will enable Gandolf and others to go ahead...
...and most important, just apply - if possible at all - the stuff to a different type of DAC!?!?
You'll figure it out.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.