[aur-general] Notification of GPL violation
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
Unfortunately this seemed to have little effect, forcing me to officially notify the ArchLinux project of this violation.
The GPLv3 in Section 5 clearly states that modified versions must be marked as such
- Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified it, and giving a relevant date.
The term conveyance is also clearly defined in the GPL:
To "propagate" a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other parties to make or receive copies. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying.
Since AUR is distributing packages to automate the process receiving modified versions of the source code, I believe this constitutes conveyance of a modified version, irregardless of where the original source code tarball is fetched from.
Please ensure the cessation of violation of the license immediately.
Regards
lawl
Sent with [ProtonMail](https://protonmail.com) Secure Email.
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org wrote:
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
The "patches" you're talking about are those to apply the correct version for packaging? There's no way any reasonable person would find that violates the licensing.
I think the patch in question has been reduced to only modifying the version 16 hours ago, before that it did a bit more than that[1]. But I guess, in the current state[2], it should be acceptable?
[1] https://aur.archlinux.org/cgit/aur.git/commit/?h=noisetorch&id=0d0edebd6... [2] https://aur.archlinux.org/cgit/aur.git/tree/main.patch?h=noisetorch
Am Mo., 17. Mai 2021 um 15:40 Uhr schrieb Doug Newgard via aur-general < aur-general@lists.archlinux.org>:
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org wrote:
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User
Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply
patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
The "patches" you're talking about are those to apply the correct version for packaging? There's no way any reasonable person would find that violates the licensing.
On Mon, 17 May 2021 15:48:09 +0200 Michael Kogan via aur-general aur-general@lists.archlinux.org wrote:
I think the patch in question has been reduced to only modifying the version 16 hours ago, before that it did a bit more than that[1]. But I guess, in the current state[2], it should be acceptable?
[1] https://aur.archlinux.org/cgit/aur.git/commit/?h=noisetorch&id=0d0edebd6... [2] https://aur.archlinux.org/cgit/aur.git/tree/main.patch?h=noisetorch
All that did was update the patch for the new version, it didn't make any changes in what it does.
And please don't top post.
The license is very clear. I am not going to debate with everyone if any individual patch is acceptable or not.
I expect the ArchLinux project will comply with the license and I'm unsubscribing again from this mailing list since there is nothing to discuss.
You have been notified.
Sent with ProtonMail Secure Email.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ On Monday, May 17, 2021 3:39 PM, Doug Newgard scimmia@archlinux.org wrote:
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org wrote:
Hello I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3). I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
The "patches" you're talking about are those to apply the correct version for packaging? There's no way any reasonable person would find that violates the licensing.
On Mon, 17 May 2021 14:11:41 +0000 lawl via aur-general aur-general@lists.archlinux.org wrote:
The license is very clear. I am not going to debate with everyone if any individual patch is acceptable or not.
I expect the ArchLinux project will comply with the license and I'm unsubscribing again from this mailing list since there is nothing to discuss.
You have been notified.
I guess the "reasonable person" part of my reply is moot.
On Mon, 2021-05-17 at 14:11 +0000, lawl via aur-general wrote:
The license is very clear. I am not going to debate with everyone if any individual patch is acceptable or not.
I expect the ArchLinux project will comply with the license and I'm unsubscribing again from this mailing list since there is nothing to discuss.
You have been notified.
The Arch Linux project does not distribute your software.
We host a repository of user contributed build scripts, we do not distribute source code or any kind of binary build from your software. The AUR *is not* "distributing packages", it is distributing build scripts. The only thing we host is this git repo:
https://aur.archlinux.org/cgit/aur.git/?h=noisetorch
For the sake of avoiding conflict, I think it would be best for the AUR maintainer for this build script to add a notice on the final software describing the modifications (it could simply point to the PKGBUILD). Though, I would like to point out that there is absolutely no basis for demanding the name to be changed.
Several people tried to have a reasonable discussion with you, to no avail, so my reply here lacks most of the courtesy I would have otherwise given and is purely technical.
Filipe Laíns
On 5/17/21 4:11 PM, lawl via aur-general wrote:
The license is very clear. I am not going to debate with everyone if any individual patch is acceptable or not.
I expect the ArchLinux project will comply with the license and I'm unsubscribing again from this mailing list since there is nothing to discuss.
You have been notified.
🍿🍿🍿
On 17/05/2021 17:53, NicoHood via aur-general wrote:
On 5/17/21 4:11 PM, lawl via aur-general wrote:
The license is very clear. I am not going to debate with everyone if any individual patch is acceptable or not.
I expect the ArchLinux project will comply with the license and I'm unsubscribing again from this mailing list since there is nothing to discuss.
You have been notified.
🍿🍿🍿
This list is not for your personal entertainment. Please keep things on-topic.
Alad
Apart of whatever noise reduction is used, the possible full gated (mute) part of this app IMO is a no-go for communication.
In (e.g. music) audio engineering this kind of gate can be used to some degree, but if you've ever talked with somebody who's phone used this kind of gate approach, you'll know how annoying it is.
Even assuming there should be no additional issues (such as e.g. possible imperfect latency, lookahead, cut off workarounds) the problem with muting by gate is:
No voice = no audio or at best noise-damped audio, unfortunately this results in an annoying pumping sound, from silence to voice+noise to silence. It doesn't work for communication.
For multi-track audio (video) recordings it does work, since noise is cut away, if no signal is available, too but if a signal+noise is available, the noise doesn't annoy, because the noise is either covered by other tracks, or by the nature of the sound, e.g. instead of vocals, a distorted guitar.
If audio recordings are done for speech only, by audio or video studios, then almost all engineers are in favour of keeping the noise, to avoid this pumping by turning noise on and off. Especially when using a "decent microphone", it's not much needed to workaround noise at all. Noise gates almost all of the times are only used, if something else does cover the annoying effect of a noise gate. IOW if e.g. another track, such as a piano backs the singer, but for a speaker only, it's most of the times a no-go.
I might be mistaken, since I never used this app. I'm sceptic that it is useful at all, due to my experiences of several decades of audio and video engineering and since I worked as engineer for one of the two famous, well known professional German microphone companies.
Danke fürs Corn-Poppen! :-)
On Mon, 2021-05-17 at 22:23 +0200, Markus Schaaf via aur-general wrote:
Danke fürs Corn-Poppen! :-)
Hi,
please, don't consider my 2 cents as bikeshedding.
"Make sure the package you want to upload is useful." - https://wiki.archlinux.org/title/AUR_submission_guidelines#Rules_of_submissi...
I question the usefulness, in relation to all the trouble with upstream.
Actually, if several people consider something as useful, while I have got doubts, I wouldn't chime in. However, in this case there's some trouble and I wonder, if it's really worth the hassle.
What actually from an audio engineering point of view is speaking for the usefulness?
Greetings from the peanut gallery Ralf
17 мая 2021 г., 23:38 +0300, Ralf Mardorf via aur-general aur-general@lists.archlinux.org, написал (-а):
On Mon, 2021-05-17 at 22:23 +0200, Markus Schaaf via aur-general wrote:
Danke fürs Corn-Poppen! :-)
Hi,
please, don't consider my 2 cents as bikeshedding.
"Make sure the package you want to upload is useful." - https://wiki.archlinux.org/title/AUR_submission_guidelines#Rules_of_submissi...
I question the usefulness, in relation to all the trouble with upstream.
Actually, if several people consider something as useful, while I have got doubts, I wouldn't chime in. However, in this case there's some trouble and I wonder, if it's really worth the hassle.
What actually from an audio engineering point of view is speaking for the usefulness?
Greetings from the peanut gallery Ralf
I'm using this software regularly and I claim its usefulness for me. I can clap my hands while speaking in my voice calls and my peers can't hear my hands clapping. When my children are running around and laughing my peers almost can't hear them.
This is enough for me to say that this software is useful.
Pasha
On Mon, 17 May 2021 23:50:51 +0300, Pasha Finkelshtein wrote:
When my children are running around and laughing my peers almost can't hear them.
This is enough for me to say that this software is useful.
Ok! Without this software, it would be an issue? IOW did you compare the result with and without using the app? If so, my apologies!
Did you sent it off-list on purpose?
Did you sent it off-list on purpose?
Oops, tricky, one of the recipient send "via aur-general".
Pardon, I missed that. So it wasn't offlist.
On 21/05/17 11:00PM, Ralf Mardorf wrote:
On Mon, 17 May 2021 23:50:51 +0300, Pasha Finkelshtein wrote:
When my children are running around and laughing my peers almost can't hear them.
This is enough for me to say that this software is useful.
Ok! Without this software, it would be an issue? IOW did you compare the result with and without using the app? If so, my apologies!
Yes, it would be an issue and yes, I've compared the result.
Did you sent it off-list on purpose?
No, I'm sorry, looks like I've misclicked
The license is very clear. I am not going to debate with
everyone if any
individual patch is acceptable or not.
Your argument holds no candle. If you distribute a software as GPL I can modify in any shape or form I want as long as I honor the license, e.g apply modifications you don't like. Things like this happen everywhere e.g. Arch other distrubtions.
If you don't want people modifying your code don't use FOSS licenses.
I've switched the PKGBUILD to a VCS source and removed any source code modifications [0]. It should now be compliant with the license and the wishes of the upstream maintainer.
[0] https://aur.archlinux.org/cgit/aur.git/commit/PKGBUILD?h=noisetorch&id=a...
On Mon, 17 May 2021 20:38:33 +0200, Justin Kromlinger via aur-general wrote:
I've switched the PKGBUILD to a VCS source and removed any source code modifications [0]. It should now be compliant with the license and the wishes of the upstream maintainer.
[0] https://aur.archlinux.org/cgit/aur.git/commit/PKGBUILD?h=noisetorch&id=a...
I still wonder, if such an app gains something. For speech intelligibility (as well as for anything else) the position of the microphone, as well as the room reflections are important. However, formants, sibilants, etc. need to be considered. Assuming you want the best possible quality for a singer, the required frequency range is more or less from 80 Hz to 12 kHz (not that many aged people are able to hear up to 16 KHz). Rattle noise, flour noise is almost damped already below 80 Hz (around 75 HZ), but you even could cut at 90 or higher Hz, even for a male voice, let alone a female voice. Unwanted high noise is an issue, if you want the full spectrum. Actually you don't need something even near to 12 KHz for speech intelligibility. You can cut high frequencies way, way lower. _If_ there should be noise around the formants range from around 150 to 3000 Hz you can't do anything at all. Consonants might require up to 12 KHz, but actually 8 KHz or lower are likely ok. Analog telephones tended to provide 300 to less than 4000 Hz. I wonder what such an app could provide other, than irritation. My guess is, that getting used to unwanted noise is less irritating, than trying to cancel unwanted noise "smart" by what ever noise reduction, since sometimes the "smart" approach must allow unwanted noise, so the unwanted noise comes and goes. Instead, if there's no way to use a good room and good microphone position, let alone a good microphone and sound card, lowpass and highpass filters should be used as defaults. An audio engineer could tweak a little bit other frequencies, but most unlikely an automation or amateur is able to do it.
Does this app really some a kind of magic?
I've switched the PKGBUILD to a VCS source and removed
any source code
modifications [0]. It should now be compliant with the
license and the
wishes of the upstream maintainer.
Does the license of the package still match? The statement conflicts with the GPL (or any FOSS license for that matter).
Also I might add that since [1][3] the license is not just GPL but extented.
If packager does modications to the software to fix its packageging e.g. [2] its author says it is a fork? ---- [1] https://github.com/lawl/NoiseTorch/commit/ 0ef8229e6fb54c50ec16dd64e24fb32e49b3b207#diff- c693279643b8cd5d248172d9c22cb7cf4ed163a3c98c8a3f69c2717edd3eacb7 [2] https://github.com/lawl/NoiseTorch/issues/124 [3] https://github.com/lawl/NoiseTorch/commit/ 8b054c20a909c9cc459e87ee8b56226c0428201b
On Mon, 17 May 2021 23:15:44 +0300 Bjoern Bidar via aur-general aur-general@lists.archlinux.org wrote:
Does the license of the package still match? The statement conflicts with the GPL (or any FOSS license for that matter).
IANAL but GPL generally seems to allow extension of the license (in reasonable matter). I would say as long as we package the upstream LICENSE file (which we now do) we're in the green, valid GPL or not.
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org said:
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
I am just looking at the AUR repository for this right now. It has no patches. It does not modify anything from the original. It builds the original with standard go tooling and options available in that toolchain (not modifications to your software) and then packages up the resulting build. The only addition is ADDING a .desktop file to that package archive to make it easier to run the application. This does not modify the software. So by your quotes below it meets the license requirements. I don't know what was in the AUR pkg repo before but it certainly seems to be fine now.
Unfortunately this seemed to have little effect, forcing me to officially notify the ArchLinux project of this violation.
The GPLv3 in Section 5 clearly states that modified versions must be marked as such
- Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified it, and giving a relevant date.
The term conveyance is also clearly defined in the GPL:
To "propagate" a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other parties to make or receive copies. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying.
Since AUR is distributing packages to automate the process receiving modified versions of the source code, I believe this constitutes conveyance of a modified version, irregardless of where the original source code tarball is fetched from.
Please ensure the cessation of violation of the license immediately.
Regards
lawl
Sent with [ProtonMail](https://protonmail.com) Secure Email.
On 2021-05-17 21:16, Carsten Haitzler via aur-general wrote:
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org said:
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
I am just looking at the AUR repository for this right now. It has no patches. It does not modify anything from the original. It builds the original with standard go tooling and options available in that toolchain (not modifications to your software) and then packages up the resulting build. The only addition is ADDING a .desktop file to that package archive to make it easier to run the application. This does not modify the software. So by your quotes below it meets the license requirements. I don't know what was in the AUR pkg repo before but it certainly seems to be fine now.
For anyone else wondering what the fuss is all about, this is a long-running bout of drama stirred up by the developer of Noisetorch [1]. They are hostile and will not be reasoned with, unfortunately.
We are under no obligation to listen to these demands as we do not distribute the software and only supply a build script for which a user may build their own package. Kowtowing to this petulance will hardly solve the problem long-term as it is clear that *any* packages not made by the developer are treated with hostility.
But I guess hashworks has already extended the olive branch; We'll have to see whether that's enough...
[1] https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
On 5/18/21 2:06 AM, Brett Cornwall via aur-general wrote:
On 2021-05-17 21:16, Carsten Haitzler via aur-general wrote:
On Mon, 17 May 2021 13:32:26 +0000 lawl via aur-general aur-general@lists.archlinux.org said:
Hello
I'm the developer of NoiseTorch (https://github.com/lawl/NoiseTorch/). I faithfully believe that the package "noisetorch" in the ArchLinux User Repository ("AUR") (https://aur.archlinux.org/packages/noisetorch/) violates my license (GPLv3).
I have the asked previous maintainer of this package to not apply patches or make it clear that this is a fork that's being conveyed. Several Arch Linux trusted users were also informed of this: https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
I am just looking at the AUR repository for this right now. It has no patches. It does not modify anything from the original. It builds the original with standard go tooling and options available in that toolchain (not modifications to your software) and then packages up the resulting build. The only addition is ADDING a .desktop file to that package archive to make it easier to run the application. This does not modify the software. So by your quotes below it meets the license requirements. I don't know what was in the AUR pkg repo before but it certainly seems to be fine now.
For anyone else wondering what the fuss is all about, this is a long-running bout of drama stirred up by the developer of Noisetorch [1]. They are hostile and will not be reasoned with, unfortunately.
We are under no obligation to listen to these demands as we do not distribute the software and only supply a build script for which a user may build their own package. Kowtowing to this petulance will hardly solve the problem long-term as it is clear that *any* packages not made by the developer are treated with hostility.
But I guess hashworks has already extended the olive branch; We'll have to see whether that's enough...
[1] https://github.com/lawl/NoiseTorch/issues/2#issuecomment-785262068
I have reviewed all commits from 811aec6641ff509bd44cd80aa872f71df502e36a forward and there was no GPLv3 violation to begin with. The patch of the version is irrelevant to the operation or use of the software.
(notwithstanding that Arch doesn't distribute anything to begin with)
Probably kinda hard for a Go programmer to understand...
participants (14)
-
alad
-
Bjoern Bidar
-
Brett Cornwall
-
Carsten Haitzler
-
David C. Rankin
-
Doug Newgard
-
Filipe Laíns
-
Justin Kromlinger
-
lawl
-
Markus Schaaf
-
Michael Kogan
-
NicoHood
-
Pasha Finkelshtein
-
Ralf Mardorf