Is ne10 just BSD or BSD and Apache?

[wookey]

I am trying to package ne10 for Debian but have stalled, because there seems to be some confusion about the licensing:

This page: http://projectne10.github.io/Ne10/doc/index.html says "The Ne10 is provided free of charge by ARM Limited and Contributors, and licensed under both New BSD License (http://opensource.org/licenses/BSD-3-Clause) and Apache 2.0 (http://www.apache.org/licenses/LICENSE-2.0)."

But the codebase contains only the BSD licence, and every file has a BSD header. The only mention in the source of Apache is in inc/NE10.h which also contains the above text.

So is it in fact all 'BSD, or Apache2, as the user sees fit', or is the Apache dual-licence mention some old thing that has dropped by the wayside? It seems unclear to me.

[wookey]

So, answering my own query as I have managed to find out inside ARM: The project is BSD only. Apache was considered early on as an alternative to the CLA, but the CLA was picked in the end. We should expunge the remaining obsolete mentions of Apache in the NE10.h file and web page.

[bgat]

Believe it or not, some of us ARE paying attention. ;-)

I'm currently carrying my own branch of NE10, I'll be very happy when it's officially packaged! Thanks for the effort!

(It didn't occur to me before now that I could have signed up to package it. My bad.)

[wookey]

On 2016-09-22 11:15 -0700, Bill Gatliff wrote:

Believe it or not, some of us ARE paying attention. ;-)

I'm currently carrying my own branch of NE10, I'll be very happy when it's officially packaged! Thanks for the effort!

Do you have patches that should go in?

(It didn't occur to me before now that I could have signed up to package it. My bad.)

Yes, you could have :-)

Good to know that there is at least one other user (people are wondering inside ARM how much use this gets, and how mchu demand for a debian package there is)

Wookey

Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

[bgat]

Wookey:

I haven't profiled it or anything, or even affirmatively chosen it: a vendor lib I'm using was linked against it, and so I got dragged in. I literally had to Google the linker symbols to figure out how to respond...

Anyhow, I've just been building and using it as-is. No patches to submit.

Let me know where to send the beer money! :-)

b.g.

[bgat]

(My "carrying my own branch" comment should have said "git submodule" instead. Still annoying compared to having it Just Work, but not the same thing as branching. Sorry for the confusion.)

[Brahmasmi]

@wookey Great to know that this will be packaged for Debian. I am not an expert in this, but I assume it might be especially helpful on ARM servers (like those by @scaleway). Also, I am not sure whether resolving #123 is relevant before packaging.

Now, I apologize in advance for going off on a tangent here. I am sorry. The reason I am dropping in from the peanut gallery is because you have some functioning communication channel with ARM. Can you do me a favour? Please pass on the rest of this as a request/mindchow for the ARM folks.

"Could ARM folks please accept pull request #84 ? That patch has been literally sitting there for 2 years now. In fact, the previous issue - #132 mentions the problem first reported in #83 .

Also, I am a bit surprised that the good folks in ARM are wondering whether this library gets any use. Case in point - Opus. If I am not wrong, @Phillip-Wang helped integrate this excellent library in the Opus codec. Opus is virtually the bedrock of all audio content being consumed, by virtue of its presence in WebRTC, WebM, Ogg Opus etc used by Youtube et al. And guess which class of devices is predominantly used to consume said audio content? Mobile. And guess which processor architecture is the dominant force in the said class of devices? Yes, you guessed it right - ARM, with an ever increasing share of NEON devices.

If anything, I personally feared that this open source foray by ARM was being killed by making it an abandonware - the last commit is literally a year old. ARM has not participated on any issue/pull request. Silence. Zilch. Nada. Nyet. I for one have tried requesting help on #84 by @ mentioning. See #127 , #123 for the pattern. It is unfair to stop supporting, go awol and then wonder if we have any users. #127 literally has an issue reported with the commitment of an ARM developer advocate to monitor the same - with no resolution.

So, I would request ARM folks to please support this excellent library and if possible, commit longterm to this project. If the suits are blocking this, please let us know. And if you do not intend to support this - please communicate that too. And in case I came across as a bit angry, I am sorry for that.

Lastly, a big thank you to everyone for this library."

[wookey]

On 2016-09-24 11:56 -0700, Brahmasmi wrote:

@wookey Great to know that this will be packaged for Debian. I am not an expert in this, but I assume it might be especially helpful on ARM servers (like those by @scaleway). Also, I am not sure whether resolving #123 is relevant before packaging.

Now, I apologize in advance for going off on a tangent here. I am sorry. The reason I am dropping in from the peanut gallery is because you have some functioning communication channel with ARM. Can you do me a favour? Please pass on the rest of this as a request/mindchow for the ARM folks.

"Could ARM folks please accept pull request #84 ? That patch has been literally sitting there for 2 years now. In fact, the previous issue - #132 mentions the problem first reported in #83 .

Apparently Phil Wang left ARM some months ago. I'm not sure who's looking after this project now. Possibly no-one, which is maybe why patches are not getting dealt with? I'll see if I can find out who is responsible. Does github tell us who has commit rights (I can see who has actually commited at https://github.com/projectNe10/Ne10/graphs/contributors )?

Also, I am a bit surprised that the good folks in ARM are wondering whether this library gets any use. Case in point - Opus. If I am not wrong, @Phillip-Wang helped integrate this excellent library in the Opus codec. Opus is virtually the bedrock of all audio content being consumed, by virtue of its presence in WebRTC, WebM, Ogg Opus etc used by Youtube et al.

WebRTC was mentioned as a place these functions are used, e.g. in Mozilla. But (AIUI) as direct functions, not actually using this library as a library.

If anything, I personally feared that this open source foray by ARM was being killed by making it an abandonware - the last commit is literally a year old. ARM has not participated on any issue/pull request. Silence. Zilch. Nada. Nyet. I for one have tried requesting help on #84 by @ mentioning. See #127 , #123 for the pattern. It is unfair to stop supporting, go awol and then wonder if we have any users. #127 literally has an issue reported with the commitment of an ARM developer advocate to monitor the same - with no resolution.

Right, given the CLA, a reasonably active maintainer is needed. I'll see what I can discover.

Wookey

Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

[mdupuy]

Apologies to all. This project has changed hands twice within ARM since I managed years ago. We just finished a discussion about removing ARM's Contributor License Agreement (CLA) requirement from those wishing to make pull requests from outside ARM. That'll be our first hurdle. I expect this to be resolved quickly and easily. The harder part is finding the right staff to manage, test and accept pull requests across multiple arch versions. We've been chronically understaffed for a few years and oversite of this project got a bit lost. We see the value in community support of this library and appreciate the interest. We hope to have the staffing issue resolved soon. I'll badger the current management of software projects like this a bit more to make them aware there is still community interest.

To be clear, this project had been dual licensed years ago but is now, simply, 3 Clause BSD. The CLA is being removed.

Thanks! Matt

On Sat, Sep 24, 2016 at 1:14 PM, wookey notifications@github.com<mailto:notifications@github.com> wrote: On 2016-09-24 11:56 -0700, Brahmasmi wrote:

@wookey Great to know that this will be packaged for Debian. I am not an expert in this, but I assume it might be especially helpful on ARM servers (like those by @scaleway). Also, I am not sure whether resolving #123 is relevant before packaging.

Now, I apologize in advance for going off on a tangent here. I am sorry. The reason I am dropping in from the peanut gallery is because you have some functioning communication channel with ARM. Can you do me a favour? Please pass on the rest of this as a request/mindchow for the ARM folks.

"Could ARM folks please accept pull request #84 ? That patch has been literally sitting there for 2 years now. In fact, the previous issue - #132 mentions the problem first reported in #83 .

Apparently Phil Wang left ARM some months ago. I'm not sure who's looking after this project now. Possibly no-one, which is maybe why patches are not getting dealt with? I'll see if I can find out who is responsible. Does github tell us who has commit rights (I can see who has actually commited at https://github.com/projectNe10/Ne10/graphs/contributors )?

Also, I am a bit surprised that the good folks in ARM are wondering whether this library gets any use. Case in point - Opus. If I am not wrong, @Phillip-Wang helped integrate this excellent library in the Opus codec. Opus is virtually the bedrock of all audio content being consumed, by virtue of its presence in WebRTC, WebM, Ogg Opus etc used by Youtube et al.

WebRTC was mentioned as a place these functions are used, e.g. in Mozilla. But (AIUI) as direct functions, not actually using this library as a library.

If anything, I personally feared that this open source foray by ARM was being killed by making it an abandonware - the last commit is literally a year old. ARM has not participated on any issue/pull request. Silence. Zilch. Nada. Nyet. I for one have tried requesting help on #84 by @ mentioning. See #127 , #123 for the pattern. It is unfair to stop supporting, go awol and then wonder if we have any users. #127 literally has an issue reported with the commitment of an ARM developer advocate to monitor the same - with no resolution.

Right, given the CLA, a reasonably active maintainer is needed. I'll see what I can discover.

Wookey

Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/projectNe10/Ne10/issues/133#issuecomment-249385209, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ABM4W0fi11XCsxMrVX8xzs1g7-fzlf9vks5qtYS4gaJpZM4J7v9E.

Matthew Du Puyhttp://community.arm.com/people/mattdupuy | Principal Software Engineer | ARM G+https://plus.google.com/+MatthewDuPuy/about | Twitterhttps://twitter.com/why_mutate_dup | Githubhttps://github.com/mdupuy | Linkedinhttps://www.linkedin.com/in/matthewdupuy | ARMhttp://community.arm.com/people/mattdupuy | Android Communityhttp://community.arm.com/groups/android-community IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

[wookey]

On 2016-09-24 22:33 -0700, Matthew DuPuy wrote:

To be clear, this project had been dual licensed years ago but is now, simply, 3 Clause BSD. The CLA is being removed.

OK. Thanks for the licence clarification. And I look forward to a CLA-free release (as I never sign CLAs, so you presumably won't include my patches until that's fixed :-) ).

I finished the packaging of this last night so there is now a (multiarched) package in the Debian 'NEW' queue, awaiting FTPmaster approval for the archive, which usually takes a few days. https://ftp-master.debian.org/new.html

It'll appear at https://tracker.debian.org/ne10 once uploaded

And I've just noticed JoeSavage's nice set of warning fixes, so I'll include those in next upload.

Wookey

Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

[Brahmasmi]

@mdupuy I was wondering if the CLA has been removed.

I ask this because from what I can see here - https://github.com/projectNe10/Ne10/blob/master/CONTRIBUTING.md#contributing-to-project-ne10, it still talks about a CLA. I may be wrong here, but there may be a pull request waiting due to this CLA - https://github.com/projectNe10/Ne10/pull/143.

Also, did we have any luck in finding folks within ARM who might be interested in supporting this project?

[wookey]

On 2017-07-05 22:28 +0000, Brahmasmi wrote:

@mdupuy I was wondering if the CLA has been removed.

I ask this because from what I can see here - https://github.com/projectNe10/ Ne10/blob/master/CONTRIBUTING.md#contributing-to-project-ne10, it still talks about a CLA. I may be wrong here, but there may be a pull request waiting due to this CLA - #143.

Also, did we have any luck in finding folks within ARM who might be interested in supporting this project?

I've heard nothing, so I'll try and find out if anyone is going to take ownership. And if not that we just say so and let people get on with it. Prod me again in a while if I drop it (ne10 is quite a long way down my list of priorities - I'm just the Debian maintainer)...

Wookey -- Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

[Brahmasmi]

@mdupuy Thank you for removing the CLA via #178 .

Have we had any luck on the staffing side of things from ARM? This project now has upwards of 600 stars, 100 volks watching it, and nearly 200 forks. Would appreciate it if ARM could actively support this project.

On a side note, I asked on the Opus codec issue list, whether Ne10 was still useful. @mjbshaw from Google/Youtube replied that he saw approximately 1-3% decrease in encoding/decoding times with Ne10 in his profiling [1].

[1] - https://github.com/xiph/opus/issues/63

[wookey]

On 2018-03-19 14:58 -0700, Brahmasmi wrote:

@mdupuy Thank you for removing the CLA via #178 .

Have we had any luck on the staffing side of things from ARM?

Not so's you'd notice. But it is regularly mentioned internally as an example of 'doing open source badly' :-)

We have managed to take the CLA away, which is progress.

This project now has upwards of 600 stars, 100 volks watching it, and nearly 200 forks. Would appreciate it if ARM could actively support this project.

Indeed. Does anyone want to volunteer to be maintainer? An actual volunteer might help significantly in terms of extracting it from ARM's dead and useless grasp.

Wookey -- Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/

[Brahmasmi]

Thank you @wookey for your reply.

I am a layman user, who knows nothing beyond trying to read documentation and compile (and that too sometimes incorrectly). So unfortunately, I would not be able to contribute towards this repo in terms of maintaining/patching it. My contribution has been limited to asking stupid questions and opening issues (and derailing threads, like this one).

Maybe @joesavage might be interested - he seems to understand Ne10, and has already made contributions.

[Phillip-Wang]

Hi all, I am looking at the pull requests now. They will be taken care of before long.

[wookey]

Thanks Phil, for stepping-up (back!) to the maintainership.