Bioconductor package flowPeaks license Artistic 1.0?

Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Dear,

I am packaging bunch of Flow Cytometry packages and `flowPeaks` says
Artistic 1.0 license.
See https://www.bioconductor.org/packages/release/bioc/manuals/flowPeaks/man/flowPeaks.pdf

And in `license.scm` there is only the 2.0.

Because I am not an expert about the licenses, which one select for the package.


Thank you in advance.

All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Ricardo Wurmus]

zimoun <zimon.toutoune@gmail.com> writes:

> I am packaging bunch of Flow Cytometry packages and `flowPeaks` says
> Artistic 1.0 license.
> See https://www.bioconductor.org/packages/release/bioc/manuals/flowPeaks/man/flowPeaks.pdf
>
> And in `license.scm` there is only the 2.0.

That’s because version 1.0 is considered non-free.  “licenses.scm” also
contains “clarified-artistic”, which is essentially the same as version
1.0 but with a few clarifications of those points that could be
interpreted as conditions making the software non-free.

-- 
Ricardo

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Ricardo,

Thank you for your quick feedback.

On Wed, 24 Jul 2019 at 23:16, Ricardo Wurmus <rekado@elephly.net> wrote:
>
> That’s because version 1.0 is considered non-free.  “licenses.scm” also
> contains “clarified-artistic”, which is essentially the same as version
> 1.0 but with a few clarifications of those points that could be
> interpreted as conditions making the software non-free.

Thank you for the explanations.

I sent an email to the maintainer to ask them to "upgrade" the
license. Let see. :-)
Otherwise, I understand.


All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Ricardo Wurmus]

zimoun <zimon.toutoune@gmail.com> writes:

> Hi Ricardo,
>
> Thank you for your quick feedback.
>
> On Wed, 24 Jul 2019 at 23:16, Ricardo Wurmus <rekado@elephly.net> wrote:
>>
>> That’s because version 1.0 is considered non-free.  “licenses.scm” also
>> contains “clarified-artistic”, which is essentially the same as version
>> 1.0 but with a few clarifications of those points that could be
>> interpreted as conditions making the software non-free.
>
> Thank you for the explanations.
>
> I sent an email to the maintainer to ask them to "upgrade" the
> license. Let see. :-)

It would be great if they could use the Clarified Artistic License
instead.  It’s really close to the Artistic 1.0, so unless they really
want the non-free interpretation of Artistic 1.0 it should be no trouble
for them to switch.

-- 
Ricardo

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Ricardo,


On Thu, 25 Jul 2019 at 14:47, Ricardo Wurmus <rekado@elephly.net> wrote:

> > On Wed, 24 Jul 2019 at 23:16, Ricardo Wurmus <rekado@elephly.net> wrote:

> >> That’s because version 1.0 is considered non-free.  “licenses.scm” also
> >> contains “clarified-artistic”, which is essentially the same as version
> >> 1.0 but with a few clarifications of those points that could be
> >> interpreted as conditions making the software non-free.

> It would be great if they could use the Clarified Artistic License
> instead.  It’s really close to the Artistic 1.0, so unless they really
> want the non-free interpretation of Artistic 1.0 it should be no trouble
> for them to switch.

I have no news from the flowPeak's maintainer and I think the package
is still in Bioconductor 3.10 but without any recent updates.

The file guix/licenses.scm contains "non-copyleft" therefore why do
not put the licenses Artistic 1.0 under this label? It will allow the
inclusion of this package -- and probable others from Bioconductor.

Well, I have read both licenses and the Clarified one does not appear
me clearer; they are both doomed!
Other said, calling Artistic 1.0 non-free in this Bioconductor case is
more a flavour of taste than a real legal issue. Especially when this
very Artistic 1.0 "qualifies as a free software license, but it may
not be a real copyleft" [1].


[1] https://www.gnu.org/licenses/license-list.en.html#PerlLicense

All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Ricardo Wurmus]

zimoun <zimon.toutoune@gmail.com> writes:

> The file guix/licenses.scm contains "non-copyleft" therefore why do
> not put the licenses Artistic 1.0 under this label? It will allow the
> inclusion of this package -- and probable others from Bioconductor.

That wouldn’t be correct.  non-copyleft is for free licenses only, and
the Artistic 1.0 does not qualify.

> Well, I have read both licenses and the Clarified one does not appear
> me clearer; they are both doomed!
> Other said, calling Artistic 1.0 non-free in this Bioconductor case is
> more a flavour of taste than a real legal issue. Especially when this
> very Artistic 1.0 "qualifies as a free software license, but it may
> not be a real copyleft" [1].
>
>
> [1] https://www.gnu.org/licenses/license-list.en.html#PerlLicense

https://www.gnu.org/licenses/license-list.en.html#ArtisticLicense says:

    “We cannot say that this is a free software license because it is
     too vague; some passages are too clever for their own good, and
     their meaning is not clear. We urge you to avoid using it, except
     as part of the disjunctive license of Perl.”

However:

https://www.gnu.org/licenses/license-list.en.html#ClarifiedArtistic

    “This license is a free software license, compatible with the
     GPL. It is the minimal set of changes needed to correct the
     vagueness of the Artistic License 1.0.”

--
Ricardo

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Ricardo,

Thank you for the quick feedback.

On Thu, 19 Dec 2019 at 18:18, Ricardo Wurmus <rekado@elephly.net> wrote:

> zimoun <zimon.toutoune@gmail.com> writes:
>
> > The file guix/licenses.scm contains "non-copyleft" therefore why do
> > not put the licenses Artistic 1.0 under this label? It will allow the
> > inclusion of this package -- and probable others from Bioconductor.
>
> That wouldn’t be correct.  non-copyleft is for free licenses only, and
> the Artistic 1.0 does not qualify.

The Perl License section says:

<<
This license is the disjunction of the Artistic License 1.0 and the
GNU GPL—in other words, you can choose either of those two licenses.
It qualifies as a free software license, but it may not be a real
copyleft. It is compatible with the GNU GPL because the GNU GPL is one
of the alternatives.

We recommend you use this license for any Perl 4 or Perl 5 package you
write, to promote coherence and uniformity in Perl programming.
Outside of Perl, we urge you not to use this license; it is better to
use just the GNU GPL.
>>

https://www.gnu.org/licenses/license-list.en.html#PerlLicense


I read "It qualifies as a free software license, but it may not be a
real copyleft." therefore it means non-copyleft.



> https://www.gnu.org/licenses/license-list.en.html#ArtisticLicense says:
>
>     “We cannot say that this is a free software license because it is
>      too vague; some passages are too clever for their own good, and
>      their meaning is not clear. We urge you to avoid using it, except
>      as part of the disjunctive license of Perl.”
>
> However:
>
> https://www.gnu.org/licenses/license-list.en.html#ClarifiedArtistic
>
>     “This license is a free software license, compatible with the
>      GPL. It is the minimal set of changes needed to correct the
>      vagueness of the Artistic License 1.0.”
>

I already know these statements. And I disagree. Currently, the
license is considered free when applied to Perl but non-free
otherwise. It does not make sense.


Well, if I understand well, as GNU Guix maintainer, you will have the
official GNU position, right?
So let discuss this official GNU position. :-)
Do you know in which mailing list can I post?


Cheers,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Ricardo Wurmus]

zimoun <zimon.toutoune@gmail.com> writes:

> I already know these statements. And I disagree. Currently, the
> [artistic] license [1.0] is considered free when applied to Perl but non-free
> otherwise. It does not make sense.

This is a misunderstanding.  The Perl license says[1]:

    It is free software; you can redistribute it and/or modify it under the terms of either:

    a) the GNU General Public License as published by the Free Software
       Foundation; either version 1, or (at your option) any later version,
       or
    b) the "Artistic License". […]

We choose the first one when redistributing Perl, so Perl is under the
GPLv1+.  The Artistic License (1.0) does not come into play at all
because we are fine with GPLv1+.

If Perl was *only* available under the Artistic License 1.0 we could not
in fact redistribute it with Guix and that would be catastrophic.
Luckily, the license allows the distribution under the terms of the GPL,
so we’re fine.

[1]: https://dev.perl.org/licenses/

--
Ricardo

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Ricardo,

On Thu, 19 Dec 2019 at 21:13, Ricardo Wurmus <rekado@elephly.net> wrote:

> This is a misunderstanding.  The Perl license says[1]:

Thank you for the pointer and the explanation.
All is clear and I am fine. :-)

Cheers,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Tobias Geerinckx-Rice]

[-- Attachment #1: Type: text/plain, Size: 2462 bytes --]

Zimoun,

Thank you for fighting for this package in Guix.  I hope upstream 
sees the light and Clarifies things.

zimoun 写道：
>  Ricardo Wurmus <rekado@elephly.net> wrote:
>> It would be great if they could use the Clarified Artistic 
>> License
>> instead.  It’s really close to the Artistic 1.0, so unless they 
>> really
>> want the non-free interpretation of Artistic 1.0 it should be 
>> no trouble
>> for them to switch.

This is the only solution.  Any other licence in licenses.scm is 
fine too.

> The file guix/licenses.scm contains "non-copyleft" therefore why 
> do
> not put the licenses Artistic 1.0 under this label? It will 
> allow the
> inclusion of this package -- and probable others from 
> Bioconductor.

‘Non-copyleft’ does not mean ‘non-free’.  All packages in Guix 
must be free.  The Artistic 1.0 licence is *not free*.[0]

I do understand your frustration & hacker instinct to ‘fix’ the 
problem in some clever way, but that's not how licences work.  The 
Artistic 1.0 story really ends here.

I'm not trying to demotivate you.  I just don't want you to waste 
your time & effort in this dead-end direction.  Bugging upstream 
until they respond is the only solution.

> Well, I have read both licenses and the Clarified one does not 
> appear
> me clearer; they are both doomed!

I hope you'll understand that I'm also not trying to be rude when 
I say (y)our personal opinions are entirely valid and absolutely 
irrelevant :-)

The FSF's legal counsel has decided that the Clarified version 
does in fact ‘correct the vagueness of of the Artistic License 
1.0’[2].

> Other said, calling Artistic 1.0 non-free in this Bioconductor 
> case is
> more a flavour of taste than a real legal issue.

No, it's a very real legal issue.  :-(

> Especially when this
> very Artistic 1.0 "qualifies as a free software license, but it 
> may
> not be a real copyleft" [1].

…but that's not this very licence, it's a completely different 
one: the (disjunct) combination of the Artistic 1.0 licence *and 
the GPL*, i.e. ‘choose one’.  The result is only free because you 
can *ignore* the Artistic 1.0 part.

Kind regards,

T G-R

[0]: 
https://www.gnu.org/licenses/license-list.en.html#ArtisticLicense
[1]: https://www.gnu.org/licenses/license-list.en.html#PerlLicense
[2]: 
https://www.gnu.org/licenses/license-list.en.html#ClarifiedArtistic

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Tobias Geerinckx-Rice]

[-- Attachment #1: Type: text/plain, Size: 918 bytes --]

Tobias Geerinckx-Rice 写道：
> zimoun 写道：
>> Other said, calling Artistic 1.0 non-free in this Bioconductor 
>> case
>> is more a flavour of taste than a real legal issue.
>
> No, it's a very real legal issue.  :-(

I should clarify: when the FSF calls the Artistic 1.0 licence 
‘vague’, that's not an aesthetic criticism.

It means that the licence is broken and fails to do what it claims 
to do: give you the licence (=freedom) to do something that would 
not otherwise be allowed by copyright law.  It means that you 
can't prove, in court, that the licence says what you thought it 
said.  It's not merely ugly, it's defective and potentially 
dangerous.

This always happens when programmers think they can write their 
own licence.  It starts with a punny name (‘artistic licence’, 
‘WTFPL’, ha ha -_-) and the result is a useless buggy mess.

Kind regards,

T G-R

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Tobias,

Thank you for the clarification.


On Thu, 19 Dec 2019 at 18:29, Tobias Geerinckx-Rice <me@tobias.gr> wrote:
>
> Tobias Geerinckx-Rice 写道：
> > zimoun 写道：
> >> Other said, calling Artistic 1.0 non-free in this Bioconductor
> >> case
> >> is more a flavour of taste than a real legal issue.
> >
> > No, it's a very real legal issue.  :-(
>
> I should clarify: when the FSF calls the Artistic 1.0 licence
> ‘vague’, that's not an aesthetic criticism.
>
> It means that the licence is broken and fails to do what it claims
> to do: give you the licence (=freedom) to do something that would
> not otherwise be allowed by copyright law.  It means that you
> can't prove, in court, that the licence says what you thought it
> said.  It's not merely ugly, it's defective and potentially
> dangerous.

I agree.
It is what I tried to express in my other email.
You explained here better than I did elsewhere. :-)

> This always happens when programmers think they can write their
> own licence.  It starts with a punny name (‘artistic licence’,
> ‘WTFPL’, ha ha -_-) and the result is a useless buggy mess.

I am on the same wavelength.


All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Tobias,

Thank you for the explanations. I agree, almost. ;-)

On Thu, 19 Dec 2019 at 18:18, Tobias Geerinckx-Rice <me@tobias.gr> wrote:

> Thank you for fighting for this package in Guix.  I hope upstream
> sees the light and Clarifies things.

The issue is that upstream has disappeared, as usual in scientific
software. Someone writes a piece of code then publishes a paper and
sometimes the requirement for publication is to be pushed in
mainstream collection of packages (Bioconductor in this case). But the
copyright holder does not maintain the code and instead write another
piece of code, try to publish a paper, etc.. Well the Reproducibility
of Science crisis.


> zimoun 写道：

> ‘Non-copyleft’ does not mean ‘non-free’.  All packages in Guix
> must be free.  The Artistic 1.0 licence is *not free*.[0]

It is not my point.

Artistic 1.0 is free and non-copyleft when applied to Perl. And it
does not make sense. A license is free or not, independently to what
it is applied to.


> I do understand your frustration & hacker instinct to ‘fix’ the
> problem in some clever way, but that's not how licences work.  The
> Artistic 1.0 story really ends here.

As I said, saying that Artistic 1.0 is non-free is really a flavour of taste.




> I'm not trying to demotivate you.  I just don't want you to waste
> your time & effort in this dead-end direction.  Bugging upstream
> until they respond is the only solution.

They will not, sadly.



> > Well, I have read both licenses and the Clarified one does not
> > appear
> > me clearer; they are both doomed!
>
> I hope you'll understand that I'm also not trying to be rude when
> I say (y)our personal opinions are entirely valid and absolutely
> irrelevant :-)

My point is: claiming that Clarified Artistic 1.0 is free and Artistic
1.0 is not is a flavour of taste.
What does it mean "free"? Well, it is defined by some GNU documents.
Then, someone reads this definition and sees if the license is
compliant with the definition. And there is an interpretation. That's
why GNU considers some license free and Debian not (or the contrary).

Well, I disagree to say Clarified Artistic 1.0 is free and Artistic
1.0 is not. They are both free or both non-free.


> The FSF's legal counsel has decided that the Clarified version
> does in fact ‘correct the vagueness of of the Artistic License
> 1.0’[2].

I understand. And I disagree. So I appeal. :-)


> > Other said, calling Artistic 1.0 non-free in this Bioconductor
> > case is
> > more a flavour of taste than a real legal issue.
>
> No, it's a very real legal issue.  :-(

Yes it is.
Because the main purpose of a license is to manage what happens in Court.



Well, I understand you are defending the official GNU position.
And currently Artisitic 1.0 will not be included in GNU Guix because
currently GNU claims that Artistic 1.0 is non-free.
I am fine with that. :-)
And I understand you block my hacky proposal of non-copyleft.
I am also fine with that. :-)

So I will appeal to FSF/GNU. ;-)


All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Tobias Geerinckx-Rice]

[-- Attachment #1: Type: text/plain, Size: 2495 bytes --]

Zimoun,

zimoun 写道：
> On Thu, 19 Dec 2019 at 18:18, Tobias Geerinckx-Rice 
> <me@tobias.gr> wrote:
>> Thank you for fighting for this package in Guix.  I hope 
>> upstream
>> sees the light and Clarifies things.
>
> The issue is that upstream has disappeared, as usual in 
> scientific
> software. Someone writes a piece of code then publishes a paper 
> and
> sometimes the requirement for publication is to be pushed in
> mainstream collection of packages (Bioconductor in this 
> case). But the
> copyright holder does not maintain the code and instead write 
> another
> piece of code, try to publish a paper, etc.. Well the 
> Reproducibility
> of Science crisis.

That is a shame.  And that while other scientists (like you) are 
working hard to make research more ‘open’ and reproducible.

However, even if they don't maintain the code they can still 
relicence it with no effort on their part.  We can still hope.

>> zimoun 写道：
>
>> ‘Non-copyleft’ does not mean ‘non-free’.  All packages in Guix
>> must be free.  The Artistic 1.0 licence is *not free*.[0]
>
> It is not my point.

I think au fond it is.  Because your point was:

> Artistic 1.0 is free and non-copyleft when applied to Perl.
  ^^^^^^^^^^^^^^^^^^^^

This just isn't true, and that's what I wanted to make clear 
above.

*Perl* is free, only because it is licenced under the GPL.  If 
 Perl were licenced only under the Artistic 1.0 licence, it would 
 not be in Guix.  I promise.

So Perl is really not relevant to this discussion at all.

If I'm still not making myself clear, I apologise & capitulate. 
We agree on all important points:

>> The FSF's legal counsel has decided that the Clarified version
>> does in fact ‘correct the vagueness of of the Artistic License
>> 1.0’[2].
>
> I understand. And I disagree. So I appeal. :-)

Hehe.  Even the FSF agrees that the Clarified version does only 
the *bare minimum* to not completely suck.  They certainly don't 
recommend it.

> Well, I understand you are defending the official GNU position.
> And currently Artisitic 1.0 will not be included in GNU Guix 
> because
> currently GNU claims that Artistic 1.0 is non-free.
> I am fine with that. :-)

Correct.  It's only the comparison with Perl that's bogus, not 
your opinions on the Artistic 1.0 licence itself.

> So I will appeal to FSF/GNU. ;-)

I admire your tenacity.  Good luck!

Kind regards,

T G-R

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Tobias,

On Thu, 19 Dec 2019 at 21:24, Tobias Geerinckx-Rice <me@tobias.gr> wrote:

[...]

> > piece of code, try to publish a paper, etc.. Well the
> > Reproducibility
> > of Science crisis.
>
> That is a shame.  And that while other scientists (like you) are
> working hard to make research more ‘open’ and reproducible.

Here 'open' is not enough. ;-)


> However, even if they don't maintain the code they can still
> relicence it with no effort on their part.  We can still hope.

The issue is really to be able to contact the author. And I am not
sure this person is even the copyright holder. (In some country, the
company/institute own the copyright even the code is not written in
office's hours.)


For example, 2 files contains:

<<
* The author of this software is Steven Fortune.  Copyright (c) 1994 by AT&T
* Bell Laboratories.
* Permission to use, copy, modify, and distribute this software for any
[...]
* This code was originally written by Stephan Fortune in C code.  I,
Shane O'Sullivan,
* have since modified it, encapsulating it in a C++ class and, fixing
memory leaks and
* adding accessors to the Voronoi Edges.
* Permission to use, copy, modify, and distribute this software for any
[...]
>>

The most of the files claim:

<<
* The author of this software is Yongchao Ge.
* Permission to use, copy, modify, and distribute this software for any
* purpose without fee is hereby granted, provided that this entire notice
* is included in all copies of any software which is or includes a copy
* or modification of this software and in all copies of the supporting
* documentation for such software.
* THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
* WARRANTY.  IN PARTICULAR, THE AUTHOR DOES NOT MAKE ANY
* REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
* OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
>>


Well, the only mention of the Artistic 1.0 license is these 3 files:
README, DESCRIPTION and vignettes/flowPeaks-guide.Rnw.


Does that mean I am allowed to reuse almost everything and repack in
another repo licensing with a "good" license?


[...]

> If I'm still not making myself clear, I apologise & capitulate.

I got it. Yes it is clear!


> > So I will appeal to FSF/GNU. ;-)
>
> I admire your tenacity.  Good luck!

Before I need to assembling the file. :-)
For example, how many packages in Bioconductor use the Artistic 1.0?


Thank for all your explanation and your time.

All the best,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Giovanni Biscuolo]

Hello zimoun,

zimoun <zimon.toutoune@gmail.com> writes:

[...]

> The issue is really to be able to contact the author. And I am not
> sure this person is even the copyright holder. (In some country, the
> company/institute own the copyright even the code is not written in
> office's hours.)
>
>
> For example, 2 files contains:

[...]

> The most of the files claim:

[...]

> For example, how many packages in Bioconductor use the Artistic 1.0?

Sorry you have to struggle with this tedious work of sorting out YALM
(Yet Another Licensing Mess), but the first thing to do in this case is
to have a list of licenses for each file/folder and see if there is a
way to **workaround** the disappearing of upstream and if needed do some
sort of _soft_ forking just to fix the missing licensing-bits

If we are lucky enough maybe the 95% of this package is free and the
remainging 5% easily replaceable with a free rewrite

WDYT?

Sorry I cannot help you sorting out the licenses in this package (next
year I'm going to refactor my timetable for Guix... but this is another
story)

Thanks! Gio'

[...]

P.S.: like Tobias, I suggest you not to spend time trying to appeal FSF
on the Artistic Licence v.1 ;-)

-- 
Giovanni Biscuolo

Xelera IT Infrastructures

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Giovanni,


On Fri, 20 Dec 2019 at 10:28, Giovanni Biscuolo <g@xelera.eu> wrote:

> zimoun <zimon.toutoune@gmail.com> writes:
>
> [...]
>
> > The issue is really to be able to contact the author. And I am not
> > sure this person is even the copyright holder. (In some country, the
> > company/institute own the copyright even the code is not written in
> > office's hours.)
> >
> >
> > For example, 2 files contains:
>
> [...]
>
> > The most of the files claim:
>
> [...]
>
> > For example, how many packages in Bioconductor use the Artistic 1.0?
>
> Sorry you have to struggle with this tedious work of sorting out YALM
> (Yet Another Licensing Mess), but the first thing to do in this case is
> to have a list of licenses for each file/folder and see if there is a
> way to **workaround** the disappearing of upstream and if needed do some
> sort of _soft_ forking just to fix the missing licensing-bits
>
> If we are lucky enough maybe the 95% of this package is free and the
> remainging 5% easily replaceable with a free rewrite

1.
This is my hope for the package flowPeak. Because it blocks my workflow at work.
Now, this package is in a personal channel but nothing provides a
guarantee that this channel would not disappear so the paper I am
working on would not be easily reproducible (in theory and
principles).
Be in the Guix tree affords more chance.

2.
This fix -- reuse all the free available code and replace the non-free
one -- do not scale.
So the question is: What is the scale we are talking about? How many
packages in Bioconductor?
If it is, say, a couple then it is doable. Or see with the people
managing Bioconductor.
If it is more, then the option is lobbying. :-)




> P.S.: like Tobias, I suggest you not to spend time trying to appeal FSF
> on the Artistic Licence v.1 ;-)

I have used frenchy bad faith rhetoric argument. ;-)

As I said, if a couple on Bioconductor are Artistic 1.0, that's ok.
Otherwise, it is an issue.
Right now, there is too much *if*. :-)


Thanks,
simon

Re: Bioconductor package flowPeaks license Artistic 1.0?

[Ricardo Wurmus]

zimoun <zimon.toutoune@gmail.com> writes:

> This fix -- reuse all the free available code and replace the non-free
> one -- do not scale.
> So the question is: What is the scale we are talking about? How many
> packages in Bioconductor?
> If it is, say, a couple then it is doable. Or see with the people
> managing Bioconductor.
> If it is more, then the option is lobbying. :-)

In my experience tt happens rarely enough that it appears to scale just
fine.  Every time it happens is a drag, of course, but only a small
percentage of CRAN and Bioconductor packages suffers from problems like
that.  That’s still an estimated dozens of packages, but it’s not quite
as bad as it may seem.

--
Ricardo

Guix and Bioconductor.

[Giovanni Biscuolo]

[-- Attachment #1: Type: text/plain, Size: 3305 bytes --]

To all Guix interested in Bioconductor,

forgive me if I raise this question here and not to "upstream", but IMHO
this issue should escalate to Bioconductor and Guix community could do
better than single package maintainers, zimoun in this case

I'm not a user of Bioconductor packages so I have no "weight" on this
matter, but I guess in Guix community there are **many** (potential?)
users of Bioconductor packages: could you please organize a "pressure
group" to convince Bioconductor be strict in their package acceptance
rules?

I fear flowPeacks will not be the last package with this kind licensing
problems

Tobias Geerinckx-Rice <me@tobias.gr> writes:

> Zimoun,

[...]

>> The issue is that upstream has disappeared, as usual in scientific
>> software. Someone writes a piece of code then publishes a paper and
>> sometimes the requirement for publication is to be pushed in
>> mainstream collection of packages (Bioconductor in this case). But
>> the copyright holder does not maintain the code and instead write
>> another piece of code, try to publish a paper, etc.. Well the
>> Reproducibility of Science crisis.
>
> That is a shame.  And that while other scientists (like you) are 
> working hard to make research more ‘open’ and reproducible.

Since «Bioconductor is committed to open source, collaborative,
distributed software development and literate, reproducible research.» [1]

The main point here is that legal aspects are an **integral part** of
reproducible research and the freedom of the developer to choose the
"open source" license he prefer should be _guided_ so he does not
involuntary harm the Bioconductor commitment to reproducible research
(redistribution is part of reproducibility).

In short: since "Clarifies Artistic License" aka "Artistic License
(Perl) 1.0" [2] (FSF approved) exists since long ago *and* since
"Artistic License v1" is not FSF approved, Biocounductor team should not
accept provide a list of accepted licenses that allows all free software
distributions to redistribute the packages (we have one if they wish :-)
)

...and yes, this means that Bioconductor does not accept an OSI approved
"Open Source" license [3], one of the **very few** cases (the only one)
in wich OSI and FSF disagrees on licenses.

To be clear, I'm pretty **sure** that an author generally does not
understand the difference between "Artistic License v1" and "Artistic
License (Perl) 1.0" and all he wants is his package will be freely
redistributable. He should be guided, in this case by Bioconductor.

WDYT?

Thanks. Gio'

[...]

[1] https://bioconductor.org/about/

[2] https://opensource.org/licenses/Artistic-Perl-1.0

    to be clear, the "diff" from Artistic License v1 is this clause
    «8.Aggregation of this Package with a commercial distribution is
    always permitted provided that the use of this Package is embedded;
    that is, when no overt attempt is made to make this Package's
    interfaces visible to the end user of the commercial
    distribution. Such use shall not be construed as a distribution of
    this Package.»

[3] Artistic License v.1 is OSI approved https://opensource.org/licenses/Artistic-1.0

-- 
Giovanni Biscuolo

Xelera IT Infrastructures

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Guix and Bioconductor.

[Ricardo Wurmus]

Giovanni Biscuolo <g@xelera.eu> writes:

> To all Guix interested in Bioconductor,
>
> forgive me if I raise this question here and not to "upstream", but IMHO
> this issue should escalate to Bioconductor and Guix community could do
> better than single package maintainers, zimoun in this case
>
> I'm not a user of Bioconductor packages so I have no "weight" on this
> matter, but I guess in Guix community there are **many** (potential?)
> users of Bioconductor packages: could you please organize a "pressure
> group" to convince Bioconductor be strict in their package acceptance
> rules?
>
> I fear flowPeacks will not be the last package with this kind licensing
> problems

It sure isn’t.  In the past I have tried to do a mass import from
Bioconductor and what slows me down the most is incorrect or non-free
licensing.  There are some packages that declare to be licensed under
Artistic 2.0, but then actually they contain data from databases that
do not permit commercial use.  Or they contain a copy of non-free tools,
or only work when those tools are present (e.g. kent tools, of which we
provide a package containing the few free tools).

It’s a pretty frustrating process to weed out these packages.

> Since «Bioconductor is committed to open source, collaborative,
> distributed software development and literate, reproducible research.» [1]

CRAN appears to be stricter about licenses (even though “strict” is
probably much too strong a word…).  Bioconductor people appear to care a
little less.

--
Ricardo

[OT] Re: Guix and Bioconductor.

[Giovanni Biscuolo]

[-- Attachment #1: Type: text/plain, Size: 2582 bytes --]

[A quick and dirty off topic rant... forgive me!]

Hello Ricardo,

thank you for the info!

...and thank you all for working on resolving this class of problems
through Guix!

Ricardo Wurmus <rekado@elephly.net> writes:

> Giovanni Biscuolo <g@xelera.eu> writes:

[...]

>> I fear flowPeacks will not be the last package with this kind licensing
>> problems
>
> It sure isn’t.  In the past I have tried to do a mass import from
> Bioconductor and what slows me down the most is incorrect or non-free
> licensing.  There are some packages that declare to be licensed under
> Artistic 2.0, but then actually they contain data from databases that
> do not permit commercial use.  Or they contain a copy of non-free tools,
> or only work when those tools are present (e.g. kent tools, of which we
> provide a package containing the few free tools).

This confirms that licensing is an integral part of reproducibility and
replicability, unfortunately a very neglected part even in academia (not
to mention "industry") :-( .  This is also part of the current science
crisis... OK stop ranting :-D

> It’s a pretty frustrating process to weed out these packages.

Let them know! (Do they know?)

«Dear Bioconductor Team, you state you are committed to bla bla
reproducible research but *some* of the research you host is
unreproducible for the simple reason some authors are ignoring licensing
issues...»

>> Since «Bioconductor is committed to open source, collaborative,
>> distributed software development and literate, reproducible research.» [1]
>
> CRAN appears to be stricter about licenses (even though “strict” is
> probably much too strong a word…).  Bioconductor people appear to care a
> little less.

I'm out from academia, but every time I talk to friends involved in
academia I'm pretty astonished by the general lack of scientific method
[1] [2] applied in academia :-O.  A little bit of metascience would help.

...and more Guix in academia is part of the solution :-D


Thanks! Gio'




[1] https://en.wikipedia.org/wiki/Reproducibility#Reproducible_research
«In 2016, Nature conducted a survey of 1,576 researchers who took a
brief online questionnaire on reproducibility in research. According to
the survey, more than 70% of researchers have tried and failed to
reproduce another scientist's experiments, and more than half have
failed to reproduce their own experiments.»

[2] https://en.wikipedia.org/wiki/Replication_crisis


-- 
Giovanni Biscuolo

Xelera IT Infrastructures

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]

Re: Bioconductor package flowPeaks license Artistic 1.0?

[zimoun]

Hi Tobias, again :-)

On Thu, 19 Dec 2019 at 18:18, Tobias Geerinckx-Rice <me@tobias.gr> wrote:

> zimoun 写道：

> > Especially when this
> > very Artistic 1.0 "qualifies as a free software license, but it
> > may
> > not be a real copyleft" [1].
>
> …but that's not this very licence, it's a completely different
> one: the (disjunct) combination of the Artistic 1.0 licence *and
> the GPL*, i.e. ‘choose one’.  The result is only free because you
> can *ignore* the Artistic 1.0 part.


Maybe I misread. And I would like to avoid any confusion.
So I have also read the French translation. Then again the English version.

Where is the License of Perl 5 and below explicitly  defined? There is
no pointer...

What I understand is: when the License of Perl 5 and below is used,
then the copyright holder chooses either the Artistic 1.0, either the
GPL. Then the License of Perl 5 and below is free but non-copyleft.

Well, it appears to me a hack. I guess that there is a lot of Perl
packages under Artistic 1.0 which seems an issue. So let create this
License of Perl 5 and below saying: choose between Artistic 1.0 or
GPL. And because you have this choice, everything is fine.


I probably misread and because it is not Guix related, I would like to
ask to GNU or FSF. Do you know where can I post an email?


All the best,
simon

Perl modules dual licensing (was Re: Bioconductor package flowPeaks license Artistic 1.0?)

[Giovanni Biscuolo]

[-- Attachment #1: Type: text/plain, Size: 2178 bytes --]

Hi zimoun

zimoun <zimon.toutoune@gmail.com> writes:

[...]

> Where is the License of Perl 5 and below explicitly  defined? There is
> no pointer...

Ricardo pointed you to https://dev.perl.org/licenses/, that is a web
version of this 
https://perl5.git.perl.org/perl5.git/blob/HEAD:/README

Perl is dual licensed at least since 1994-10-17	(see the README history
[1]

> What I understand is: when the License of Perl 5

There is no "License of Perl 5", it is Perl 5 that is dual licensed

The same dual license scheme is usually (usually?!?) adopted by Perl
modules, at least those on CPAN
http://www.cpan.org/misc/cpan-faq.html#How_is_Perl_licensed

> and below is used, then the copyright holder chooses either the
> Artistic 1.0, either the GPL.  Then the License of Perl 5 and below is
> free but non-copyleft.

Since there is no "License of Perl 5" that license cannot be qualified
:-)

> Well, it appears to me a hack. I guess that there is a lot of Perl
> packages under Artistic 1.0 which seems an issue.

I don't know how many packages/modules are distributed only using
Artistic License 1.0, but please consider that as I said above that
*many* are dual licensed.

The fact that Perl modules are (must?) commonly dual licensed is
somewhat a mystery to me, but I do not care :-D

> So let create this License of Perl 5 and below saying: choose between
> Artistic 1.0 or GPL. And because you have this choice, everything is
> fine.
>
> I probably misread

No, you do not misread: dual licensing is used in many situation and is
non a "hack", it's the decision of the copyright holder to allow
different legal uses of the software

In this particular case, **fortunately** the dual licensing was adopted
"since the beginning" to fix the problems with Artistic License

[...]

Last things about names: since Oct 2019 [2] Perl 5 is Perl (Perl 4 is
gone long ago) and Perl 6 is Raku, so finally there is no more need to
say "Perl 5" :-)

Ciao! Gio'


[1] https://perl5.git.perl.org/perl5.git/history/HEAD:/README

[2] https://lwn.net/Articles/802329/

-- 
Giovanni Biscuolo

Xelera IT Infrastructures

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]