Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Dejan Muhamedagic]

On Thu, May 05, 2016 at 04:53:57PM +0200, Jan Pokorný wrote:
> On 05/04/16 12:33 +0200, Dejan Muhamedagic wrote:
> > On Wed, Mar 30, 2016 at 05:27:20PM +0200, Jan Pokorný wrote:
> >> On 24/03/16 17:18 +0100, Jan Pokorný wrote:
> >>> On 22/03/16 19:18 +0100, Dejan Muhamedagic wrote:
>  On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
> > On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
> >> So I move to change it to GPLv2+, for everything that is a "program",
> >> and LGPLv2.1 for everything that may be viewed as a library.
> >> 
> >> At least that's how I will correct the wording in the
> >> affected files in the heartbeat mercurial.
> > 
> > In the light of the presented historic excursion, that feels natural.
> > 
> > Assuming no licensors want to speak up, the question now stands:
> > Is it the same conclusion that has been reached by booth and sbd
> > package maintainers (Dejan and Andrew respectively, if I follow what's
> > authoritative nowadays properly) and are these willing to act on it to
> > prevent the mentioned ambiguous interpretation once forever?
>  
>  Yes, that's all fine with me.
>  
> > I will be happy to provide actual patches,
>  
>  Even better :)
> >>> 
> >>> Added the "maint: clarify GPLv2.1+ -> GPLv2+ in the license notices"
> >>> (e294fa2) commit into https://github.com/ClusterLabs/booth/pull/23
> >>> if that's OK with you, Dejan.
> >> 
> >> I hope we are all on the same page as Andrew went ahead there (thanks).
> >> Alas, I've noticed there were some subtleties neglected in there so,
> >> with regrets, a separate (and hopefully final) pull request:
> >> 
> >> https://github.com/ClusterLabs/booth/pull/24
> > 
> > This got merged too. Thanks!
> 
> Neverending story, it seems.  Regrettably, please accept also
> https://github.com/ClusterLabs/booth/pull/33 to call this license
> clarification effort complete, Dejan.

Will take a look.

Cheers,

Dejan

> -- 
> Jan (Poki)



> ___
> Developers mailing list
> Developers@clusterlabs.org
> http://clusterlabs.org/mailman/listinfo/developers


___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 05/04/16 12:33 +0200, Dejan Muhamedagic wrote:
> On Wed, Mar 30, 2016 at 05:27:20PM +0200, Jan Pokorný wrote:
>> On 24/03/16 17:18 +0100, Jan Pokorný wrote:
>>> On 22/03/16 19:18 +0100, Dejan Muhamedagic wrote:
 On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
>> So I move to change it to GPLv2+, for everything that is a "program",
>> and LGPLv2.1 for everything that may be viewed as a library.
>> 
>> At least that's how I will correct the wording in the
>> affected files in the heartbeat mercurial.
> 
> In the light of the presented historic excursion, that feels natural.
> 
> Assuming no licensors want to speak up, the question now stands:
> Is it the same conclusion that has been reached by booth and sbd
> package maintainers (Dejan and Andrew respectively, if I follow what's
> authoritative nowadays properly) and are these willing to act on it to
> prevent the mentioned ambiguous interpretation once forever?
 
 Yes, that's all fine with me.
 
> I will be happy to provide actual patches,
 
 Even better :)
>>> 
>>> Added the "maint: clarify GPLv2.1+ -> GPLv2+ in the license notices"
>>> (e294fa2) commit into https://github.com/ClusterLabs/booth/pull/23
>>> if that's OK with you, Dejan.
>> 
>> I hope we are all on the same page as Andrew went ahead there (thanks).
>> Alas, I've noticed there were some subtleties neglected in there so,
>> with regrets, a separate (and hopefully final) pull request:
>> 
>> https://github.com/ClusterLabs/booth/pull/24
> 
> This got merged too. Thanks!

Neverending story, it seems.  Regrettably, please accept also
https://github.com/ClusterLabs/booth/pull/33 to call this license
clarification effort complete, Dejan.

-- 
Jan (Poki)


pgpNOD1IwzaMT.pgp
Description: PGP signature
___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Dejan Muhamedagic]

On Wed, Mar 30, 2016 at 05:27:20PM +0200, Jan Pokorný wrote:
> On 24/03/16 17:18 +0100, Jan Pokorný wrote:
> > On 22/03/16 19:18 +0100, Dejan Muhamedagic wrote:
> >> On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
> >>> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
>  So I move to change it to GPLv2+, for everything that is a "program",
>  and LGPLv2.1 for everything that may be viewed as a library.
>  
>  At least that's how I will correct the wording in the
>  affected files in the heartbeat mercurial.
> >>> 
> >>> In the light of the presented historic excursion, that feels natural.
> >>> 
> >>> Assuming no licensors want to speak up, the question now stands:
> >>> Is it the same conclusion that has been reached by booth and sbd
> >>> package maintainers (Dejan and Andrew respectively, if I follow what's
> >>> authoritative nowadays properly) and are these willing to act on it to
> >>> prevent the mentioned ambiguous interpretation once forever?
> >> 
> >> Yes, that's all fine with me.
> >> 
> >>> I will be happy to provide actual patches,
> >> 
> >> Even better :)
> > 
> > Added the "maint: clarify GPLv2.1+ -> GPLv2+ in the license notices"
> > (e294fa2) commit into https://github.com/ClusterLabs/booth/pull/23
> > if that's OK with you, Dejan.
> 
> I hope we are all on the same page as Andrew went ahead there (thanks).
> Alas, I've noticed there were some subtleties neglected in there so,
> with regrets, a separate (and hopefully final) pull request:
> 
> https://github.com/ClusterLabs/booth/pull/24

This got merged too. Thanks!

Dejan

> -- 
> Jan (Poki)



> ___
> Developers mailing list
> Developers@clusterlabs.org
> http://clusterlabs.org/mailman/listinfo/developers


___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 21/03/16 22:03 +0100, Jan Pokorný wrote:
> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
>> So I move to change it to GPLv2+, for everything that is a "program",
>> and LGPLv2.1 for everything that may be viewed as a library.
>> 
>> At least that's how I will correct the wording in the
>> affected files in the heartbeat mercurial.
> 
> In the light of the presented historic excursion, that feels natural.
> 
> Assuming no licensors want to speak up, the question now stands:
> Is it the same conclusion that has been reached by booth and sbd
> package maintainers (Dejan and Andrew respectively, if I follow what's
> authoritative nowadays properly) and are these willing to act on it to
> prevent the mentioned ambiguous interpretation once forever?
> 
> I will be happy to provide actual patches, but that's not a decision
> for me to take.
> 
> In fact, Andrew has already expressed his inclination:
> http://oss.clusterlabs.org/pipermail/developers/2016-March/000182.html

Here we go to address this in sbd:
https://github.com/ClusterLabs/sbd/pull/3

-- 
Jan (Poki)


pgpKN0x64mRXU.pgp
Description: PGP signature
___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 24/03/16 17:18 +0100, Jan Pokorný wrote:
> On 22/03/16 19:18 +0100, Dejan Muhamedagic wrote:
>> On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
>>> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
 So I move to change it to GPLv2+, for everything that is a "program",
 and LGPLv2.1 for everything that may be viewed as a library.
 
 At least that's how I will correct the wording in the
 affected files in the heartbeat mercurial.
>>> 
>>> In the light of the presented historic excursion, that feels natural.
>>> 
>>> Assuming no licensors want to speak up, the question now stands:
>>> Is it the same conclusion that has been reached by booth and sbd
>>> package maintainers (Dejan and Andrew respectively, if I follow what's
>>> authoritative nowadays properly) and are these willing to act on it to
>>> prevent the mentioned ambiguous interpretation once forever?
>> 
>> Yes, that's all fine with me.
>> 
>>> I will be happy to provide actual patches,
>> 
>> Even better :)
> 
> Added the "maint: clarify GPLv2.1+ -> GPLv2+ in the license notices"
> (e294fa2) commit into https://github.com/ClusterLabs/booth/pull/23
> if that's OK with you, Dejan.

I hope we are all on the same page as Andrew went ahead there (thanks).
Alas, I've noticed there were some subtleties neglected in there so,
with regrets, a separate (and hopefully final) pull request:

https://github.com/ClusterLabs/booth/pull/24

-- 
Jan (Poki)


pgpvGDEwkecNJ.pgp
Description: PGP signature
___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 22/03/16 19:18 +0100, Dejan Muhamedagic wrote:
> On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
>> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
>>> So I move to change it to GPLv2+, for everything that is a "program",
>>> and LGPLv2.1 for everything that may be viewed as a library.
>>> 
>>> At least that's how I will correct the wording in the
>>> affected files in the heartbeat mercurial.
>> 
>> In the light of the presented historic excursion, that feels natural.
>> 
>> Assuming no licensors want to speak up, the question now stands:
>> Is it the same conclusion that has been reached by booth and sbd
>> package maintainers (Dejan and Andrew respectively, if I follow what's
>> authoritative nowadays properly) and are these willing to act on it to
>> prevent the mentioned ambiguous interpretation once forever?
> 
> Yes, that's all fine with me.
> 
>> I will be happy to provide actual patches,
> 
> Even better :)

Added the "maint: clarify GPLv2.1+ -> GPLv2+ in the license notices"
(e294fa2) commit into https://github.com/ClusterLabs/booth/pull/23
if that's OK with you, Dejan.

-- 
Jan (Poki)


pgpUkyq9jrOa3.pgp
Description: PGP signature
___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Dejan Muhamedagic]

On Mon, Mar 21, 2016 at 10:03:12PM +0100, Jan Pokorný wrote:
> On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
> > On Thu, Mar 17, 2016 at 07:12:19PM +0100, Dejan Muhamedagic wrote:
> >> Hi Jan,
> >> 
> >> On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
> >>> Hello all,
> >>> 
> >>> during latest reviews of packages building on core cluster
> >>> infrastructure, it turned out there is a frequented (viral?) issue
> >>> with source files declaring unusual licence: GPLv2.1+.
> >> 
> >> Yes, I'd say that it's all coming from a single source. I suspect
> >> that nobody's looking at the license, just copies another source
> >> file from the same project. Anyway, that's what I did in booth.
> >> 
> >> Who created the first file with this non-existent license is
> >> anybody's guess. It could probably be traced, but I doubt that
> >> it'd help in any way.
> > 
> > Actually it might.
> > 
> > I think that what happened was this:
> 
> Lars, many thanks for providing the following historical context.
> It seems to provide enough background to guide the interpretation
> of dubious license notices towards the original intentions.
> 
> Let me premise that from the new perspective, it might have been
> a bit preposterous trying to contact every and each (potential)
> licensor of the affected files (got a couple of bounces[*], anyway),
> just followed the transparency principle, turning them to stand-by
> should any coordination be required...
> I am sorry if you were looped in without solicitation, especially if
> neither project is of your interest anymore.
> 
> > in the early days of heartbeat, way back when,
> > source code got "batch tagged" with the license statement:
> > http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/4a67fde00b0b#l1.10
> > 2000/07/26 05:17:18
> > 
> > Most stuff got tagged with the LGPL 2.1.
> > 
> > Some time later, someone noticed that in some cases,
> > a "program" is not a "library", and tried to re-tag
> > e.g. "api_test.c" with the GPL 2,
> > but without properly taking the actual suggested GPL 2 stanza,
> > but by simply dropping "Lesser" and changing "library" to "software".
> > http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/bc508513c4dc#l2.10
> > 2000/08/31 05:23:36
> > 
> >  :-(
> 
> I was made aware of this is downright supported by LGPL 2.1 provision:
> 
>   3. You may opt to apply the terms of the ordinary GNU General Public
>   License instead of this License to a given copy of the Library. To do
>   this, you must alter all the notices that refer to this License, so
>   that they refer to the ordinary GNU General Public License, version
>   2, instead of to this License. (If a newer version than version 2 of
>   the ordinary GNU General Public License has appeared, then you can
>   specify that version instead if you wish.) Do not make any other
>   change in these notices.
> 
>   Once this change is made in a given copy, it is irreversible for that
>   copy, so the ordinary GNU General Public License applies to all
>   subsequent copies and derivative works made from that copy.
> 
>   This option is useful when you wish to copy part of the code of the
>   Library into a program that is not a library. 
> 
> So one might think this is just it's imperfect execution.
> 
> > Both changes predate the GPLv3 by seven years.
> 
> Hence the timeless, causation-ignorant, plain logic implication of GPLv3+
> cannot simply hold under the demonstrated reality-check.
> 
> > From there it propagated to ipfail.c and attrd.c, which both became
> > *the* template files to start from when writing daemons and extensions
> > using the API.
> > 
> > Developers quickly browse their "template",
> > their "auto-correct" filter reads "GPL 2",
> > which they are content with,
> > and in good faith they hack away.
> > 
> > I think it is safe to assume that any developer copying from there meant
> > to "stay in project" regarding the licensing.
> 
> Agreed on good will/no surprises being presumed in free software
> projects, leading to code being anything that really matters in such
> circles.  Licensing (and other) paper work is frequently played down
> to cargo cult minimum[**].  Unfortunately, a minor inaccuracy there
> can go 15+ years down the road unnoticed, as we can observe right now :-)
> 
> > So I move to change it to GPLv2+, for everything that is a "program",
> > and LGPLv2.1 for everything that may be viewed as a library.
> > 
> > At least that's how I will correct the wording in the
> > affected files in the heartbeat mercurial.
> 
> In the light of the presented historic excursion, that feels natural.
> 
> Assuming no licensors want to speak up, the question now stands:
> Is it the same conclusion that has been reached by booth and sbd
> package maintainers (Dejan and Andrew respectively, if I follow what's
> authoritative nowadays properly) and are these willing to act on it to
> prevent the mentioned ambiguous interpretation once forever?

Yes, that's all fine with me.

> I will be 

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 18/03/16 16:16 +0100, Lars Ellenberg wrote:
> On Thu, Mar 17, 2016 at 07:12:19PM +0100, Dejan Muhamedagic wrote:
>> Hi Jan,
>> 
>> On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
>>> Hello all,
>>> 
>>> during latest reviews of packages building on core cluster
>>> infrastructure, it turned out there is a frequented (viral?) issue
>>> with source files declaring unusual licence: GPLv2.1+.
>> 
>> Yes, I'd say that it's all coming from a single source. I suspect
>> that nobody's looking at the license, just copies another source
>> file from the same project. Anyway, that's what I did in booth.
>> 
>> Who created the first file with this non-existent license is
>> anybody's guess. It could probably be traced, but I doubt that
>> it'd help in any way.
> 
> Actually it might.
> 
> I think that what happened was this:

Lars, many thanks for providing the following historical context.
It seems to provide enough background to guide the interpretation
of dubious license notices towards the original intentions.

Let me premise that from the new perspective, it might have been
a bit preposterous trying to contact every and each (potential)
licensor of the affected files (got a couple of bounces[*], anyway),
just followed the transparency principle, turning them to stand-by
should any coordination be required...
I am sorry if you were looped in without solicitation, especially if
neither project is of your interest anymore.

> in the early days of heartbeat, way back when,
> source code got "batch tagged" with the license statement:
> http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/4a67fde00b0b#l1.10
> 2000/07/26 05:17:18
> 
> Most stuff got tagged with the LGPL 2.1.
> 
> Some time later, someone noticed that in some cases,
> a "program" is not a "library", and tried to re-tag
> e.g. "api_test.c" with the GPL 2,
> but without properly taking the actual suggested GPL 2 stanza,
> but by simply dropping "Lesser" and changing "library" to "software".
> http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/bc508513c4dc#l2.10
> 2000/08/31 05:23:36
> 
>  :-(

I was made aware of this is downright supported by LGPL 2.1 provision:

  3. You may opt to apply the terms of the ordinary GNU General Public
  License instead of this License to a given copy of the Library. To do
  this, you must alter all the notices that refer to this License, so
  that they refer to the ordinary GNU General Public License, version
  2, instead of to this License. (If a newer version than version 2 of
  the ordinary GNU General Public License has appeared, then you can
  specify that version instead if you wish.) Do not make any other
  change in these notices.

  Once this change is made in a given copy, it is irreversible for that
  copy, so the ordinary GNU General Public License applies to all
  subsequent copies and derivative works made from that copy.

  This option is useful when you wish to copy part of the code of the
  Library into a program that is not a library. 

So one might think this is just it's imperfect execution.

> Both changes predate the GPLv3 by seven years.

Hence the timeless, causation-ignorant, plain logic implication of GPLv3+
cannot simply hold under the demonstrated reality-check.

> From there it propagated to ipfail.c and attrd.c, which both became
> *the* template files to start from when writing daemons and extensions
> using the API.
> 
> Developers quickly browse their "template",
> their "auto-correct" filter reads "GPL 2",
> which they are content with,
> and in good faith they hack away.
> 
> I think it is safe to assume that any developer copying from there meant
> to "stay in project" regarding the licensing.

Agreed on good will/no surprises being presumed in free software
projects, leading to code being anything that really matters in such
circles.  Licensing (and other) paper work is frequently played down
to cargo cult minimum[**].  Unfortunately, a minor inaccuracy there
can go 15+ years down the road unnoticed, as we can observe right now :-)

> So I move to change it to GPLv2+, for everything that is a "program",
> and LGPLv2.1 for everything that may be viewed as a library.
> 
> At least that's how I will correct the wording in the
> affected files in the heartbeat mercurial.

In the light of the presented historic excursion, that feels natural.

Assuming no licensors want to speak up, the question now stands:
Is it the same conclusion that has been reached by booth and sbd
package maintainers (Dejan and Andrew respectively, if I follow what's
authoritative nowadays properly) and are these willing to act on it to
prevent the mentioned ambiguous interpretation once forever?

I will be happy to provide actual patches, but that's not a decision
for me to take.

In fact, Andrew has already expressed his inclination:
http://oss.clusterlabs.org/pipermail/developers/2016-March/000182.html


Thanks for taking this seriously.


[*] see the difference of CC field between original and this message
 

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Jan Pokorný]

On 17/03/16 14:34 -0500, Ken Gaillot wrote:
> On 03/17/2016 12:47 PM, Jan Pokorný wrote:
>> Affected packages breakdown is at the bottom, including
>> (possibly non-exhaustive) contributors that knowingly or unknowingly
>> contributed under that file-local license.
> 
> Since this is merely a clarification of the intended license, and not a
> change to a different license, I believe it should be sufficient to ask
> the intent of the original package author without having to contact
> every file author/contributor.

1. IANAL, I prefer to err on the safe side

2. contacting merely the contributors to the affected files
   was attempted

-- 
Jan (Poki)


pgpQfu0KmIxL5.pgp
Description: PGP signature
___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Lars Ellenberg]

On Thu, Mar 17, 2016 at 07:12:19PM +0100, Dejan Muhamedagic wrote:
> Hi Jan,
> 
> On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
> > Hello all,
> > 
> > during latest reviews of packages building on core cluster
> > infrastructure, it turned out there is a frequented (viral?) issue
> > with source files declaring unusual licence: GPLv2.1+.
> 
> Yes, I'd say that it's all coming from a single source. I suspect
> that nobody's looking at the license, just copies another source
> file from the same project. Anyway, that's what I did in booth.
> 
> Who created the first file with this non-existent license is
> anybody's guess. It could probably be traced, but I doubt that
> it'd help in any way.

Actually it might.

I think that what happened was this:

in the early days of heartbeat, way back when,
source code got "batch tagged" with the license statement:
http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/4a67fde00b0b#l1.10
2000/07/26 05:17:18

Most stuff got tagged with the LGPL 2.1.

Some time later, someone noticed that in some cases,
a "program" is not a "library", and tried to re-tag
e.g. "api_test.c" with the GPL 2,
but without properly taking the actual suggested GPL 2 stanza,
but by simply dropping "Lesser" and changing "library" to "software".
http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/bc508513c4dc#l2.10
2000/08/31 05:23:36

 :-(

Both changes predate the GPLv3 by seven years.

>From there it propagated to ipfail.c and attrd.c, which both became
*the* template files to start from when writing daemons and extensions
using the API.

Developers quickly browse their "template",
their "auto-correct" filter reads "GPL 2",
which they are content with,
and in good faith they hack away.

I think it is safe to assume that any developer copying from there meant
to "stay in project" regarding the licensing.

So I move to change it to GPLv2+, for everything that is a "program",
and LGPLv2.1 for everything that may be viewed as a libraray.

At least that's how I will correct the wording in the
affected files in the heartbeat mercurial.

-- 
: Lars Ellenberg
: LINBIT | Keeping the Digital World Running
: DRBD -- Heartbeat -- Corosync -- Pacemaker
: R&D, Integration, Ops, Consulting, Support

DRBD® and LINBIT® are registered trademarks of LINBIT

___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Dejan Muhamedagic]

Hi Jan,

On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
> Hello all,
> 
> during latest reviews of packages building on core cluster
> infrastructure, it turned out there is a frequented (viral?) issue
> with source files declaring unusual licence: GPLv2.1+.

Yes, I'd say that it's all coming from a single source. I suspect
that nobody's looking at the license, just copies another source
file from the same project. Anyway, that's what I did in booth.

Who created the first file with this non-existent license is
anybody's guess. It could probably be traced, but I doubt that
it'd help in any way.

> Affected packages breakdown is at the bottom, including
> (possibly non-exhaustive) contributors that knowingly or unknowingly
> contributed under that file-local license.
> 
> Now, it is highly questionable what was asserted by this license
> reference by particular contributors.  The fact is that such a license
> does not exist.  So the logical implication and view of the situation
> is that affected files are effectively licensed under GPLv3+.
> 
> The other possible view is that it's actually a typo arising from
> LGPL2.1 vs. GPLv2 confusion, and then again it's unclear which one
> should apply.

As you raised this issue earlier in one github pull request, I
looked today into the matter and reached the same conclusion
(which is not much of a conclusion).

> Therefore I would like to start moving towards resolution of this
> issue by soliciting feedback amongst affected contributors (CC'd)
> which direction is preferred:
> 
> 1. settle down on GPLv2+ (or LGPL2.1+), which will likely require
>whole relicensing process, i.e., collecting sign-offs on this by all
>contributors in question
> 
> 2. clarify that indeed GPLv3+ was meant and adjust the respective
>clauses in the source code
> 
> 3. keep the status quo, leave it in the shadow zone allowing for
>ambiguous interpretations that may leave potential contributors
>away from the project(s)
> 
> Apparently, both affected packages, booth and sbd, can choose its
> own direction independently of the other package.

I don't have an opinion on which resolution to pursue and I'm not
sure if I care at all.

Thanks,

Dejan

> * * *
> 
> Affected packages breakdown
> 
> package (considered repo, branch):
> - file: GPLv2.1+ source
>   (rest of contributors)
> 
> 
> booth (https://github.com/ClusterLabs/booth, master):
> - src/attr.c: Dejan Muhamedagic
>   ()
> - src/attr.h: Dejan Muhamedagic
>   ()
> - src/auth.c: Dejan Muhamedagic
>   ()
> - src/auth.h: Dejan Muhamedagic
>   ()
> - src/booth.h: Jiaju Zhang
>   (Adam Spiers, Dejan Muhamedagic, Dongmao Zhang, Philipp Marek)
> - src/config.c: Jiaju Zhang
>   (Adam Spiers, Daniel Gollub, Dejan Muhamedagic, Kazunori INOUE,
>   Philipp Marek, Yuichi SEINO, Yuusuke Iida)
> - src/config.h: Jiaju Zhang
>   (Adam Spiers, Dejan Muhamedagic, Philipp Marek)
> - src/handler.c: Philipp Marek
>   (Dejan Muhamedagic)
> - src/handler.h: Philipp Marek
>   (Dejan Muhamedagic)
> - src/inline-fn.h: Philipp Marek
>   (Dejan Muhamedagic)
> - src/log.h: Jiaju Zhang
>   (Dejan Muhamedagic, Philipp Marek)
> - src/main.c: Jiaju Zhang
>   (Adam Spiers, Daniel Gollub, Dejan Muhamedagic, Dongmao Zhang,
>   Guangliang Zhao, Philipp Marek, Xia Li, Yuich SEINO, Yuusuke Iida)
> - src/pacemaker.c: Jiaju Zhang
>   (Daniel Gollub, Dejan Muhamedagic, Philipp Marek)
> - src/pacemaker.h: Jiaju Zhang
>   (Dejan Muhamedagic, Philipp Marek)
> - src/raft.c: Philipp Marek
>   (Dejan Muhamedagic)
> - src/raft.h: Philipp Marek 
>   (Dejan Muhamedagic)
> - src/request.c: Dejan Muhamedagic
>   ()
> - src/request.h: Dejan Muhamedagic
>   ()
> - src/ticket.c: Jiaju Zhang
>   (Adam Spiers, Dejan Muhamedagic, Dongmao Zhang, Guangliang Zhao,
>   Philipp Marek, Yuusuke Iida)
> - src/ticket.h: Jiaju Zhang
>   (Adam Spiers, Dejan Muhamedagic, Dongmao Zhang, Philipp Marek)
> - src/timer.c: Dejan Muhamedagic
>   (Jiaju Zhang, Philipp Marek)
> - src/timer.h: Dejan Muhamedagic
>   (Jiaju Zhang, Philipp Marek)
> - src/transport.c: Jiaju Zhang
>   (Adam Spiers, Dejan Muhamedagic, Dongmao Zhang, Philipp Marek,
>Yuich SEINO)
> - src/transport.h: Jiaju Zhang
>   (Dejan Muhamedagic, Philipp Marek)
> (see also
> https://github.com/ClusterLabs/booth/pull/23#issuecomment-193957362)
> 
> 
> sbd (https://github.com/ClusterLabs/sbd, master):
> - agent/sbd: Lars Marowsky-Bree
>   ()
> - man/sbd.8.pod: Lars Marowsky-Bree
>   ()
> - src/sbd-common.c: Lars Marowsky-Bree
>   (Andrew Beekhof)
> - src/sbd.h: Lars Marowsky-Bree
>   (aaronknister, Andrew Beekhof)
> - src/sbd-inquisitor.c: Andrew Beekhof
>   (aaronknister, Lars Ellenberg)
> - src/sbd-md.c: xwhu
>   (Andrew Beekhof, Lars Marowsky-Bree)
> - src/sbd.sh: Lars Marowsky-Bree
>   ()
> - tests/regressions.sh: Lars Marowsky-Bree
>   ()
> 
> -- 
> Jan (Poki)



___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Oyvind Albrigtsen]

On 17/03/16 22:05, Andrew Beekhof wrote:


On 18 Mar 2016, at 5:23 AM, Philipp Marek > wrote:


On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:

Hello all,

during latest reviews of packages building on core cluster
infrastructure, it turned out there is a frequented (viral?) issue
with source files declaring unusual licence: GPLv2.1+.

Affected packages breakdown is at the bottom, including
(possibly non-exhaustive) contributors that knowingly or unknowingly
contributed under that file-local license.

Now, it is highly questionable what was asserted by this license
reference by particular contributors.  The fact is that such a license
does not exist.  So the logical implication and view of the situation
is that affected files are effectively licensed under GPLv3+.

The other possible view is that it's actually a typo arising from
LGPL2.1 vs. GPLv2 confusion, and then again it's unclear which one
should apply.

Therefore I would like to start moving towards resolution of this
issue by soliciting feedback amongst affected contributors (CC'd)
which direction is preferred:

1. settle down on GPLv2+ (or LGPL2.1+), which will likely require
  whole relicensing process, i.e., collecting sign-offs on this by all
  contributors in question

2. clarify that indeed GPLv3+ was meant and adjust the respective
  clauses in the source code

3. keep the status quo, leave it in the shadow zone allowing for
  ambiguous interpretations that may leave potential contributors
  away from the project(s)

Apparently, both affected packages, booth and sbd, can choose its
own direction independently of the other package.

I'm okay with both, with a slight preference to 2 (moving to GPLv3+).


Please no, anything but v3

IMO the GPLv2.1+ suggests that parts of the code could be e.g. GPLv3 
depending on what the author of specific parts of the code decides, so I 
think the question is should it be GPLv2 or LGPL2.1.


My (brief) look at the header lines via "git blame" says

   commit 54dc3dc2be33cd977866ce63d8cc74934d21405d
   Author: Jiaju Zhang mailto:jjzh...@suse.de>>
   Date:   Wed Aug 24 12:07:22 2011 +0800

   Initial check-in of the Booth Cluster Ticket Manager.

so it predates my contributions by quite some time - I'm not sure how 
much

I've got to say in this matter ;)


Regards,

Phil

--
Philipp Marek
LINBIT | Keeping the Digital World Running
DRBD - Heartbeat - Corosync - Pacemaker

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.




___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers


___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Philipp Marek]

On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
> Hello all,
> 
> during latest reviews of packages building on core cluster
> infrastructure, it turned out there is a frequented (viral?) issue
> with source files declaring unusual licence: GPLv2.1+.
> 
> Affected packages breakdown is at the bottom, including
> (possibly non-exhaustive) contributors that knowingly or unknowingly
> contributed under that file-local license.
> 
> Now, it is highly questionable what was asserted by this license
> reference by particular contributors.  The fact is that such a license
> does not exist.  So the logical implication and view of the situation
> is that affected files are effectively licensed under GPLv3+.
> 
> The other possible view is that it's actually a typo arising from
> LGPL2.1 vs. GPLv2 confusion, and then again it's unclear which one
> should apply.
> 
> Therefore I would like to start moving towards resolution of this
> issue by soliciting feedback amongst affected contributors (CC'd)
> which direction is preferred:
> 
> 1. settle down on GPLv2+ (or LGPL2.1+), which will likely require
>whole relicensing process, i.e., collecting sign-offs on this by all
>contributors in question
> 
> 2. clarify that indeed GPLv3+ was meant and adjust the respective
>clauses in the source code
> 
> 3. keep the status quo, leave it in the shadow zone allowing for
>ambiguous interpretations that may leave potential contributors
>away from the project(s)
> 
> Apparently, both affected packages, booth and sbd, can choose its
> own direction independently of the other package.
I'm okay with both, with a slight preference to 2 (moving to GPLv3+).

My (brief) look at the header lines via "git blame" says

commit 54dc3dc2be33cd977866ce63d8cc74934d21405d
Author: Jiaju Zhang 
Date:   Wed Aug 24 12:07:22 2011 +0800

Initial check-in of the Booth Cluster Ticket Manager.

so it predates my contributions by quite some time - I'm not sure how much 
I've got to say in this matter ;)


Regards,

Phil

-- 
Philipp Marek
LINBIT | Keeping the Digital World Running
DRBD - Heartbeat - Corosync - Pacemaker

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.

___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Andrew Beekhof]

> On 18 Mar 2016, at 5:23 AM, Philipp Marek  wrote:
> 
> On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
>> Hello all,
>> 
>> during latest reviews of packages building on core cluster
>> infrastructure, it turned out there is a frequented (viral?) issue
>> with source files declaring unusual licence: GPLv2.1+.
>> 
>> Affected packages breakdown is at the bottom, including
>> (possibly non-exhaustive) contributors that knowingly or unknowingly
>> contributed under that file-local license.
>> 
>> Now, it is highly questionable what was asserted by this license
>> reference by particular contributors.  The fact is that such a license
>> does not exist.  So the logical implication and view of the situation
>> is that affected files are effectively licensed under GPLv3+.
>> 
>> The other possible view is that it's actually a typo arising from
>> LGPL2.1 vs. GPLv2 confusion, and then again it's unclear which one
>> should apply.
>> 
>> Therefore I would like to start moving towards resolution of this
>> issue by soliciting feedback amongst affected contributors (CC'd)
>> which direction is preferred:
>> 
>> 1. settle down on GPLv2+ (or LGPL2.1+), which will likely require
>>   whole relicensing process, i.e., collecting sign-offs on this by all
>>   contributors in question
>> 
>> 2. clarify that indeed GPLv3+ was meant and adjust the respective
>>   clauses in the source code
>> 
>> 3. keep the status quo, leave it in the shadow zone allowing for
>>   ambiguous interpretations that may leave potential contributors
>>   away from the project(s)
>> 
>> Apparently, both affected packages, booth and sbd, can choose its
>> own direction independently of the other package.
> I'm okay with both, with a slight preference to 2 (moving to GPLv3+).

Please no, anything but v3

> 
> My (brief) look at the header lines via "git blame" says
> 
>commit 54dc3dc2be33cd977866ce63d8cc74934d21405d
>Author: Jiaju Zhang mailto:jjzh...@suse.de>>
>Date:   Wed Aug 24 12:07:22 2011 +0800
> 
>Initial check-in of the Booth Cluster Ticket Manager.
> 
> so it predates my contributions by quite some time - I'm not sure how much 
> I've got to say in this matter ;)
> 
> 
> Regards,
> 
> Phil
> 
> -- 
> Philipp Marek
> LINBIT | Keeping the Digital World Running
> DRBD - Heartbeat - Corosync - Pacemaker
> 
> DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.

___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Ken Gaillot]

On 03/17/2016 12:47 PM, Jan Pokorný wrote:
> Affected packages breakdown is at the bottom, including
> (possibly non-exhaustive) contributors that knowingly or unknowingly
> contributed under that file-local license.

Since this is merely a clarification of the intended license, and not a
change to a different license, I believe it should be sufficient to ask
the intent of the original package author without having to contact
every file author/contributor.


___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers

Re: [ClusterLabs Developers] [booth][sbd] GPLv2.1+ clarification request

[Andrew Beekhof]

Agree completely 

Sent from a mobile device 

> On 19 Mar 2016, at 2:16 AM, Lars Ellenberg  wrote:
> 
>> On Thu, Mar 17, 2016 at 07:12:19PM +0100, Dejan Muhamedagic wrote:
>> Hi Jan,
>> 
>>> On Thu, Mar 17, 2016 at 06:47:37PM +0100, Jan Pokorný wrote:
>>> Hello all,
>>> 
>>> during latest reviews of packages building on core cluster
>>> infrastructure, it turned out there is a frequented (viral?) issue
>>> with source files declaring unusual licence: GPLv2.1+.
>> 
>> Yes, I'd say that it's all coming from a single source. I suspect
>> that nobody's looking at the license, just copies another source
>> file from the same project. Anyway, that's what I did in booth.
>> 
>> Who created the first file with this non-existent license is
>> anybody's guess. It could probably be traced, but I doubt that
>> it'd help in any way.
> 
> Actually it might.
> 
> I think that what happened was this:
> 
> in the early days of heartbeat, way back when,
> source code got "batch tagged" with the license statement:
> http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/4a67fde00b0b#l1.10
> 2000/07/26 05:17:18
> 
> Most stuff got tagged with the LGPL 2.1.
> 
> Some time later, someone noticed that in some cases,
> a "program" is not a "library", and tried to re-tag
> e.g. "api_test.c" with the GPL 2,
> but without properly taking the actual suggested GPL 2 stanza,
> but by simply dropping "Lesser" and changing "library" to "software".
> http://hg.linux-ha.org/heartbeat-STABLE_3_0/rev/bc508513c4dc#l2.10
> 2000/08/31 05:23:36
> 
> :-(
> 
> Both changes predate the GPLv3 by seven years.
> 
> From there it propagated to ipfail.c and attrd.c, which both became
> *the* template files to start from when writing daemons and extensions
> using the API.
> 
> Developers quickly browse their "template",
> their "auto-correct" filter reads "GPL 2",
> which they are content with,
> and in good faith they hack away.
> 
> I think it is safe to assume that any developer copying from there meant
> to "stay in project" regarding the licensing.
> 
> So I move to change it to GPLv2+, for everything that is a "program",
> and LGPLv2.1 for everything that may be viewed as a libraray.
> 
> At least that's how I will correct the wording in the
> affected files in the heartbeat mercurial.
> 
> -- 
> : Lars Ellenberg
> : LINBIT | Keeping the Digital World Running
> : DRBD -- Heartbeat -- Corosync -- Pacemaker
> : R&D, Integration, Ops, Consulting, Support
> 
> DRBD® and LINBIT® are registered trademarks of LINBIT
> 
> ___
> Developers mailing list
> Developers@clusterlabs.org
> http://clusterlabs.org/mailman/listinfo/developers

___
Developers mailing list
Developers@clusterlabs.org
http://clusterlabs.org/mailman/listinfo/developers