Discussion:
VAFS, GFS2 and the GPL, was: Re: VSI OpenVMS Roadmap: V9.2 is x86-64 only
(too old to reply)
Simon Clubley
2020-09-24 12:59:11 UTC
Permalink
VSI need a new file-system for VMS.
Yes
Which was going to be provided by the VAFS work VSI were doing and
there have been briefings about VAFS in the past.

What is below is a major surprise, at least to me.
VSI figured out that it was much lower cost to lift an existing
open source file-system than creating one themselves from scratch.
Yes
Ok, this is a major surprise, given the previous announcements about
VAFS. When I wrote my comments, I thought GFS2 was going to be something
in addition to VAFS and was added for some customer-specific purpose.

Does this mean the VAFS project is now dead ?
GSF2 seems like a good fit for VMS.
Yes
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any mention
of any linking exceptions or similar.

How do you reconcile the GPL licence for GFS2 with VMS if you are going
to integrate GFS2 into the VMS kernel ?

What mechanisms are you using to stop the VMS kernel code from falling
under the GPL licence if you are integrating GFS2 into VMS kernel code ?

As far as I can see, this is a similar problem to trying to use GNU Readline
in a closed source program as GNU readline is pure GPL.

The last paragraph of the introduction here:

https://tiswww.case.edu/php/chet/readline/rltop.html

talks about the conditions for using Readline in a program.

Is there some linking exception or similar for GFS2 that I have missed
or have VSI come up with a way to integrate the GFS2 code into VMS
without causing at least of some of the existing VMS code to now be
covered by the GPL ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-09-24 17:13:42 UTC
Permalink
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any mention
of any linking exceptions or similar.
How do you reconcile the GPL licence for GFS2 with VMS if you are going
to integrate GFS2 into the VMS kernel ?
What mechanisms are you using to stop the VMS kernel code from falling
under the GPL licence if you are integrating GFS2 into VMS kernel code ?
Is there some linking exception or similar for GFS2 that I have missed
or have VSI come up with a way to integrate the GFS2 code into VMS
without causing at least of some of the existing VMS code to now be
covered by the GPL ?
Interesting question.

I hope VSI thought about that.

I believe they can use two different approaches to work
around this:
* put in a layer between VMS and GFS2 - the "NVidia model"
* claim it is program and whatever parts of VMS it needs
fall under the system library exception in GPL

GPL V2:

However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.

GPL V3:

The “System Libraries” of an executable work include anything, other
than the work as a whole, that (a) is included in the normal form of
packaging a Major Component, but which is not part of that Major
Component, and (b) serves only to enable use of the work with that Major
Component, or to implement a Standard Interface for which an
implementation is available to the public in source code form. A “Major
Component”, in this context, means a major essential component (kernel,
window system, and so on) of the specific operating system (if any) on
which the executable work runs, or a compiler used to produce the work,
or an object code interpreter used to run it.

The “Corresponding Source” for a work in object code form means all the
source code needed to generate, install, and (for an executable work)
run the object code and to modify the work, including scripts to control
those activities. However, it does not include the work's System
Libraries, or general-purpose tools or generally available free programs
which are used unmodified in performing those activities but which are
not part of the work. For example, Corresponding Source includes
interface definition files associated with source files for the work,
and the source code for shared libraries and dynamically linked
subprograms that the work is specifically designed to require, such as
by intimate data communication or control flow between those subprograms
and other parts of the work.

Arne
Simon Clubley
2020-09-28 12:43:03 UTC
Permalink
Post by Arne Vajhøj
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any mention
of any linking exceptions or similar.
How do you reconcile the GPL licence for GFS2 with VMS if you are going
to integrate GFS2 into the VMS kernel ?
What mechanisms are you using to stop the VMS kernel code from falling
under the GPL licence if you are integrating GFS2 into VMS kernel code ?
Is there some linking exception or similar for GFS2 that I have missed
or have VSI come up with a way to integrate the GFS2 code into VMS
without causing at least of some of the existing VMS code to now be
covered by the GPL ?
Interesting question.
I hope VSI thought about that.
I've had another look at this and I don't immediately see how the options
below are viable ways to get GPL code into the VMS kernel without causing
at least some of the existing VMS kernel code to now fall until the GPL.
Post by Arne Vajhøj
I believe they can use two different approaches to work
* put in a layer between VMS and GFS2 - the "NVidia model"
The problem with this one is that the situation is reversed from the
Nvidia case. In the Nvidia case, Nvidia themselves own the code that
they want to stop from becoming GPL when it is inserted into a GPL kernel.

In the GFS2 case, VSI want to use pure GPL V2 code that they do not own
within a non-GPL kernel (the VMS kernel). This needs to happen in a way
that does not make the rest of the VMS kernel now fall until the GPL as
a result.

If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Post by Arne Vajhøj
* claim it is program and whatever parts of VMS it needs
fall under the system library exception in GPL
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.
This one doesn't work because of the last bit. The VMS kernel will
accompany the GFS2 executables as part of the installation media
and GFS2 will be tied into the VMS kernel itself.

The example I keep coming back to is that GNU Readline (which is pure GPL
and not LGPL) makes any program which uses it fall under a GPL-compatible
licence.

Can someone here from VSI confirm the following with regards to GFS2:

1) that they have realised the viral implications of the pure GPL licence
in that code bound with a pure GPL routine requires that code to also be
released under a GPL-compatible licence ?

2) That the GPL is _NOT_ the LGPL ? The LGPL is what you use when you want
to allow your open source library to be used in a wider range of code,
not just code that is licenced under a GPL-compatible licence.

However, the GFS2 source code appears to be licenced under a pure GPL licence.

3) If VSI have come up with a clever way to allow GPL code to be used
within a non-GPL kernel, it would be nice if they could outline the
approach here. It would be very interesting to see how they have done it.

Thanks,

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-09-28 13:01:56 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any mention
of any linking exceptions or similar.
How do you reconcile the GPL licence for GFS2 with VMS if you are going
to integrate GFS2 into the VMS kernel ?
What mechanisms are you using to stop the VMS kernel code from falling
under the GPL licence if you are integrating GFS2 into VMS kernel code ?
Is there some linking exception or similar for GFS2 that I have missed
or have VSI come up with a way to integrate the GFS2 code into VMS
without causing at least of some of the existing VMS code to now be
covered by the GPL ?
Interesting question.
I hope VSI thought about that.
I've had another look at this and I don't immediately see how the options
below are viable ways to get GPL code into the VMS kernel without causing
at least some of the existing VMS kernel code to now fall until the GPL.
Post by Arne Vajhøj
I believe they can use two different approaches to work
* put in a layer between VMS and GFS2 - the "NVidia model"
The problem with this one is that the situation is reversed from the
Nvidia case. In the Nvidia case, Nvidia themselves own the code that
they want to stop from becoming GPL when it is inserted into a GPL kernel.
In the GFS2 case, VSI want to use pure GPL V2 code that they do not own
within a non-GPL kernel (the VMS kernel). This needs to happen in a way
that does not make the rest of the VMS kernel now fall until the GPL as
a result.
GPL does not operate with a caller/called concept - only on a
combined concept, so it should not matter for GPL.
Post by Simon Clubley
If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Not necessarily.

If usage is good according to GPL the the authors permission is
not needed.

There are obviously benefits from getting the permission though. If
FSF sue them and GFS2 people say they do not consider it a violation,
then the case would be a lot easier.
Post by Simon Clubley
Post by Arne Vajhøj
* claim it is program and whatever parts of VMS it needs
fall under the system library exception in GPL
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.
This one doesn't work because of the last bit. The VMS kernel will
accompany the GFS2 executables as part of the installation media
and GFS2 will be tied into the VMS kernel itself.
VSI-GFS2.PCSI

:-)
Post by Simon Clubley
The example I keep coming back to is that GNU Readline (which is pure GPL
and not LGPL) makes any program which uses it fall under a GPL-compatible
licence.
That is a very traditional library with no shim and no system
library exception and a general understanding in the industry that
linking with such a library is combining.

Arne
Simon Clubley
2020-09-28 18:37:45 UTC
Permalink
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
Interesting question.
I hope VSI thought about that.
I've had another look at this and I don't immediately see how the options
below are viable ways to get GPL code into the VMS kernel without causing
at least some of the existing VMS kernel code to now fall until the GPL.
Post by Arne Vajhøj
I believe they can use two different approaches to work
* put in a layer between VMS and GFS2 - the "NVidia model"
The problem with this one is that the situation is reversed from the
Nvidia case. In the Nvidia case, Nvidia themselves own the code that
they want to stop from becoming GPL when it is inserted into a GPL kernel.
In the GFS2 case, VSI want to use pure GPL V2 code that they do not own
within a non-GPL kernel (the VMS kernel). This needs to happen in a way
that does not make the rest of the VMS kernel now fall until the GPL as
a result.
GPL does not operate with a caller/called concept - only on a
combined concept, so it should not matter for GPL.
Post by Simon Clubley
If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Not necessarily.
If usage is good according to GPL the the authors permission is
not needed.
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.

As such, it's not what the GPL says; it's what the owners of the GFS2
code _believe_ the GPL says that matters when they decide whether what
VSI are doing is acceptable.
Post by Arne Vajhøj
There are obviously benefits from getting the permission though. If
FSF sue them and GFS2 people say they do not consider it a violation,
then the case would be a lot easier.
It would be a lot better however if VSI simply state upfront why they
believe they can integrate pure GPL code into the VMS kernel without
a problem. They might be planning something that makes the problem go
away or there might simply be an angle which has been missed so far.
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
* claim it is program and whatever parts of VMS it needs
fall under the system library exception in GPL
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.
This one doesn't work because of the last bit. The VMS kernel will
accompany the GFS2 executables as part of the installation media
and GFS2 will be tied into the VMS kernel itself.
VSI-GFS2.PCSI
:-)
Not that simple. Have you looked at the VMS kernel internals ?

I've never seen the VMS source code, but it's clear from the I&DS
that the VMS kernel is a monolithic mass of closely bound code, at
least when compared to the clean interfaces and plug-in modules
that Linux has.

That approach also doesn't work if you ship VSI-GFS2.PCSI in the same
installation kit as the VMS installation kit even if VSI add a public
filesystem interface similar to that which Linux has.

There's also the very open question about what happens when a vendor
loads a GPL filesystem module into the VMS kernel using this new interface
when the same vendor supplies the operating system. Would this be seen
as a deliberate attempt to circumvent the requirements of the GPL ?
Post by Arne Vajhøj
Post by Simon Clubley
The example I keep coming back to is that GNU Readline (which is pure GPL
and not LGPL) makes any program which uses it fall under a GPL-compatible
licence.
That is a very traditional library with no shim and no system
library exception and a general understanding in the industry that
linking with such a library is combining.
And that's why it's such a good example. For example (and I honestly
don't know the answer to this question) could you get around the
GNU Readline GPL requirements by placing all the GNU Readline code
into a .so module and then using dlopen() at runtime to map it into
the already running program ?

My feeling is to say I don't think so as that should be treated as a
deliberate attempt to get around the GPL requirements and the author
of the program would have written code to deliberately load GPL code
into their non-GPL program via dlopen().

That would appear to be similar to the same vendor supplying both the
operating system and the plug-in modules when the operating system is
designed to take advantage of the plug-in modules.

I would really like to see VSI's take on this. They may have a simple
and/or clever answer that allows them to do this without forcing at
least some of the VMS kernel code to go under a GPL-compatible licence.

Actually, after I wrote the above, I decided to have a look at the GPL
FAQ before posting and came across this:

https://www.gnu.org/licenses/gpl-faq.en.html#NFUseGPLPlugins

That would seem to support what I am saying in the above Readline example.

It would also seem to indirectly imply that parts of the VMS kernel
would become covered by the GPL when the same vendor provides the
operating system and the plug-in because VSI would be distributing
both the main program (the VMS kernel) and the plug-in (the GFS2
filesystem).

If you read the linked question here:

https://www.gnu.org/licenses/gpl-faq.en.html#GPLPlugins

then I would argue that the VMS kernel and the GFS2 filesystem would
become a single combined program under that definition, especially when
you consider that the primary use for GFS2 is with VMS clustering,
which is tightly integrated into the VMS kernel.

It would be nice to hear from VSI about how they plan to integrate
GFS2 into VMS because they may have a plan that addresses all of the
above issues.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-09-28 19:06:15 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
Interesting question.
I hope VSI thought about that.
I've had another look at this and I don't immediately see how the options
below are viable ways to get GPL code into the VMS kernel without causing
at least some of the existing VMS kernel code to now fall until the GPL.
Post by Arne Vajhøj
I believe they can use two different approaches to work
* put in a layer between VMS and GFS2 - the "NVidia model"
The problem with this one is that the situation is reversed from the
Nvidia case. In the Nvidia case, Nvidia themselves own the code that
they want to stop from becoming GPL when it is inserted into a GPL kernel.
In the GFS2 case, VSI want to use pure GPL V2 code that they do not own
within a non-GPL kernel (the VMS kernel). This needs to happen in a way
that does not make the rest of the VMS kernel now fall until the GPL as
a result.
GPL does not operate with a caller/called concept - only on a
combined concept, so it should not matter for GPL.
Post by Simon Clubley
If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Not necessarily.
If usage is good according to GPL the the authors permission is
not needed.
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
I agree that it is currently not clear.
Post by Simon Clubley
As such, it's not what the GPL says; it's what the owners of the GFS2
code _believe_ the GPL says that matters when they decide whether what
VSI are doing is acceptable.
No no no.

In its core it only matters if if the usage is legal under
under the license given per legal interpretation of the license.

License granter is bound by the license.
Post by Simon Clubley
Post by Arne Vajhøj
There are obviously benefits from getting the permission though. If
FSF sue them and GFS2 people say they do not consider it a violation,
then the case would be a lot easier.
It would be a lot better however if VSI simply state upfront why they
believe they can integrate pure GPL code into the VMS kernel without
a problem. They might be planning something that makes the problem go
away or there might simply be an angle which has been missed so far.
True.
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
* claim it is program and whatever parts of VMS it needs
fall under the system library exception in GPL
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.
This one doesn't work because of the last bit. The VMS kernel will
accompany the GFS2 executables as part of the installation media
and GFS2 will be tied into the VMS kernel itself.
VSI-GFS2.PCSI
:-)
Not that simple. Have you looked at the VMS kernel internals ?
I've never seen the VMS source code, but it's clear from the I&DS
that the VMS kernel is a monolithic mass of closely bound code, at
least when compared to the clean interfaces and plug-in modules
that Linux has.
It matters a lot if the legally important aspect is the term
"accompanies".
Post by Simon Clubley
There's also the very open question about what happens when a vendor
loads a GPL filesystem module into the VMS kernel using this new interface
when the same vendor supplies the operating system. Would this be seen
as a deliberate attempt to circumvent the requirements of the GPL ?
The argument that something is illegal even though it follows
the license because it is done that way to circumvent the license will
be very difficult to make stick.
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
The example I keep coming back to is that GNU Readline (which is pure GPL
and not LGPL) makes any program which uses it fall under a GPL-compatible
licence.
That is a very traditional library with no shim and no system
library exception and a general understanding in the industry that
linking with such a library is combining.
And that's why it's such a good example.
I question whether a well known but non similar example is a
good example.
Post by Simon Clubley
For example (and I honestly
don't know the answer to this question) could you get around the
GNU Readline GPL requirements by placing all the GNU Readline code
into a .so module and then using dlopen() at runtime to map it into
the already running program ?
That would not avoid GPL requirements.

Static vs dynamic linking does not matter for GPL.
Post by Simon Clubley
My feeling is to say I don't think so as that should be treated as a
deliberate attempt to get around the GPL requirements and the author
of the program would have written code to deliberately load GPL code
into their non-GPL program via dlopen().
It is not a matter about circumventing. The stuff is simply
not less combined because it is dynamic linked instead of
static linked.

Static vs dynamic does not matter for GPL.

It matters for LGPL, but that is different.
Post by Simon Clubley
Actually, after I wrote the above, I decided to have a look at the GPL
https://www.gnu.org/licenses/gpl-faq.en.html#NFUseGPLPlugins
That would seem to support what I am saying in the above Readline example.
It would also seem to indirectly imply that parts of the VMS kernel
would become covered by the GPL when the same vendor provides the
operating system and the plug-in because VSI would be distributing
both the main program (the VMS kernel) and the plug-in (the GFS2
filesystem).
https://www.gnu.org/licenses/gpl-faq.en.html#GPLPlugins
then I would argue that the VMS kernel and the GFS2 filesystem would
become a single combined program under that definition, especially when
you consider that the primary use for GFS2 is with VMS clustering,
which is tightly integrated into the VMS kernel.
The mechanisms of GPL are rather well known.

If there is no linking exception then the only two approaches
I can see is the either the shim or the system library exception.

Arne
Simon Clubley
2020-09-29 12:21:33 UTC
Permalink
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Not necessarily.
If usage is good according to GPL the the authors permission is
not needed.
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
I agree that it is currently not clear.
Post by Simon Clubley
As such, it's not what the GPL says; it's what the owners of the GFS2
code _believe_ the GPL says that matters when they decide whether what
VSI are doing is acceptable.
No no no.
Yes yes yes, but for a reason you have not mentioned, but which I was
thinking of when I wrote that.
Post by Arne Vajhøj
In its core it only matters if if the usage is legal under
under the license given per legal interpretation of the license.
License granter is bound by the license.
From a legal point of view, you are correct. From a practical point
of view, you are not correct.

If the GFS2 copyright holders believe what VSI are doing is wrong,
even if VSI ultimately turn out to be correct, then those copyright
holders can make VSI's life a misery until the situation is finally
resolved and cause major uncertainty in the VMS community until it
is finally resolved.

For proof, I refer you to what happened with SCO and the uncertainty
that caused in the Linux community for a while.

That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-09-29 13:23:16 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
If VSI are going down this path, they also need to get the owners of the
GFS2 code to agree that this is an acceptable use that is compatible with
the GPL.
Not necessarily.
If usage is good according to GPL the the authors permission is
not needed.
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
I agree that it is currently not clear.
Post by Simon Clubley
As such, it's not what the GPL says; it's what the owners of the GFS2
code _believe_ the GPL says that matters when they decide whether what
VSI are doing is acceptable.
No no no.
Yes yes yes, but for a reason you have not mentioned, but which I was
thinking of when I wrote that.
Post by Arne Vajhøj
In its core it only matters if if the usage is legal under
under the license given per legal interpretation of the license.
License granter is bound by the license.
From a legal point of view, you are correct. From a practical point
of view, you are not correct.
Having a legal right also helps with the practical right.
Post by Simon Clubley
If the GFS2 copyright holders believe what VSI are doing is wrong,
even if VSI ultimately turn out to be correct, then those copyright
holders can make VSI's life a misery until the situation is finally
resolved and cause major uncertainty in the VMS community until it
is finally resolved.
For proof, I refer you to what happened with SCO and the uncertainty
that caused in the Linux community for a while.
Well - Linux ended up a lot better than SCO, so that would not be
a bad precedence.

Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
Post by Simon Clubley
That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.
Clarity is always good.

But my guess is that VSI will say absolutely nothing about the
topic.

VSI legal team will not want to prepare legal statements for
c.o.v/I-V.

And VSI technical guys do not want to post on a legal matter
you potentially could end up in court.

Arne
Craig A. Berry
2020-09-29 14:09:51 UTC
Permalink
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
If the entire codebase actually has just one copyright holder, then as
far as I know they could re-license it to VSI under whatever terms they
want and the GPL of the open source version wouldn't apply.
Arne Vajhøj
2020-09-29 14:35:50 UTC
Permalink
Post by Craig A. Berry
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
If the entire codebase actually has just one copyright holder, then as
far as I know they could re-license it to VSI under whatever terms they
want and the GPL of the open source version wouldn't apply.
GFS2 is a "Redhat thing"

But a quick look at the source code shows both Redhat copright
and other copyright. Not sure now much of the other copyright is
original GFS2 contributions and how much is something GFS2 team
reused from other Linux stuff (some must be in the latter category as
some copyright notices are older than GFS2).

Messy.

Arne
Simon Clubley
2020-09-29 17:25:18 UTC
Permalink
Post by Arne Vajhøj
Post by Simon Clubley
If the GFS2 copyright holders believe what VSI are doing is wrong,
even if VSI ultimately turn out to be correct, then those copyright
holders can make VSI's life a misery until the situation is finally
resolved and cause major uncertainty in the VMS community until it
is finally resolved.
For proof, I refer you to what happened with SCO and the uncertainty
that caused in the Linux community for a while.
Well - Linux ended up a lot better than SCO, so that would not be
a bad precedence.
Linux also went through a lot of hassle and uncertainty before it got
to that stage and some people ended up paying Linux licence fees to SCO
in the meantime.
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
https://www.ft.com/content/6f02ce76-e1d6-45d8-b27a-0491281c2507

https://www.networkworld.com/article/2300069/ibm-sues-amazon-for-patent-infringement.html

http://techrights.org/2016/03/03/ibm-swpats-weaponised/
Post by Arne Vajhøj
Post by Simon Clubley
That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.
Clarity is always good.
But my guess is that VSI will say absolutely nothing about the
topic.
VSI legal team will not want to prepare legal statements for
c.o.v/I-V.
And VSI technical guys do not want to post on a legal matter
you potentially could end up in court.
But why ? If there's not an issue, why doesn't VSI simply say
how they plan to integrate the code ? It could make for a _very_
informative discussion about how to use GPL code within a kernel
which is not GPL compatible if VSI have figured out how to do that.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Simon Clubley
2020-09-29 17:34:42 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
https://www.ft.com/content/6f02ce76-e1d6-45d8-b27a-0491281c2507
When I click on this link directly, it turns out to be paywalled.

Searching for "ibm patent lawsuits" in Google (without the quotes)
brings up the article with a title of "IBM takes Airbnb to court over
historic patents". If I click on the Google result, I am able to see
the article. It looks like they let you through the paywall if you
have a Google referrer in the HTTP headers.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-09-29 17:45:14 UTC
Permalink
Post by Simon Clubley
Post by Simon Clubley
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
https://www.ft.com/content/6f02ce76-e1d6-45d8-b27a-0491281c2507
When I click on this link directly, it turns out to be paywalled.
Searching for "ibm patent lawsuits" in Google (without the quotes)
brings up the article with a title of "IBM takes Airbnb to court over
historic patents". If I click on the Google result, I am able to see
the article. It looks like they let you through the paywall if you
have a Google referrer in the HTTP headers.
patents <> copyright

And even for patents I would not consider IBM thatvery aggressive. They
have filed for most patents for like 25 years in row, so a few cases is
to be expected.

And even though Redhat is now part of IBM then IBM has stated that
they will let Redhat operate rather independently.

Arne
Arne Vajhøj
2020-09-29 17:39:55 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.
Clarity is always good.
But my guess is that VSI will say absolutely nothing about the
topic.
VSI legal team will not want to prepare legal statements for
c.o.v/I-V.
And VSI technical guys do not want to post on a legal matter
you potentially could end up in court.
But why ? If there's not an issue, why doesn't VSI simply say
how they plan to integrate the code ?
Employees does not comment public on a matter that specifically
involves their companies and could end up in court.

That could provide ammunition for those that want to sue.

And turn employees into ex-employees.
Post by Simon Clubley
It could make for a _very_
informative discussion about how to use GPL code within a kernel
which is not GPL compatible if VSI have figured out how to do that.
Sure.

But but that will not make it happen.

Arne
John Dallman
2020-09-29 18:28:00 UTC
Permalink
Post by Simon Clubley
Linux also went through a lot of hassle and uncertainty before it
got to that stage and some people ended up paying Linux licence
fees to SCO in the meantime.
Not that much hassle, as I experienced it, and not many people paid. It
became clear after a while that SCO's case was largely built on wishful
thinking.

They refused to provide examples of the code that they claimed had been
copied from them, on the grounds that they would then be replaced. Well,
yes. But that would not remove the fact that their code had been pirated,
if, in fact, it had. If it had, they could have made a case for damages.
But they wanted to turn it into /owning/ Linux, and they didn't actually
have any grounds under the law for that.

John
Arne Vajhøj
2020-09-29 18:52:26 UTC
Permalink
Post by John Dallman
Post by Simon Clubley
Linux also went through a lot of hassle and uncertainty before it
got to that stage and some people ended up paying Linux licence
fees to SCO in the meantime.
Not that much hassle, as I experienced it, and not many people paid. It
became clear after a while that SCO's case was largely built on wishful
thinking.
They refused to provide examples of the code that they claimed had been
copied from them, on the grounds that they would then be replaced. Well,
yes. But that would not remove the fact that their code had been pirated,
if, in fact, it had. If it had, they could have made a case for damages.
But they wanted to turn it into /owning/ Linux, and they didn't actually
have any grounds under the law for that.
I believe the cases were very messy.

SCO claimed that they owned Unix and Linux violated Unix
copyright. That claim was rejected when the courts decided
that Novell owned Unix not SCO.

They also claimed that some IBM code contributed to Linux
was in violation of a business agreement between IBM and
them. Those claims got mostly thrown out by the court and
that case died.

At some point in time SCO also tried to get GPL deemed
unconstitutional. But they dropped that.

Arne
Stephane Tougard
2020-09-30 13:35:27 UTC
Permalink
Post by John Dallman
They refused to provide examples of the code that they claimed had been
copied from them, on the grounds that they would then be replaced. Well,
yes. But that would not remove the fact that their code had been pirated,
if, in fact, it had. If it had, they could have made a case for damages.
But they wanted to turn it into /owning/ Linux, and they didn't actually
have any grounds under the law for that.
SCO was not even the owner of the code they were claiming was stolen
by the Linux community, they just had a license on this code, but Novel
was the legitimate owner.

Novel dropped the case, SCO just went bankrupt, killed by their own
stupidity and the very bad publicity around this this story.

In fact, SCO was not SCO, it was Caldera and Caldera has sold a full
Linux distribution for years under the GPL license. Just because of
that, their claim was stupid. They sold during years a software under GPL
license and after years they would have said that GPL can not apply
because they own the code (code they don't own in the first place).

The whole story was just ludicrous.
Stephane Tougard
2020-09-30 13:30:03 UTC
Permalink
Post by Simon Clubley
Linux also went through a lot of hassle and uncertainty before it got
to that stage and some people ended up paying Linux licence fees to SCO
in the meantime.
Some people, such as Microsoft, supported SCO in their stupid claim and
purchased some license to financially help them.
Post by Simon Clubley
But why ? If there's not an issue, why doesn't VSI simply say
how they plan to integrate the code ? It could make for a _very_
informative discussion about how to use GPL code within a kernel
which is not GPL compatible if VSI have figured out how to do that.
They clearly can not without the express authorization of all the
copyrights holders, which may make a lot of people as any guy who sent
a "one line patch" is a copyright holder.

The LGPL is a license made to use a library directly linked with a
non free binary (I guess it could apply to a module and a kernel), but
the GPL itself does not allow that.

The only way you can merge non free software and GPL software is when
they run as independant software (different PID) and they communicate
through an IPC/RPC such as file, pipe, TCP cnx ...

However, the reverse is possible, it's possible for a free kernel (such
as Linux) to use a proprietary module. Still, this is very arguable, but
it exists anyway, so it's possible.

Usually, proprietary Operating System borrowing code from Free world use
*BSD as a source.
Stephane Tougard
2020-09-30 13:20:32 UTC
Permalink
Post by Arne Vajhøj
Post by Simon Clubley
For proof, I refer you to what happened with SCO and the uncertainty
that caused in the Linux community for a while.
Well - Linux ended up a lot better than SCO, so that would not be
a bad precedence.
There was never any real uncertainty anyway. The claim of SCO was
indefensible and delusional at best.

The Debian community worked on the version of the distribution using
FreeBSD kernel, just in case, but that was, as far as I know, the most
important side effect.
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
Post by Simon Clubley
That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.
GFS2 seems to be GPL2, which means that it can not be included inside a
non free kernel.
Simon Clubley
2020-10-01 12:19:49 UTC
Permalink
Post by Stephane Tougard
Post by Arne Vajhøj
Post by Simon Clubley
For proof, I refer you to what happened with SCO and the uncertainty
that caused in the Linux community for a while.
Well - Linux ended up a lot better than SCO, so that would not be
a bad precedence.
There was never any real uncertainty anyway. The claim of SCO was
indefensible and delusional at best.
The Debian community worked on the version of the distribution using
FreeBSD kernel, just in case, but that was, as far as I know, the most
important side effect.
Post by Arne Vajhøj
Also note that the GFS2 copyright holder Redhat (IBM) is not known
for being aggressive with law suits.
Post by Simon Clubley
That is why VSI need to be very clear about how they intend to integrate
GFS2 into the VMS kernel and they need to do it in a way that the GFS2
copyright holders do not have a problem with.
GFS2 seems to be GPL2, which means that it can not be included inside a
non free kernel.
I've just sent email to VSI asking them how they intend to handle
this problem, since I am no closer to understanding how you can insert
GPL V2 code directly into the VMS kernel in a way that complies with
the GPL V2 licence terms.

I'm assuming VSI have a valid and possibly creative answer. I'm just
no closer to seeing what that answer is.

I'll post a response here if I get a reply.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Simon Clubley
2020-10-08 12:20:20 UTC
Permalink
Post by Simon Clubley
Post by Stephane Tougard
GFS2 seems to be GPL2, which means that it can not be included inside a
non free kernel.
I've just sent email to VSI asking them how they intend to handle
this problem, since I am no closer to understanding how you can insert
GPL V2 code directly into the VMS kernel in a way that complies with
the GPL V2 licence terms.
I'm assuming VSI have a valid and possibly creative answer. I'm just
no closer to seeing what that answer is.
I'll post a response here if I get a reply.
Just to let people know: so far, I have not had any reply from VSI.

I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.

I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Jan-Erik Söderholm
2020-10-08 12:56:16 UTC
Permalink
Post by Simon Clubley
Post by Simon Clubley
Post by Stephane Tougard
GFS2 seems to be GPL2, which means that it can not be included inside a
non free kernel.
I've just sent email to VSI asking them how they intend to handle
this problem, since I am no closer to understanding how you can insert
GPL V2 code directly into the VMS kernel in a way that complies with
the GPL V2 licence terms.
I'm assuming VSI have a valid and possibly creative answer. I'm just
no closer to seeing what that answer is.
I'll post a response here if I get a reply.
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
Simon.
I do not understand why you expect that VSI would reply to you about this.
Simon Clubley
2020-10-08 17:42:49 UTC
Permalink
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.

VSI are talking about integrating code that appears to be pure GPL V2
(without any apparent linking or other exceptions) directly into the
core of VMS (which is not licenced using a GPL V2 compatible licence).

This is so unusual that I consider it reasonable to ask VSI how they
have managed to do this and for them to explain how they have managed
to integrate what is apparently pure GPL V2 code directly into VMS
while still keeping to the requirements of the GPL licence.

Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.

However, given how unusual this is, I still believe that VSI should
explain how they have done this so that people can make sure there
isn't an aspect to this that VSI might have missed.

Is anyone else here interested in knowing how VSI have managed to
integrate the GFS2 code into VMS ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Chris Townley
2020-10-08 19:02:41 UTC
Permalink
Post by Simon Clubley
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.
VSI are talking about integrating code that appears to be pure GPL V2
(without any apparent linking or other exceptions) directly into the
core of VMS (which is not licenced using a GPL V2 compatible licence).
This is so unusual that I consider it reasonable to ask VSI how they
have managed to do this and for them to explain how they have managed
to integrate what is apparently pure GPL V2 code directly into VMS
while still keeping to the requirements of the GPL licence.
Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.
However, given how unusual this is, I still believe that VSI should
explain how they have done this so that people can make sure there
isn't an aspect to this that VSI might have missed.
Is anyone else here interested in knowing how VSI have managed to
integrate the GFS2 code into VMS ?
Simon.
Is it not possible they have negotiated to buy a license to use the code that is suitable for their purpose from the owners?


Chris
Jan-Erik Söderholm
2020-10-08 20:03:31 UTC
Permalink
Post by Simon Clubley
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.
Anything VSI does is "unusual" in the IT world when thinking of
the market share of OpenVMS.
Post by Simon Clubley
Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.
Agree. Whatever they do, I just assume that it is OK. I will not ask.
Bill Gunshannon
2020-10-08 21:56:08 UTC
Permalink
Post by Chris Townley
Post by Simon Clubley
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.
VSI are talking about integrating code that appears to be pure GPL V2
(without any apparent linking or other exceptions) directly into the
core of VMS (which is not licenced using a GPL V2 compatible licence).
This is so unusual that I consider it reasonable to ask VSI how they
have managed to do this and for them to explain how they have managed
to integrate what is apparently pure GPL V2 code directly into VMS
while still keeping to the requirements of the GPL licence.
Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.
However, given how unusual this is, I still believe that VSI should
explain how they have done this so that people can make sure there
isn't an aspect to this that VSI might have missed.
Is anyone else here interested in knowing how VSI have managed to
integrate the GFS2 code into VMS ?
Simon.
Is it not possible they have negotiated to buy a license to use the code that is suitable for their purpose from the owners?
It would be good if people were told this. Would you want to
to be betting your business on a company that was setting
themselves up for a potentially expensive legal battle?

bill
Simon Clubley
2020-10-09 12:21:26 UTC
Permalink
Post by Bill Gunshannon
Post by Chris Townley
Is it not possible they have negotiated to buy a license to use the code that is suitable for their purpose from the owners?
It would be good if people were told this. Would you want to
to be betting your business on a company that was setting
themselves up for a potentially expensive legal battle?
And that is exactly why I am surprised that VSI have not answered,
especially if it is a nice simple answer.

A good number of companies get uneasy when the licence and usage terms
on a product they are interested in are not clear, and for good reasons.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-10-09 12:49:12 UTC
Permalink
Post by Simon Clubley
Post by Bill Gunshannon
It would be good if people were told this. Would you want to
to be betting your business on a company that was setting
themselves up for a potentially expensive legal battle?
And that is exactly why I am surprised that VSI have not answered,
especially if it is a nice simple answer.
A good number of companies get uneasy when the licence and usage terms
on a product they are interested in are not clear, and for good reasons.
That may be the only way to get an answer from VSI.

Call them and say that you want to buy 100 licenses if you can get
GFS2 but that you need evidence that it is OK.

Then VSI senior management will tell legal to provide an
explanation.

A comp.os.vms discussion is just not the same.

Arne

Simon Clubley
2020-10-09 12:17:46 UTC
Permalink
Post by Chris Townley
Is it not possible they have negotiated to buy a license to use the code that is suitable for their purpose from the owners?
That's certainly possible, although I don't know how much the source
code would cost for such a licence.

But do they buy the really old version from when RedHat integrated it
into Linux over a decade ago or do they buy the current version which
may have a whole range of copyright owners due to the various
contributions and changes to Linux over the years ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Scott Dorsey
2020-10-09 12:26:21 UTC
Permalink
Is it not possible they have negotiated to buy a license to use the code th=
at is suitable for their purpose from the owners?
You don't negotiate with rms. You might as well argue with a rock.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Simon Clubley
2020-10-09 12:38:42 UTC
Permalink
Post by Scott Dorsey
Is it not possible they have negotiated to buy a license to use the code th=
at is suitable for their purpose from the owners?
You don't negotiate with rms. You might as well argue with a rock.
Red Hat are the original copyright owners after buying the product,
so they could in theory dual-licence that original version for a
sufficient amount of money.

However, how many other copyright owners might there be over the
last decade or so as a result of additional changes to the code
after it was integrated into Linux ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Dave Froble
2020-10-09 00:42:04 UTC
Permalink
Post by Simon Clubley
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.
VSI are talking about integrating code that appears to be pure GPL V2
(without any apparent linking or other exceptions) directly into the
core of VMS (which is not licenced using a GPL V2 compatible licence).
This is so unusual that I consider it reasonable to ask VSI how they
have managed to do this and for them to explain how they have managed
to integrate what is apparently pure GPL V2 code directly into VMS
while still keeping to the requirements of the GPL licence.
Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.
However, given how unusual this is, I still believe that VSI should
explain how they have done this so that people can make sure there
isn't an aspect to this that VSI might have missed.
Is anyone else here interested in knowing how VSI have managed to
integrate the GFS2 code into VMS ?
Simon.
I haven't been paying much attention. I have a question. Has VSI said
that they are using the existing GFS2 code, or, just the design, which
they perhaps are implementing?

If they write their own, where would be the problem?
--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: ***@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486
Simon Clubley
2020-10-09 12:34:06 UTC
Permalink
Post by Dave Froble
I haven't been paying much attention. I have a question. Has VSI said
that they are using the existing GFS2 code, or, just the design, which
they perhaps are implementing?
If they write their own, where would be the problem?
If they write their own from a formal public specification and don't need
to use the GPL code to understand how it works then there is no problem.

But if VSI are going to the major effort of doing that, then why don't
they simply complete the VAFS work instead, which is a product that
they would fully control and which would be a native VMS product ?

BTW, I have not been able to find a public specification for the current
GFS2 filesystem (not the original GFS filesystem). Does anyone have
a link to the current formal _GFS2_ specification ?

Thanks,

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Arne Vajhøj
2020-10-09 12:46:14 UTC
Permalink
Post by Simon Clubley
Post by Jan-Erik Söderholm
Post by Simon Clubley
Just to let people know: so far, I have not had any reply from VSI.
I asked again yesterday in case the original query had been forgotten
about, but I have not heard anything back from that either.
I, for one, would like to know how VSI have achieved this. It could
make for quite an interesting discussion.
I do not understand why you expect that VSI would reply to you about this.
Because based on their own roadmap, what VSI are doing is very unusual.
This is so unusual that I consider it reasonable to ask VSI how they
have managed to do this and for them to explain how they have managed
to integrate what is apparently pure GPL V2 code directly into VMS
while still keeping to the requirements of the GPL licence.
"reasonable" is not a particular relevant term.

It is all about what makes sense for a company.

VSI technical people will not want to comment public on a
legal matter.

VSI legal people will not want to comment public on a legal
matter unless they have to.

Companies don't do that.
Post by Simon Clubley
Note that my working assumption is that VSI _have_ found some way to
do this that they believe still complies with the GPL.
However, given how unusual this is, I still believe that VSI should
explain how they have done this so that people can make sure there
isn't an aspect to this that VSI might have missed.
I don't think VSI legal team will consider comp.os.vms the right
place to consult if any doubt. There are are legal companies
specializing in open source licenses.
Post by Simon Clubley
Is anyone else here interested in knowing how VSI have managed to
integrate the GFS2 code into VMS ?
Probably lots.

But our curiosity probably doesn't matter much to VSI legal team.

Arne
Dave Froble
2020-09-28 19:08:23 UTC
Permalink
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: ***@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486
Phillip Helbig (undress to reply)
2020-09-29 09:24:37 UTC
Permalink
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
Read some of the "Gnu philosophy" stuff by Stallman. He actually says
that closed-source software is a "crime against humanity". It is said
that so many have jumped on the open-source bandwagon for such trivial
reasons, rather than stepping back and asking themselves if it isn't
worth paying a bit more to get some sort of sensible product.
Phillip Helbig (undress to reply)
2020-09-29 09:28:20 UTC
Permalink
Post by Phillip Helbig (undress to reply)
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
Read some of the "Gnu philosophy" stuff by Stallman. He actually says
that closed-source software is a "crime against humanity". It is said
said ---> sad
Post by Phillip Helbig (undress to reply)
that so many have jumped on the open-source bandwagon for such trivial
reasons, rather than stepping back and asking themselves if it isn't
worth paying a bit more to get some sort of sensible product.
Bill Gunshannon
2020-09-29 11:33:01 UTC
Permalink
Post by Phillip Helbig (undress to reply)
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
Read some of the "Gnu philosophy" stuff by Stallman. He actually says
that closed-source software is a "crime against humanity". It is said
that so many have jumped on the open-source bandwagon for such trivial
reasons, rather than stepping back and asking themselves if it isn't
worth paying a bit more to get some sort of sensible product.
Even more absurd when you know what led to his actions. But I am
sure, if it hasn't been already, IT history will be re-written to
make him the savior of the industry rather than the idiot he is.

bill
Simon Clubley
2020-09-29 12:30:43 UTC
Permalink
Post by Phillip Helbig (undress to reply)
Read some of the "Gnu philosophy" stuff by Stallman. He actually says
that closed-source software is a "crime against humanity". It is said
that so many have jumped on the open-source bandwagon for such trivial
reasons, rather than stepping back and asking themselves if it isn't
worth paying a bit more to get some sort of sensible product.
I agree some of his views are extreme, but OTOH he was absolutely
spot on when he wrote The Right to Read:

https://en.wikipedia.org/wiki/The_Right_to_Read

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Simon Clubley
2020-09-29 12:28:49 UTC
Permalink
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
The GPL licence terms are very clear and are designed to meet a very
well documented goal. The person/team/company responsible for the project
you are interested in have decided that they wish to release their work
under those terms because they agree with the goals of the GPL.

If the GPL terms are unacceptable to you, there is nothing stopping you
from designing and implementing your own version of a GPL project under
different licence terms or implementing the same functionality in a
different way.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Dave Froble
2020-09-29 14:30:26 UTC
Permalink
Post by Simon Clubley
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
The GPL licence terms are very clear and are designed to meet a very
well documented goal. The person/team/company responsible for the project
you are interested in have decided that they wish to release their work
under those terms because they agree with the goals of the GPL.
If the GPL terms are unacceptable to you, there is nothing stopping you
from designing and implementing your own version of a GPL project under
different licence terms or implementing the same functionality in a
different way.
Simon.
It is not the specific terms that I object to as much as the mentality
behind them. Frankly, I consider the concept one that a thief might have.

Stallman and the chicoms seem to have much in common.
--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: ***@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486
Arne Vajhøj
2020-09-29 14:41:32 UTC
Permalink
Post by Dave Froble
Post by Simon Clubley
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
The GPL licence terms are very clear and are designed to meet a very
well documented goal. The person/team/company responsible for the project
you are interested in have decided that they wish to release their work
under those terms because they agree with the goals of the GPL.
If the GPL terms are unacceptable to you, there is nothing stopping you
from designing and implementing your own version of a GPL project under
different licence terms or implementing the same functionality in a
different way.
It is not the specific terms that I object to as much as the mentality
behind them.  Frankly, I consider the concept one that a thief might have.
Stallman and the chicoms seem to have much in common.
It is something for something.

Big Profit Inc. let you use their library in your product
if you pay them royalties for each copy you sell.

GPL project XYZ let you use their library for free
in your product if you give your users access to
source code, rigts to modify and right to redistribute.

It is just rights instead of money.

In both cases you can say no. Nobody is forcing you to
buy from Big Profits Inc or use the software from GPL project
XYZ.

Arne
Dave Froble
2020-09-29 16:12:25 UTC
Permalink
Post by Arne Vajhøj
Post by Dave Froble
Post by Simon Clubley
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
The GPL licence terms are very clear and are designed to meet a very
well documented goal. The person/team/company responsible for the project
you are interested in have decided that they wish to release their work
under those terms because they agree with the goals of the GPL.
If the GPL terms are unacceptable to you, there is nothing stopping you
from designing and implementing your own version of a GPL project under
different licence terms or implementing the same functionality in a
different way.
It is not the specific terms that I object to as much as the mentality
behind them. Frankly, I consider the concept one that a thief might have.
Stallman and the chicoms seem to have much in common.
It is something for something.
Big Profit Inc. let you use their library in your product
if you pay them royalties for each copy you sell.
GPL project XYZ let you use their library for free
in your product if you give your users access to
source code, rigts to modify and right to redistribute.
It is just rights instead of money.
No it is not, because I've never seen anything about proportions. If
99% of some product is proprietary, and 1% is GPL, then it is absurd for
a vendor to give up 99% of the IP just for 1% of GPL IP.
--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: ***@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486
Arne Vajhøj
2020-09-29 16:19:47 UTC
Permalink
Post by Arne Vajhøj
Post by Dave Froble
Post by Simon Clubley
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
The GPL licence terms are very clear and are designed to meet a very
well documented goal. The person/team/company responsible for the project
you are interested in have decided that they wish to release their work
under those terms because they agree with the goals of the GPL.
If the GPL terms are unacceptable to you, there is nothing stopping you
from designing and implementing your own version of a GPL project under
different licence terms or implementing the same functionality in a
different way.
It is not the specific terms that I object to as much as the mentality
behind them.  Frankly, I consider the concept one that a thief might have.
Stallman and the chicoms seem to have much in common.
It is something for something.
Big Profit Inc. let you use their library in your product
if you pay them royalties for each copy you sell.
GPL project XYZ let you use their library for free
in your product if you give your users access to
source code, rigts to modify and right to redistribute.
It is just rights instead of money.
No it is not, because I've never seen anything about proportions.  If
99% of some product is proprietary, and 1% is GPL, then it is absurd for
a vendor to give up 99% of the IP just for 1% of GPL IP.
They give up 0% of the IP. They still own the copyright 100%.

But yes it does not distinguish between 1% GPL, 50% GPL and 99% GPL.

On the other hand then Big Profit Inc. does not care about
whether their library is 1%, 50% or 99% either.

Arne
Simon Clubley
2020-09-29 17:44:11 UTC
Permalink
Post by Dave Froble
Post by Arne Vajhøj
It is something for something.
Big Profit Inc. let you use their library in your product
if you pay them royalties for each copy you sell.
GPL project XYZ let you use their library for free
in your product if you give your users access to
source code, rigts to modify and right to redistribute.
It is just rights instead of money.
No it is not, because I've never seen anything about proportions. If
99% of some product is proprietary, and 1% is GPL, then it is absurd for
a vendor to give up 99% of the IP just for 1% of GPL IP.
Why ? It's their product so they get to decide the terms of use
unless they are in some kind of regulated or monopoly market where
external rules dictate the terms under which they can make the
product available.

In this situation you are free to invest the time and resources
to make your own version of the product if you can't find an
alternative. This is what you would have to do anyway if the
GPL product did not exist and you needed the functionality.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Phillip Helbig (undress to reply)
2020-09-29 19:47:47 UTC
Permalink
Post by Dave Froble
Post by Arne Vajhøj
It is something for something.
Big Profit Inc. let you use their library in your product
if you pay them royalties for each copy you sell.
GPL project XYZ let you use their library for free
in your product if you give your users access to
source code, rigts to modify and right to redistribute.
It is just rights instead of money.
No it is not, because I've never seen anything about proportions. If
99% of some product is proprietary, and 1% is GPL, then it is absurd for
a vendor to give up 99% of the IP just for 1% of GPL IP.
If you don't like it, don't do it.

It's like people who complain about how bad Windows is, yet use it
voluntarily. :-|
Stephane Tougard
2020-09-30 13:38:00 UTC
Permalink
Post by Dave Froble
No it is not, because I've never seen anything about proportions. If
99% of some product is proprietary, and 1% is GPL, then it is absurd for
a vendor to give up 99% of the IP just for 1% of GPL IP.
Rewrite the 1% then, it's simpler.

Beside that, no there is nothing about proportion, if your software is
1% GPL, it's 100% GPL. If you don't agree with it, don't use it. There
is a lot of free code out there who is not under GPL.
John Dallman
2020-09-30 18:16:00 UTC
Permalink
There is a lot of free code out there who is not under GPL.
See the BSD licenses, the MIT license, the Apache license, and so on. The
GPL is the only widely-used license that can require you to open-source
code linked against it AFAIK; many of the other licenses cause no
problems at all with their code being linked into commercial software.

John
Arne Vajhøj
2020-09-30 18:26:06 UTC
Permalink
Post by John Dallman
There is a lot of free code out there who is not under GPL.
See the BSD licenses, the MIT license, the Apache license, and so on. The
GPL is the only widely-used license that can require you to open-source
code linked against it AFAIK; many of the other licenses cause no
problems at all with their code being linked into commercial software.
There are not many strong copyleft licenses and GPL is by far the most
widely used.

Number two is probably GPL's twin brother AGPL.

Arne
Arne Vajhøj
2020-09-29 13:43:43 UTC
Permalink
Post by Dave Froble
Post by Simon Clubley
It is extremely unclear whether this use could ever be allowed under
the GPL, even with an interface, without requiring at least some of the
VMS code to be opened up under a GPL-compatible licence.
Am I the only one who thinks the "if you use any of my code, which I've
made public, then I own all your code" is just a bit distasteful?
It does not claim any ownership of your code, but it does
put some restrictions on your licensing.

It is a something for something model. You get source code
but then you also need to provide source code.

There are basically 3 types of open source licenses:

1) permissive licenses like BSD, MIT, Apache etc.

The basic philosophy is "I give you permission to
use my stuff - you can use it for open source or
for closed source as you want and you can keep your
modifications open source or closed source as you want".

There are sometimes a requirement for recognizing the
usage.

2) weak copy left like LGPL

The basic philosophy is "I give you permission to
use my stuff - you can use it for open source or for
closed source as you want, but my stuff stays open source
you have to make your modifications open source and
you need to enable users to upgrade my stuff independently of
yours".

3) strong copy left like GPL

The basic philosophy is "I give you permission to
use my stuff but all products using it must also
be open source including modifications".

Usually (like GPL) the restriction is only in place when you
distribute your product to others - you can do whatever you want
for internal usage.

Arne
Richard Whalen
2020-09-28 13:06:13 UTC
Permalink
Post by Simon Clubley
Post by Arne Vajhøj
However, as a special exception, the source code distributed need not
include anything that is normally distributed (in either source or
binary form) with the major components (compiler, kernel, and so on) of
the operating system on which the executable runs, unless that component
itself accompanies the executable.
This one doesn't work because of the last bit. The VMS kernel will
accompany the GFS2 executables as part of the installation media
and GFS2 will be tied into the VMS kernel itself.
I do not know any more about VSI's plans than anyone else that has viewed the roadmap.
I also am writing my personal, non-legal (I'm not a lawyer) opinion.

What if the GFS code and the code that interfaces it to the VMS kernel were delivered as a layered product, and hence not delivered with the VMS kernel?
Then, the GPL license might apply to the interface code, but not to the code in the VMS kernel that it interfaces to.
Stephen Hoffman
2020-09-24 17:39:52 UTC
Permalink
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any
mention of any linking exceptions or similar.
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/LICENSES/exceptions/GCC-exception-2.0
--
Pure Personal Opinion | HoffmanLabs LLC
Simon Clubley
2020-09-24 17:58:11 UTC
Permalink
Post by Stephen Hoffman
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any
mention of any linking exceptions or similar.
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/LICENSES/exceptions/GCC-exception-2.0
This is the exception that Stephen refers to above:

|SPDX-Exception-Identifier: GCC-exception-2.0
|SPDX-URL: https://spdx.org/licenses/GCC-exception-2.0.html
|SPDX-Licenses: GPL-2.0, GPL-2.0+, GPL-2.0-only, GPL-2.0-or-later
|Usage-Guide:
| This exception is used together with one of the above SPDX-Licenses to
| allow linking the compiled version of code to non GPL compliant code.
| To use this exception add it with the keyword WITH to one of the
| identifiers in the SPDX-Licenses tag:
| SPDX-License-Identifier: <SPDX-License> WITH GCC-exception-2.0
|License-Text:
|
|In addition to the permissions in the GNU Library General Public License,
|the Free Software Foundation gives you unlimited permission to link the
|compiled version of this file into combinations with other programs, and to
|distribute those programs without any restriction coming from the use of
|this file. (The General Public License restrictions do apply in other
|respects; for example, they cover modification of the file, and
|distribution when not linked into another program.)

I'm not convinced that this fully covers the possible use of GFS2 within VMS.

For example, it talks about linking compiled code with non-GPL compliant
code. What if you have to directly integrate the GFS2 source code within
the XQP source code ?

Furthermore, if we look at one specific file, chosen at random, from
this GFS2 source tree that Stephen points to:

https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/fs/gfs2/bmap.c

|/*
| * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
| * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
| *
| * This copyrighted material is made available to anyone wishing to use,
| * modify, copy, or redistribute it subject to the terms and conditions
| * of the GNU General Public License version 2.
| */

That looks to be a pure GPL V2 licence with no linking or other exceptions.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Craig A. Berry
2020-09-24 19:14:38 UTC
Permalink
Post by Simon Clubley
Post by Stephen Hoffman
Post by Simon Clubley
As far as I can tell, GFS2 is GPL, not LGPL, and I can't find any
mention of any linking exceptions or similar.
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/LICENSES/exceptions/GCC-exception-2.0
|SPDX-Exception-Identifier: GCC-exception-2.0
|SPDX-URL: https://spdx.org/licenses/GCC-exception-2.0.html
|SPDX-Licenses: GPL-2.0, GPL-2.0+, GPL-2.0-only, GPL-2.0-or-later
| This exception is used together with one of the above SPDX-Licenses to
| allow linking the compiled version of code to non GPL compliant code.
| To use this exception add it with the keyword WITH to one of the
| SPDX-License-Identifier: <SPDX-License> WITH GCC-exception-2.0
|
|In addition to the permissions in the GNU Library General Public License,
|the Free Software Foundation gives you unlimited permission to link the
|compiled version of this file into combinations with other programs, and to
|distribute those programs without any restriction coming from the use of
|this file. (The General Public License restrictions do apply in other
|respects; for example, they cover modification of the file, and
|distribution when not linked into another program.)
I'm not convinced that this fully covers the possible use of GFS2 within VMS.
For example, it talks about linking compiled code with non-GPL compliant
code. What if you have to directly integrate the GFS2 source code within
the XQP source code ?
Furthermore, if we look at one specific file, chosen at random, from
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/fs/gfs2/bmap.c
|/*
| * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
| * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
| *
| * This copyrighted material is made available to anyone wishing to use,
| * modify, copy, or redistribute it subject to the terms and conditions
| * of the GNU General Public License version 2.
| */
That looks to be a pure GPL V2 licence with no linking or other exceptions.
Right. I was poking around earlier in the same directory tree Hoff
pointed to and it has boilerplate for all sorts of licenses, including 3
kinds of BSD and MIT. But I don't think that means you can pick
whatever license you want. As far as I can tell, you would have to find
the string "SPDX-Exception-Identifier: GCC-exception-2.0" in each and
every source file you need in order for that exception to apply. But
instead what I seem to be finding from a couple of random samples looks
more like:

// SPDX-License-Identifier: GPL-2.0
/*
* cacheinfo support - processor cache information via sysfs
*
* Based on arch/x86/kernel/cpu/intel_cacheinfo.c
* Author: Sudeep Holla <***@arm.com>
*/
Simon Clubley
2020-09-25 12:36:25 UTC
Permalink
Post by Craig A. Berry
Post by Simon Clubley
Furthermore, if we look at one specific file, chosen at random, from
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/fs/gfs2/bmap.c
|/*
| * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
| * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
| *
| * This copyrighted material is made available to anyone wishing to use,
| * modify, copy, or redistribute it subject to the terms and conditions
| * of the GNU General Public License version 2.
| */
That looks to be a pure GPL V2 licence with no linking or other exceptions.
Right. I was poking around earlier in the same directory tree Hoff
pointed to and it has boilerplate for all sorts of licenses, including 3
kinds of BSD and MIT. But I don't think that means you can pick
whatever license you want. As far as I can tell, you would have to find
the string "SPDX-Exception-Identifier: GCC-exception-2.0" in each and
every source file you need in order for that exception to apply. But
instead what I seem to be finding from a couple of random samples looks
// SPDX-License-Identifier: GPL-2.0
/*
* cacheinfo support - processor cache information via sysfs
*
* Based on arch/x86/kernel/cpu/intel_cacheinfo.c
*/
Yes, that's why I was puzzled by the licence Stephen posted.

I thought he was seeing something I had missed, but the sampling of
files I have looked at all have the pure GPL V2 notice at the top
of the file.

I really hope that VSI have come up with a way to isolate the GFS2
source code from the rest of VMS in a way that's acceptable to the
owners of the GFS2 source code so that we don't get into GPL arguments
about the rest of the VMS kernel.

And you are correct, just because there's a set of licences in a common
directory, that doesn't mean you can automatically apply whatever licence
you choose to whatever source code you wish.

It's what is at the top of each source file that matters and the
only reason there's a collection of licences in a common directory
somewhere is because not all source files in a tree may have the
same licence.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Walking destinations on a map are further away than they appear.
Chris
2020-09-29 00:10:14 UTC
Permalink
Post by Simon Clubley
Post by Craig A. Berry
Post by Simon Clubley
Furthermore, if we look at one specific file, chosen at random, from
https://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git/tree/fs/gfs2/bmap.c
|/*
| * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
| * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
| *
| * This copyrighted material is made available to anyone wishing to use,
| * modify, copy, or redistribute it subject to the terms and conditions
| * of the GNU General Public License version 2.
| */
That looks to be a pure GPL V2 licence with no linking or other exceptions.
Right. I was poking around earlier in the same directory tree Hoff
pointed to and it has boilerplate for all sorts of licenses, including 3
kinds of BSD and MIT. But I don't think that means you can pick
whatever license you want. As far as I can tell, you would have to find
the string "SPDX-Exception-Identifier: GCC-exception-2.0" in each and
every source file you need in order for that exception to apply. But
instead what I seem to be finding from a couple of random samples looks
// SPDX-License-Identifier: GPL-2.0
/*
* cacheinfo support - processor cache information via sysfs
*
* Based on arch/x86/kernel/cpu/intel_cacheinfo.c
*/
Yes, that's why I was puzzled by the licence Stephen posted.
I thought he was seeing something I had missed, but the sampling of
files I have looked at all have the pure GPL V2 notice at the top
of the file.
I really hope that VSI have come up with a way to isolate the GFS2
source code from the rest of VMS in a way that's acceptable to the
owners of the GFS2 source code so that we don't get into GPL arguments
about the rest of the VMS kernel.
And you are correct, just because there's a set of licences in a common
directory, that doesn't mean you can automatically apply whatever licence
you choose to whatever source code you wish.
It's what is at the top of each source file that matters and the
only reason there's a collection of licences in a common directory
somewhere is because not all source files in a tree may have the
same licence.
Simon.
Some people make a lot of fuss abut licensing violations, a sort of
professional ethics situation, but if a vendor tries to side step or
play fast and loose with with other's licensing conditions, would you
trust them about anything else ?. Trust is everything in business etc..

Commercial vendors, who slagged off open source as being amateur for
years, have meantime used it to make a lot of money. Some have
put something back into the pot, but many have not...

Chris





and most people stay
honest and do the right thing.
Loading...