"Open Core" and the Ada Community

"Open Core" and the Ada Community

[David Botton]

This article is written by Bradley M. Kuhn who is in charge of enforcing the GPL for the Linux Kernel

http://ebb.org/bkuhn/blog/2009/10/16/open-core-shareware.html


My own words:

The spirit of GPL has long since died with corporations using the GPL to restrict freedoms for corporate gains and to cripple new development and innovation.

There needs to be an Open Source Code of Ethics. A clause added to the BSD/MIT license saying using this software binds you to the shame clause.

I no longer see a point in choosing GPL over BSD/MIT licenses. Those that have a community ethic will act accordingly and be good citizens of open source, the rest use the GPL as a weapon to harm the community.

May as well open the door and remove blocks from our code and the honest be honest and the abusers be shamed.

David Botton

Re: "Open Core" and the Ada Community

[Dirk Heinrichs]

David Botton wrote:

> The spirit of GPL has long since died with corporations using the GPL to
> restrict freedoms for corporate gains and to cripple new development and
> innovation.

I don't think it has. The essential point in Kuhn's post is that people can 
still fork the GPL-licensed codebase and re-implement the proprietary parts 
they need themselves. That's still impossible with a BSD licensed codebase 
that has been turned proprietary.

So there's no crippling of new development and innovation at all. It's just 
a bit slowed down ;)

I'm still wondering about the "...and the Ada Community" part of the 
subject, though.

Bye...

	Dirk
-- 
Dirk Heinrichs <dirk.heinrichs@altum.de>
GPG Public Key CB614542 | Jabber: dirk.heinrichs@altum.de
Tox: heini@toxme.se
Sichere Internetkommunikation: http://www.retroshare.org
Privacy Handbuch: https://www.privacy-handbuch.de

Re: "Open Core" and the Ada Community

[David Botton]

<<I don't think it has. The essential point in Kuhn's post is that people can 
still fork the GPL-licensed codebase and re-implement the proprietary parts 
they need themselves. That's still impossible with a BSD licensed codebase 
that has been turned proprietary.>>

No that was not the essential point of the article at all. It was that despite being GPL it is possible to use the GPL in ways that restric others rights.

His "essential point" was that developers should not allow homogeonization of the copyright to help reduce the risks of that. "start contributing to the code with your own copyrights, licensed under GPL"

The Ada community has seen claims of full rights for AdaCore to do as they please and enable and empower their claims to various licensing scenarios they use.

<<So there's no crippling of new development and innovation at all. It's just 
a bit slowed down ;)>>

When developers know their work will be encumbered for use by others in licensing games or otherwise the "core" has been booby trapped they hold back from development and innovation. It is not just slowed down it is often stopped or never started.

<<Bye...>>
I am a proponent and author of open source for Ada, one of the few remaining. Someone has to care and speak up regularly. Things can not get swept under the rug.

David Botton

Re: "Open Core" and the Ada Community

[David Botton]

<<I'm still wondering about the "...and the Ada Community" part of the 
subject, though.>>

His words and what will be our response:

"The software freedom world is more complex than it once was. Ten years ago, we advocates could tell people to look for the GPL label and know that the software would automatically be part of a freedom-friendly, software sharing community. Not all GPL'd software is created equal anymore" ... "The new advice is: judge the freedom of your codebase not only on its license, but also on the diversity of the community that contributes to it."

So as an Ada community, what are we doing to turn Ada in to a "freedom-friendly, software sharing community." 

AdaCore has taken some very small positive steps with the move of some software to github. Certainly it will be hard for them to play make pretend and say you have to ignore the headers and the runtime linking exception is void in this case.

What positive steps can we do to for FSF GNAT? Placing it on footing and raising awareness I've done with GetAdaNow.com and the mascot, etc.

However we need a real team taking charge of a fork of FSF GNAT and a foundation to become "true" stewards of the full freedoms GNAT version.

I am the wrong person to do the fork or would have done it already. My skills lay in other areas of SE then compilers. Are there others willing to get involved with such a fork?

David Botton

Re: "Open Core" and the Ada Community

[Florian Weimer]

* David Botton:

> I no longer see a point in choosing GPL over BSD/MIT licenses. Those
> that have a community ethic will act accordingly and be good citizens
> of open source, the rest use the GPL as a weapon to harm the
> community.

This doesn't work if the community owns the copyright.

The immediate problem often is copyright assignment with asymmetric
licensing terms to some central party.  But I don't think this is what
happens in the cases you are so concerned about because the assignment
is either to the wrong party, or no assignment takes place at all.

The real problem is a lack of a diverse development community, and no
interest on the side of the primary (or original) contributor/project
steward to create a such a community (which is somewhat understandable
because it as a lot of work).  Once there is an actual development
community that brings value to the primary contributor, licensing
matters less because power is shared in other ways.

(With “development community” I mean people collaborating to improve
the software itself, not merely using it.)

In the general case, this is not related to “open core” at all.
People will call your free software “open core” even if it is *always*
GPL (even for paying customers), and all features are available in
public versions downloadable for free.

Re: "Open Core" and the Ada Community

[Simon Wright]

David Botton <david@botton.com> writes:

> However we need a real team taking charge of a fork of FSF GNAT and a
> foundation to become "true" stewards of the full freedoms GNAT
> version.

I don't see the point of this, really. FSF GNAT has the freedoms you
need. The things that don't have the freedoms are some of the libraries;
xmlada (though an unrestricted source is available), gnatcoll, asis.

It's annoying that there's no public repo for gprbuild, but the annual
GPL releases are pretty reliable (with the odd patch).

It can take a while for changes to work their way through the system,
especially when the subject is not on the maintainers' (i.e.AdaCore's)
main line: for example, GCC 6 currently allows us to compile a
restricted runtime (i.e., embedded) that supports finalization but not
exceptions, but doesn't let us bind it[1].

I'm not sure how long it takes for an AdaCore employee to get to grips
with GNAT compiler internals; looks pretty daunting to me. Tinkering
with tools and libraries is more my thing!

[1] https://gcc.gnu.org/ml/gcc-patches/2015-12/msg01951.html

Re: "Open Core" and the Ada Community

[Jacob Sparre Andersen]

Florian Weimer wrote:

> The real problem is a lack of a diverse development community, and no
> interest on the side of the primary (or original) contributor/project
> steward to create a such a community (which is somewhat understandable
> because it as a lot of work).  Once there is an actual development
> community that brings value to the primary contributor, licensing
> matters less because power is shared in other ways.
>
> (With “development community” I mean people collaborating to improve
> the software itself, not merely using it.)

You are probably right about this.  If we can manage to get there with
GNAT (or another Ada compiler), I can't see how the specific choice of
Open Source licence will matter much, since there will be a real cost
(lost man-power) if one walls in the development.

But who has the spare time to learn enough to contribute to GNAT?  Or to
write an Ada compiler from scratch?

How large is GNAT?  I've been told that learning to find your way around
in a (specific) 2M SLOC Ada application takes about a year (full time).
How long time does it take to learn to find your way around in GNAT?

Greetings,

Jacob
-- 
"Banning open source would have immediate, broad, and
 strongly negative impacts on the ability of many sensitive
 and security-focused DOD groups to protect themselves
 against cyberattacks"                        -- Mitre Corp.

Re: "Open Core" and the Ada Community

[Dirk Heinrichs]

David Botton wrote:

> His "essential point" was that developers should not allow homogeonization
> of the copyright to help reduce the risks of that. "start contributing to
> the code with your own copyrights, licensed under GPL"

Hmm, don't know if that would apply to german people at all, since copyright 
is more or less a US thing. In germany, we have the "Urheberrecht", which 
could be translated as "creators right". And this, per definition, is not 
re-assignable.

Bye...

	Dirk
-- 
Dirk Heinrichs <dirk.heinrichs@altum.de>
GPG Public Key CB614542 | Jabber: dirk.heinrichs@altum.de
Tox: heini@toxme.se
Sichere Internetkommunikation: http://www.retroshare.org
Privacy Handbuch: https://www.privacy-handbuch.de

Re: "Open Core" and the Ada Community

[Florian Weimer]

* Jacob Sparre Andersen:

> You are probably right about this.  If we can manage to get there with
> GNAT (or another Ada compiler), I can't see how the specific choice of
> Open Source licence will matter much, since there will be a real cost
> (lost man-power) if one walls in the development.
>
> But who has the spare time to learn enough to contribute to GNAT?

It's a strategic decision to go down that right, it doesn't have to be
spare time.

> How large is GNAT?

The front end and build tools appear to be around 360 kSLOC of Ada,
plus 240 kSLOC for the libraries (which includes some 10 kSLOC of C
sources).  It's all in one directory, so I had to use file-name-based
heuristics to separate the library sources.

The interface to the GCC middle-end is 20 kSLOC of C++.

The GCC middle-end itself is a further 650 KSLOC of C++.  Back-ends
vary in size from 10 kSLOC to 90 kSLOC (the latter is the one for i386
and x86_64, unfortunately).

GCC is fairly modular.  It is possible to make useful changes even if
you do not know exactly what you are doing.  There's code review
before you are allowed to commit anything.

Re: "Open Core" and the Ada Community

[Luke A. Guest]

Florian Weimer <fw@deneb.enyo.de> wrote:

> GCC is fairly modular.  It is possible to make useful changes even if
> you do not know exactly what you are doing.  There's code review
> before you are allowed to commit anything.
> 

Expect zero help or at most a sentence which comes across as sarcasm from
the guys who actually know this stuff.

Re: "Open Core" and the Ada Community

[Florian Weimer]

* Luke A. Guest:

> Florian Weimer <fw@deneb.enyo.de> wrote:
>
>> GCC is fairly modular.  It is possible to make useful changes even if
>> you do not know exactly what you are doing.  There's code review
>> before you are allowed to commit anything.
>
> Expect zero help or at most a sentence which comes across as sarcasm from
> the guys who actually know this stuff.

For the changes I wanted to get in, the process was a bit tedious, but
in the end the requested changes were improvements.

Re: "Open Core" and the Ada Community

[Paul Rubin]

Florian Weimer <fw@deneb.enyo.de> writes:
> For the changes I wanted to get in, the process was a bit tedious, but
> in the end the requested changes were improvements.

I haven't had serious trouble either, and I got reasonable advice from
the developers on the IRC channel.