Release testing and the relationship between 'bzr selftest' and plugins

[mbp]

On 27 March 2012 02:49, Jelmer Vernooij <jelmer@...> wrote:
>>     > Shipping the plugins with core also has a few issues:
>>
>>     >  * lp:bzr  (AFAIK) falls under the contributor agreement, and all code
>>     > (with some exceptions) is (C) Canonical. Most plugins have different
>>     > copyright holders.
>>
>> Hmm, this one is hard.

This was a big thread, and I did not have time to read it all properly
yet, but I will answer this one:

This code already exists, in separate trees.  We distribute some of
them in .debs or (?) in installers.  We're willing to ship non-CA/CLA
code.

We also have code in the main tree in eg bzrlib/util that's
partitioned from bzr itself and under different copyright and/or a
different (compatible) licence.

"core plugins" in bzrlib/plugins/ are clearly separate, and just a
different tactic for distributing plugin code.

Having a copyright steward still makes sense for core code and for
plugins that are primarily developed by Canonical.  Distributing
plugins not under CLA/CA also makes sense.

If we copy code from one project to another, eg from a plugin to the
core or vice versa, we have to be sure we have legal permission to do
that, but it's never been a practical problem.  Specifically if there
is code from a plugin that ought to be in core we would need the
author at that point to do the CLA, they haven't already.

Adding all these up: if it makes technical sense for them to be core
plugins in the main tree, copyright isn't a reason not to merge them.

--
Martin

[David Ingamells-2]

I fail to see the connection between copyright policies and stupid
changes. IMHO there exists a huge amount of copyrighted code in the
internet that is quite stupid. Legal "caveat emptor" protections are
different but he was talking about copyright.

On 02/04/12 06:21, Stephen J. Turnbull wrote:

> On Fri, Mar 30, 2012 at 8:52 PM, Ben Finney<ben+bazaar@...>  wrote:
>
>> Please, Bazaar developers, convince Canonical that your project is being
>> hurt by this silliness, and move to the “inbound == outbound” copyright
>> policy of the vast majority of successful free-software projects
>> <URL:https://lwn.net/Articles/414051/>.
> That's a *terrible* reason for do anything.  "If  two million penguins
> do a stupid thing, it is still a stupid thing."  Most free software
> projects, successful or not, (a) simply are not mission-critical to
> anybody, and (b) simply do not want to deal with legal issues and
> generally try to pretend that they'll go away if ignored (even in the
> case of XEmacs, which is downstream of the most rigidly policed legal
> policies in free software).
>
> Whether this particular change is stupid, of course, is going to be up
> to the judgment of the Canonical legal staff.
>
> The right reason for making this change is that optional plugins are,
> well, optional plugins, the core runs perfectly well without them, and
> the direction of derivation and therefore the legal issues are thus
> separable from the core.  That argument might even carry some weight
> with Mark and the Canonical leagle beagles.
>
> Regards,
> Steve
>
>

--

*David Ingamells * Test Architect *Mapscape *

Henri Wijnmalenweg 5b  | 5657 EP Eindhoven  | The Netherlands
Phone [phone]  | Fax [phone]
www.mapscape.eu <http://www.mapscape.eu/>

Mapscape B.V. is ISO9001:2008 certified.
This e-mail and any attachment may contain corporate proprietary
information and may only be read, copied and used by the intended
recipient. If you have received it by mistake, please notify us
immediately by reply e-mail and delete this e-mail and its attachments
from your system. We believe but do not warrant that this message and
any attachments are virus free.
Mapscape B.V. is registered at the Kamer van Koophandel Oost-Brabant
located in Eindhoven, The Netherlands, number 17210210.

[Ben Finney-11]

"Stephen J. Turnbull" <stephen@...> writes:

> On Fri, Mar 30, 2012 at 8:52 PM, Ben Finney <ben+bazaar@...> wrote:
>
> > Please, Bazaar developers, convince Canonical that your project is
> > being hurt by this silliness, and move to the “inbound == outbound”
> > copyright policy of the vast majority of successful free-software
> > projects <URL:https://lwn.net/Articles/414051/>.
>
> That's a *terrible* reason for do anything. "If two million penguins
> do a stupid thing, it is still a stupid thing."

The paragraph you chose to quote, above, was not giving reason. The
reason given was in most of the message that you snipped and chose not
to respond to.

To put it more briefly, then: Code which people want in Bazaar, which
Canonical already has all the legal permission they need to include and
distribute, is under a cloud of uncertainly solely because of this
self-imposed obstacle.

That's an argument for removing the obstacle, and it's one which the
Bazaar developers, as people being inconvenienced by this policy and
with presumably more influence than outsiders, are in a position to
present to Canonical.

--
 \       “If you go flying back through time and you see somebody else |
  `\   flying forward into the future, it's probably best to avoid eye |
_o__)                                           contact.” —Jack Handey |
Ben Finney

[Stephen J. Turnbull]

On Mon, Apr 2, 2012 at 5:02 PM, Ben Finney <ben+bazaar@...> wrote:

> The paragraph you chose to quote, above, was not giving reason. The
> reason given was in most of the message that you snipped and chose not
> to respond to.

And I'm sure that the Canonical developers know those reasons, if only
because you've been at pains to explain them on several occasions
before.  But it's the lawyers, or their boss, that you need to
convince, not the developers.

> To put it more briefly, then: Code which people want in Bazaar, which
> Canonical already has all the legal permission they need to include and
> distribute,

Granted.  The lawyers think Canonical needs more than permission; why,
I don't know.  I am familiar, however, with the "risk to the project"
rationale that RMS puts forward, and I posted a few words to address
that.  Your arguments do not.

> is under a cloud of uncertainly

Nonsense.  Under an assignment or contributor agreement policy, it's
clear what the status of that code is: not distributed by the project
as part of the core product but you're free to get it yourself (and
may even be distributed as Ubuntu packages in many cases).   That's
true of a lot of code. Sure, to some extent such a policy is a PITA,
and arguably, it's a bad idea.  But unclear, it is not.

> solely because of this self-imposed obstacle.

A self-imposed obstacle which the seminal legal minds in free software
(ie, RMS and his lawyer) have also chosen to impose on several core
GNU projects.  Many other projects (pretty much any project big enough
to have a $PROJECT Foundation, and not a few smaller than that) also
require either an assignment or a contributor agreement.

> That's an argument for removing the obstacle, and it's one which the
> Bazaar developers, as people being inconvenienced by this policy and
> with presumably more influence than outsiders, are in a position to
> present to Canonical.

Sure, and I know that some Canonical employees have at the very least
presented the argument to their bosses, with little effect at the
time.  As far as I know, on this list you're singing to the preacher
(or however that old saw goes).  My point is simply that talking about
what unlawyered projects do isn't going to help the case, nor will
overstating the valid points that you do make.  You should avoid that.

[Stefan Monnier]

>> > A few others that come to mind (I haven't checked them against the
>> > rationale you give above though):
>> > - bzr-xmloutput,
>> This has non-Canonical copyright holders.
>> > - bzr-search,
>> Has non-Canonical copyright holders.
>> > - bzr-rebase ?
>> bzr-rebase was renamed to bzr-rewrite. It has non-Canonical copyright
>> holders.
> The only reason “has non-Canonical copyright holders” is mentioned here
> is, I assume, because of Canonical's stubborn insistence on having a
> specially privileged position over other contributors to the Bazaar code
> base <URL:https://lwn.net/Articles/359013/>.

FWIW, getting copyright agreements signed is not always as hard as some
people think.


        Stefan

[John Yates-6]

Stefan Nonnier wrote:

> FWIW, getting copyright agreements signed is not always as hard as some
people think.

As far as I can tell if you work at IBM it is nearly impossible.  Yes,
they understand open source and support contributing thereto.  But "giving
away" IP to which they have a claim?  Ain't gonna happen!

/john

[Stefan Monnier]

>> FWIW, getting copyright agreements signed is not always as hard as
>> some people think.
> As far as I can tell if you work at IBM it is nearly impossible.

The two statements are not mutually exclusive.

> Yes,  they understand open source and support contributing thereto.
> But "giving  away" IP to which they have a claim?  Ain't gonna happen!

"grep IBM copyright.list" on fencepost.gnu.org suggests that it's at
least not completely impossible (of course, it doesn't mean it's easy
either).


        Stefan