Gitbox

Gitbox

[Daniel Searles]

Hi All,

I'm curious if Gitbox is in violation of the Git's GPL.
http://www.gitboxapp.com/

Thank you,
Daniel Searles
Tech Support
Media Lab, Inc.

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 11:16 AM, Daniel Searles wrote:

> I'm curious if Gitbox is in violation of the Git's GPL.
> http://www.gitboxapp.com/

 From <http://www.gitboxapp.com/faq.html#gpl>:

"The application package contains a compiled official version 1.7.3.2  
of the Git content tracker. Gitbox neither links (statically or  
dynamically) against Git nor uses custom interfaces to interact with  
it. Git binaries are provided for your convenience only. You can  
replace  bundled binaries with any compatible build of the current  
version of Git or newer."

That appears to comply with both the letter and spirit of the GPL.

Josh

Re: Gitbox

[Drew Northup]

On Wed, 2011-04-13 at 11:16 -0700, Daniel Searles wrote:
> Hi All,
> 
> I'm curious if Gitbox is in violation of the Git's GPL.
> http://www.gitboxapp.com/
> 
> Thank you,
> Daniel Searles
> Tech Support
> Media Lab, Inc.

Well, it appears that he may have implemented "git" on his own, or at
the very least he's trying to imply that he did.

http://www.gitboxapp.com/faq.html#gpl

Perhaps he'll have a more enlightening reply (please use reply all).

-- 
-Drew Northup
________________________________________________
"As opposed to vegetable or mineral error?"
-John Pescatore, SANS NewsBites Vol. 12 Num. 59

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 11:59 AM, Drew Northup wrote:

> On Wed, 2011-04-13 at 11:16 -0700, Daniel Searles wrote:
>> Hi All,
>>
>> I'm curious if Gitbox is in violation of the Git's GPL.
>> http://www.gitboxapp.com/
>>
>> Thank you,
>> Daniel Searles
>> Tech Support
>> Media Lab, Inc.
>
> Well, it appears that he may have implemented "git" on his own, or at
> the very least he's trying to imply that he did.
>
> http://www.gitboxapp.com/faq.html#gpl

He very clearly says that he ships Git, not a reimplementation.

Josh

Re: Gitbox

[Drew Northup]

On Wed, 2011-04-13 at 12:09 -0700, Joshua Juran wrote:
> On Apr 13, 2011, at 11:59 AM, Drew Northup wrote:
> 
> > On Wed, 2011-04-13 at 11:16 -0700, Daniel Searles wrote:
> >> Hi All,
> >>
> >> I'm curious if Gitbox is in violation of the Git's GPL.
> >> http://www.gitboxapp.com/
> >>
> >> Thank you,
> >> Daniel Searles
> >> Tech Support
> >> Media Lab, Inc.
> >
> > Well, it appears that he may have implemented "git" on his own, or at
> > the very least he's trying to imply that he did.
> >
> > http://www.gitboxapp.com/faq.html#gpl
> 
> He very clearly says that he ships Git, not a reimplementation.

He also clearly stated that he's not calling git command-line tools...

"Gitbox neither links (statically or dynamically) against Git nor uses
custom interfaces to interact with it. Git binaries are provided for
your convenience only."

I'm hoping he'll have a more complete statement for all of us.

-- 
-Drew Northup
________________________________________________
"As opposed to vegetable or mineral error?"
-John Pescatore, SANS NewsBites Vol. 12 Num. 59

Re: Gitbox

[Jakub Narebski]

Daniel Searles <dsearles@medialab.com> writes:

> Hi All,
> 
> I'm curious if Gitbox is in violation of the Git's GPL.
> http://www.gitboxapp.com/

It is graphical _interface_ for Git; I assume that it _uses_ Git.
Using a program, as opposed to including it (making derivative work),
is not covered by GPLv2.

It's not only proprietary Git GUI: there is also Sprout, SmartGit,
SourceTree (I think: my information might be outdated).
-- 
Jakub Narebski
Poland
ShadeHawk on #git

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 12:16 PM, Drew Northup wrote:

> On Wed, 2011-04-13 at 12:09 -0700, Joshua Juran wrote:
>
>> On Apr 13, 2011, at 11:59 AM, Drew Northup wrote:
>>
>>> On Wed, 2011-04-13 at 11:16 -0700, Daniel Searles wrote:
>>
>>>
>>>> I'm curious if Gitbox is in violation of the Git's GPL.
>>>> http://www.gitboxapp.com/
>>>
>>> Well, it appears that he may have implemented "git" on his own, or  
>>> at
>>> the very least he's trying to imply that he did.
>>>
>>> http://www.gitboxapp.com/faq.html#gpl
>>
>> He very clearly says that he ships Git, not a reimplementation.
>
> He also clearly stated that he's not calling git command-line tools...

I wonder why he would ship them, then.

> "Gitbox neither links (statically or dynamically) against Git nor uses
> custom interfaces to interact with it. Git binaries are provided for
> your convenience only."

Calling execve() is not dynamic linking, that I'm aware of.

> I'm hoping he'll have a more complete statement for all of us.

I think you've misinterpreted his words, but I agree that a  
clarification to set minds at ease would be appropriate.

Josh

Re: Gitbox

[Randal L. Schwartz]

>>>>> "Drew" == Drew Northup <drew.northup@maine.edu> writes:

Drew> He also clearly stated that he's not calling git command-line tools...

Drew> "Gitbox neither links (statically or dynamically) against Git nor uses
Drew> custom interfaces to interact with it. Git binaries are provided for
Drew> your convenience only."

Huh?  How do you get "not using git CLI" there?  He's not using *custom*
interfaces, but the *standard* interfaces are more than enough.

He's made new porcelein, and using the plumbing directly.  That's
exactly why we have plumbing in git, right?

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.posterous.com/ for Smalltalk discussion

Re: Gitbox

[Drew Northup]

On Wed, 2011-04-13 at 12:41 -0700, Randal L. Schwartz wrote:
> >>>>> "Drew" == Drew Northup <drew.northup@maine.edu> writes:
> 
> Drew> He also clearly stated that he's not calling git command-line tools...
> 
> Drew> "Gitbox neither links (statically or dynamically) against Git nor uses
> Drew> custom interfaces to interact with it. Git binaries are provided for
> Drew> your convenience only."
> 
> Huh?  How do you get "not using git CLI" there?  He's not using *custom*
> interfaces, but the *standard* interfaces are more than enough.
> 
> He's made new porcelein, and using the plumbing directly.  That's
> exactly why we have plumbing in git, right?

If that's what he's doing then why does he state that the git binaries
are "for your convenience only"--strongly implying his program will work
in the absence thereof? I'm hoping he just made a typo--as what he said
is there in plain English, and is somewhat contradicted by the next
sentence. If he's indeed using the plumbing then the binaries are
required--which is indeed what they are there for. 

I think that I've already proved myself to be one of the list's resident
grammar nazis. If you wish to dispute my interpretation of his grammar
please contact me directly and don't bother the rest of the list.

I suspect that when he gets a chance he'll be able to clarify what he's
doing well enough on his own.

-- 
-Drew Northup
________________________________________________
"As opposed to vegetable or mineral error?"
-John Pescatore, SANS NewsBites Vol. 12 Num. 59

Re: Gitbox

[Taylor Hedberg]

Drew Northup, Wed, Apr 13, 2011 at 04:04:49PM -0400:
> If that's what he's doing then why does he state that the git binaries
> are "for your convenience only"--strongly implying his program will
> work in the absence thereof? I'm hoping he just made a typo--as what
> he said is there in plain English, and is somewhat contradicted by the
> next sentence. If he's indeed using the plumbing then the binaries are
> required--which is indeed what they are there for. 

I didn't detect that implication at all. I took it to mean that he is
distributing Git alongside his frontend because it would not function
properly without it. It's a "convenience" in the sense that if he didn't
bundle the Git binaries, you'd have to install them separately in order
to use his product. Why would he bother to include Git itself if the
product worked just fine without it?


> I think that I've already proved myself to be one of the list's
> resident grammar nazis. If you wish to dispute my interpretation of
> his grammar please contact me directly and don't bother the rest of
> the list.

Sending this to the list anyway, because it seems germane to the overall
discussion.

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 1:04 PM, Drew Northup wrote:

> On Wed, 2011-04-13 at 12:41 -0700, Randal L. Schwartz wrote:
>>>>>>> "Drew" == Drew Northup <drew.northup@maine.edu> writes:
>>
>> Drew> He also clearly stated that he's not calling git command-line  
>> tools...
>>
>> Drew> "Gitbox neither links (statically or dynamically) against Git  
>> nor uses
>> Drew> custom interfaces to interact with it. Git binaries are  
>> provided for
>> Drew> your convenience only."
>>
>> Huh?  How do you get "not using git CLI" there?  He's not using  
>> *custom*
>> interfaces, but the *standard* interfaces are more than enough.
>>
>> He's made new porcelein, and using the plumbing directly.  That's
>> exactly why we have plumbing in git, right?
>
> If that's what he's doing then why does he state that the git binaries
> are "for your convenience only"

I suspect he considers it convenient to download the app and use it  
immediately without having to install Git manually, but wants to  
clarify that GitBox works with later versions as well.  In other  
words, Git 1.7.3.2 is provided for *convenience*, not due to  
compatibility restrictions.

> --strongly implying his program will work
> in the absence thereof?

Strongly implying?  I read the exact opposite -- GitBox clearly  
depends on an external git binary.

> I'm hoping he just made a typo--as what he said
> is there in plain English, and is somewhat contradicted by the next
> sentence.

It's ambiguous.  The next sentence clarifies it, ruling out one  
interpretation and therefore confirming the other.  In regular  
expressions this is called backtracking.

> If you wish to dispute my interpretation of his grammar
> please contact me directly and don't bother the rest of the list.

I decline your request; I think it's better to have a public record of  
the issue being settled, if for no other reason than that others don't  
waste time making points that I've already made.

Josh

Re: Gitbox

[Junio C Hamano]

merlyn@stonehenge.com (Randal L. Schwartz) writes:

>>>>>> "Drew" == Drew Northup <drew.northup@maine.edu> writes:
>
> Drew> He also clearly stated that he's not calling git command-line tools...
>
> Drew> "Gitbox neither links (statically or dynamically) against Git nor uses
> Drew> custom interfaces to interact with it. Git binaries are provided for
> Drew> your convenience only."
>
> Huh?  How do you get "not using git CLI" there?  He's not using *custom*
> interfaces, but the *standard* interfaces are more than enough.
>
> He's made new porcelein, and using the plumbing directly.  That's
> exactly why we have plumbing in git, right?

That agrees with my reading of the page.

GPLv2 section 3 (c) allows distribution of compiled executable without
source by only pointing at the place the distributor got an offer to the
source code when the distributor got the compiled executable it is passing
on, and that is quite close to (but not exactly is --- I think the
distributor compiled the executable, not passing on) what the Gitbox site
is doing (i.e. distributing executable without distributing the source).

However, 3 (c) is allowed only for noncommercial distributor that passes
compiled executable along from its upstream source, so that part may need
to be fixed, I suspect, to comply with section 3 by going in either
subsections 3 (a) or 3 (b) route.

IANAL of course.

Re: Gitbox

[Randal L. Schwartz]

>>>>> "Daniel" == Daniel Searles <dsearles@medialab.com> writes:

Daniel> It seems to me that since gitbox is useless without git it is
Daniel> definitely in violation of the GPL. My interpretation of the GPL
Daniel> is based off of the following article:
Daniel> http://clisp.cvs.sourceforge.net/viewvc/clisp/clisp/doc/Why-CLISP-is-under-GPL

And note the author of that opinion, RMS, who would always rule in favor
of more things needing to be under the GPL as the One True License.

And it's also about readline.a, which *has* to be linked into a binary
to make it work.  None of the git code is being *linked* in.

Particularly, I see this phrase in the actual license of Git (version 2,
not version 3 under which Readline is distributed):

    Thus, it is not the intent of this section to claim rights or contest
    your rights to work written entirely by you; rather, the intent is to
    exercise the right to control the distribution of derivative or
    collective works based on the Program.

    In addition, mere aggregation of another work not based on the Program
    with the Program (or with a work based on the Program) on a volume of
    a storage or distribution medium does not bring the other work under
    the scope of this License.

I'd consider the git binary distribution to be an aggregation with
gitbox, not a "linking", presuming that one of the following has also
happened:

      3. You may copy and distribute the Program (or a work based on it,
    under Section 2) in object code or executable form under the terms of
    Sections 1 and 2 above provided that you also do one of the following:

        a) Accompany it with the complete corresponding machine-readable
        source code, which must be distributed under the terms of Sections
        1 and 2 above on a medium customarily used for software interchange;
        or,

        b) Accompany it with a written offer, valid for at least three
        years, to give any third party, for a charge no more than your
        cost of physically performing source distribution, a complete
        machine-readable copy of the corresponding source code, to be
        distributed under the terms of Sections 1 and 2 above on a medium
        customarily used for software interchange; or,

        c) Accompany it with the information you received as to the offer
        to distribute corresponding source code.  (This alternative is
        allowed only for noncommercial distribution and only if you
        received the program in object code or executable form with such
        an offer, in accord with Subsection b above.)

I cannot confirm that one of these has been done, but let's say it has.

As already said in this thread, if "execve()" is considered a "linking",
then *everything* in a typical Linux distro would *have* to be GPL.
That's patently not the case.

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.posterous.com/ for Smalltalk discussion

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 3:17 PM, Daniel Searles wrote:

> I apologize for the empty email. My SMTP server is having problems.  
> That is why I'm going to use this email address to write to the list.

Gah, stop doing that!  :-)

> I feel that gitbox and perhaps other commercial tools for git are in  
> violation of the GPL simply since they rely on git in order to be  
> useful. Take git away from gitbox and it serves no purpose. The  
> thread in the following link goes into depth with regards to a  
> program relying on code that is under a GPL license. The conclusion  
> made in the thread may be due to an older version of the GPL. Could  
> it be that gitbox isn't in violation of the GPL since git uses GPL  
> version 2.0 exclusively?
>
> http://clisp.cvs.sourceforge.net/viewvc/clisp/clisp/doc/Why-CLISP-is-under-GPL


My shell scripts that call git are also useless without Git.  Am I not  
allowed to distribute my scripts under non-GPL terms?  (And what about  
a script that calls the nonexistent blort utility, and then someone  
writes blort and distributes it under the GPL?)  As I understand it,  
the GPL (and the thread you mention) address linking, not utility.

RMS seemed to win this debate by arguing that his position was better  
for the community.  I have colleagues who understand that Git is the  
best VCS available but are concerned about having to train users who  
are used to IDEs.  I've used a couple open source Git viewers and  
would hesitate to recommend them.  The option to use GitBox makes  
adoption of Git more likely in this case.

CVS is a GNU project.  Did FSF go after proprietary cvs wrappers?

Josh

Re: Gitbox

[Tim Smith]

On Apr 13, 2011, at 1:04 PM, Drew Northup wrote:
> On Wed, 2011-04-13 at 12:41 -0700, Randal L. Schwartz wrote:
>> He's made new porcelein, and using the plumbing directly.  That's
>> exactly why we have plumbing in git, right?
> 
> If that's what he's doing then why does he state that the git binaries
> are "for your convenience only"--strongly implying his program will work
> in the absence thereof? I'm hoping he just made a typo--as what he said

Yes, it will work without needing the git binaries he includes. If you already have git binaries (e.g., from macports) you can use those. Hence, the ones he provides are for your convenience, letting you use Gitbox without having to deal with finding and installing git yourself.

-- 
--tzs

Re: Gitbox

[Randal L. Schwartz]

>>>>> "Junio" == Junio C Hamano <gitster@pobox.com> writes:

Junio> However, 3 (c) is allowed only for noncommercial distributor that passes
Junio> compiled executable along from its upstream source, so that part may need
Junio> to be fixed, I suspect, to comply with section 3 by going in either
Junio> subsections 3 (a) or 3 (b) route.

It's interesting also to note that the GPL uses the term "noncommercial"
without ever defining it.  Could gitbox argue that it is charging for
his code, and not for git, and therefore, it's a "noncommercial"
distribution of git which just happens to be aggregated?

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.posterous.com/ for Smalltalk discussion

Re: Gitbox

[Chris Perkins]

Let's look at this at a slighty different way. Let's say someone writes
a GUI wrapper for Git, bundles it with Git, and then offers for sale a
new proprietary SVC system. They list off all the wonderful features
that it has.  On the back page of their website is a small 'Licenses'
disclosure and the source code to Git comes with the download buried in
a subdirectory.  None of the users realize the software is using Git.

Is that a violation of the GPL? I would say that it absolutely is. A
proprietary software is being built atop the hard labor of the GPL
authored software. Hundreds contributed to the Open Source effort, one
profits by burying their work in his with the minimal effort.

So, I guess the question is, how is that scenario different than Gitbox?
I guess Gitbox doesn't present itself as something else, but code-wise
it is no different. Without the Git binary there is no Gitbox.


(Randal L Schwartz writes):
>And it's also about readline.a, which *has* to be linked into a binary
>to make it work.  None of the git code is being *linked* in.

The GPL doesn't mention .a files versus .o versus .so.   _Linked_ is not
the issue.  The issue is one work being _based_ on another, and the word
_based_ is the one that appears in the GPL, including the section you
quoted.

Section 2 b) of GPL 2 plainly states that any work based on a GPL
licensed product must be distributed free of charge and be GPL:

  2 b) You must cause any work that you distribute or publish,
  that in whole or in part contains or is derived from the Program
  or any part thereof, to be licensed as a whole at no charge to
  all third parties under the terms of this License.

And the closing explanation of the GPL 2 says the same in even more
frank and plain language. These are the last three lines. 'your program'
refers to the program being placed under GPL (ie Git), they are written
to someone considering releasing software under the GPL:

  This General Public License does not permit incorporating your
  program into proprietary programs. If your program is a subroutine
  library, you may consider it more useful to permit linking
  proprietary applications with the library. If this is what you want
  to do, use the GNU Lesser General Public License instead of this
  License.



(Joshua Juran writes):
> My shell scripts that call git are also useless without Git.  Am I not
> allowed to distribute my scripts under non-GPL terms?  

I don't know the answer to this. Maybe they must be distributed as GPL.


(Joshua Juran writes):
> CVS is a GNU project.  Did FSF go after proprietary cvs wrappers?

I don't know, but I doubt they did.  And it is a good point that failure
to defend a copyright can often be argued in court as tacit disavowal of
that copyright.




Chris

Re: Gitbox

[Randal L. Schwartz]

>>>>> "Chris" == Chris Perkins <cperkins@medialab.com> writes:

Chris> Is that a violation of the GPL? I would say that it absolutely
Chris> is.

Which part, then?  Name chapter and verse.

Maybe the spirit, as you see it.  But which letter?

But if the GPL is to be treated as a legally enforceable document, "live
by the sword, and die by the sword".  GPL-desk-pounders can't have it
both ways.

(This is also why it's not "GNU/Linux"... if RMS had wanted that, it
should have been written into the license.  He forgot, so the rest is
history. It's just Linux, dammit.)

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.posterous.com/ for Smalltalk discussion

Re: Gitbox

[Chris Perkins]

On 4/13/11 11:02 PM, Randal L. Schwartz wrote:
>>>>>> "Chris" == Chris Perkins <cperkins@medialab.com> writes:
>
> Chris> Is that a violation of the GPL? I would say that it absolutely
> Chris> is.
>
> Which part, then?  Name chapter and verse.

Section 2 b). I quoted it in whole in my email:

  2 b) You must cause any work that you distribute or publish,
  that in whole or in part contains or is derived from the Program
  or any part thereof, to be licensed as a whole at no charge to
  all third parties under the terms of this License.

and the last paragraph in the "how to" section is even more frank.


> But if the GPL is to be treated as a legally enforceable document, "live
> by the sword, and die by the sword".  GPL-desk-pounders can't have it
> both ways.

I'm not a GPL-desk-pounder. I write commercial software for a living.
But I take the efforts I put into my software very seriously. If I were
to release software as free software under a license like the GPL I
would not appreciate someone attempting to profit off my hard work by
putting a bow on it and selling it as their own.

Don't get me wrong. I don't see Gitbox as just a 'bow' atop Git.
Frankly, the type of wrapping that Gitbox is doing seems perfectly
reasonable to me. Gitbox is providing real value to users, it's not
disguising or hiding Git and it's not pretending to be provide the work
of Git as its own and it isn't trying to pretend to be anything that it
isn't.

However, that said, I still think it's a violation of the GPL. That
license lays it out very clearly without much room for interpretation.
You can't make commercial works based on GPL licensed software.  Without
Git there is no Gitbox. End of story.

I'm not a lawyer, were I then perhaps I'd know solidly one way or other.
 I'm not the author nor one of the contributors to Git. Were I then
perhaps I might feel differently about Gitbox being 'perfectly reasonable'.

But I will add this, some on our team would very much like to
incorporate Git into one of our own upcoming commercial products.
Code-wise in the same manner as Gitbox has done. But we aren't going to
do this, because the GPL license for Git clearly states that we cannot.

Chris

Re: Gitbox

[Randal L. Schwartz]

>>>>> "Chris" == Chris Perkins <cperkins@medialab.com> writes:

>> Which part, then?  Name chapter and verse.

Chris> Section 2 b). I quoted it in whole in my email:

Chris>   2 b) You must cause any work that you distribute or publish,
Chris>   that in whole or in part contains or is derived from the Program
Chris>   or any part thereof, to be licensed as a whole at no charge to
Chris>   all third parties under the terms of this License.

Broken pre-condition.  See the beginning of 2:

    2. You may modify your copy or copies of the Program or any portion of
    it, thus forming a work based on the Program, and copy and distribute
    such modifications or work under the terms of Section 1 above, provided
    that you also meet all of these conditions:

Gitbox is not *modifying*, or creating a work *based on the Program*.
Separate copyrights.  The rest of your argument thus falls apart.

Chris> However, that said, I still think it's a violation of the GPL. That
Chris> license lays it out very clearly without much room for
Chris> interpretation.

Bullcrap. See above.  Get your preconditions right, and I'll play along.

Chris> You can't make commercial works based on GPL licensed software.

Sure you can.  RedHat makes money somehow.  Don't play the "we must be
starving artists for the cause of GPL" on me.

Chris> But I will add this, some on our team would very much like to
Chris> incorporate Git into one of our own upcoming commercial products.
Chris> Code-wise in the same manner as Gitbox has done. But we aren't going to
Chris> do this, because the GPL license for Git clearly states that we
Chris> cannot.

Maybe you need better lawyers then.

-- 
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.posterous.com/ for Smalltalk discussion

Re: Gitbox

[Joshua Juran]

On Apr 13, 2011, at 7:28 PM, Chris Perkins wrote:

> You can't make commercial works based on GPL licensed software.   
> Without
> Git there is no Gitbox. End of story.
>
> I'm not a lawyer, were I then perhaps I'd know solidly one way or  
> other.

Or, you could consult a lawyer.  Or you could see what Git authors and/ 
or the FSF think about it.

> But I will add this, some on our team would very much like to
> incorporate Git into one of our own upcoming commercial products.

I imagine some of your users would like that as well.

> Code-wise in the same manner as Gitbox has done.

As well as ProjectBuilder/Xcode.  The same company that released their  
Objective-C compiler front end to GCC under the GPL, demonstrating  
Steve Jobs' willingness to comply with RMS' interpretation of the GPL,  
also created their own proprietary GUI front end to GCC tools.   
ProjectBuilder/Xcode has existed for over two decades, and to my  
knowledge no objection has been lodged.  Clearly, there is a critical  
difference between linking and invoking.

> But we aren't going to
> do this, because the GPL license for Git clearly states that we  
> cannot.

Then you may lose some of your users to competitors who more  
accurately interpret the GPL and are clear what rights they've been  
granted.

Josh

Re: Gitbox

[Sitaram Chamarty]

On Thu, Apr 14, 2011 at 1:34 AM, Drew Northup <drew.northup@maine.edu> wrote:

> If that's what he's doing then why does he state that the git binaries
> are "for your convenience only"--strongly implying his program will work
> in the absence thereof? I'm hoping he just made a typo--as what he said

Not at all; as others have explained, you're reading too much into that.

I believe what he is doing is often called "mere aggregation".
Perfectly legal, IMO.

Re: Gitbox

[Vincent van Ravesteijn]

> I believe what he is doing is often called "mere aggregation".
> Perfectly legal, IMO.

I guess he just shouldn't sell it as a Version Control App, but only as a Repository Manager.

So, the description: "Gitbox is a Git repository manager" is clearly saying that it is just a repository manager (what some people like to call a 'bow' depending on their mood)

The subtitle of his app: "The missing version control on a Mac" raises (maybe too much) the impression that he has written a version control system.

Vincent

Re: Gitbox

[Drew Northup]

On Thu, 2011-04-14 at 15:22 +0530, Sitaram Chamarty wrote:
> On Thu, Apr 14, 2011 at 1:34 AM, Drew Northup <drew.northup@maine.edu> wrote:
> 
> > If that's what he's doing then why does he state that the git binaries
> > are "for your convenience only"--strongly implying his program will work
> > in the absence thereof? I'm hoping he just made a typo--as what he said
> 
> Not at all; as others have explained, you're reading too much into that.

I deal with stupid user tricks on a daily basis. Most people that I've
met that aren't very computer savvy (including a number of web
programmers) would interpret what he wrote to mean "I'm not going to use
the command-line tools, so I can delete them and free up space on my
disk" (despite the fact that I'm pretty sure all of us actually agree
that's not what he intended). Sating that "Gitbox requires the Git
command-line tools and other binaries," and that "I have included
version xxx with Gitbox for your convenience, but you may use any newer
compatible version as well," would be far more clear. (The key word is
"requires.") You'd be surprised (or, well, probably not in your personal
case) how many programmers would benefit from it being stated that
clearly.

> I believe what he is doing is often called "mere aggregation".
> Perfectly legal, IMO.

I'm pretty sure that's what's going on as well. We've been copying him
this whole time so if he feels the need to speak up he can do so.

-- 
-Drew Northup
________________________________________________
"As opposed to vegetable or mineral error?"
-John Pescatore, SANS NewsBites Vol. 12 Num. 59

Re: Gitbox

[Martin Langhoff]

On Wed, Apr 13, 2011 at 9:58 PM, Chris Perkins <cperkins@medialab.com> wrote:
> Let's look at this at a slighty different way. Let's say someone writes
> a GUI wrapper for Git, bundles it with Git, and then offers for sale a
> new proprietary SVC system. They list off all the wonderful features
> that it has.  On the back page of their website is a small 'Licenses'
> disclosure and the source code to Git comes with the download buried in
> a subdirectory.  None of the users realize the software is using Git.
>
> Is that a violation of the GPL? I would say that it absolutely is.

It absolutely is not. Lots of companies do this, and it is perfectly
kosher -- either bundle the src somewhere or offer a link to download
the source somewhere.

While IANAL, and specifically not _your_ lawyer, I have been in this
field for >10 years, and studied law @ masters level on software
licensing. You are reading the GPL wrong, and you're not aware of
widespread industry practices around it.

Anyone who is curious about this gitbox thing, and interested in
*facts* instead of fiction, could advance our knowledge with a simple
procedure:

 - Download the "free" version (or payfor the paid version!). It's a
zipfile, no need to hurt any Macs.

 - See if it includes the src or a link to download the src -- it will
probably be in a corner of the documentation or license. Maybe there's
an offer to provide the src in a different way, but a download link is
the usual trick.

 - Does the link work? Can you effectively get the src?

 - Does the src match the binaries you got?

cheers,



m
-- 
 martin.langhoff@gmail.com
 martin@laptop.org -- Software Architect - OLPC
 - ask interesting questions
 - don't get distracted with shiny stuff  - working code first
 - http://wiki.laptop.org/go/User:Martinlanghoff

Re: Gitbox

[Victor Engmark]

On 04/14/2011 03:55 PM, Martin Langhoff wrote:
> On Wed, Apr 13, 2011 at 9:58 PM, Chris Perkins <cperkins@medialab.com> wrote:
>> Let's look at this at a slighty different way. Let's say someone writes
>> a GUI wrapper for Git, bundles it with Git, and then offers for sale a
>> new proprietary SVC system. They list off all the wonderful features
>> that it has.  On the back page of their website is a small 'Licenses'
>> disclosure and the source code to Git comes with the download buried in
>> a subdirectory.  None of the users realize the software is using Git.
>>
>> Is that a violation of the GPL? I would say that it absolutely is.
> 
> It absolutely is not. Lots of companies do this, and it is perfectly
> kosher -- either bundle the src somewhere or offer a link to download
> the source somewhere.
> 
> While IANAL, and specifically not _your_ lawyer, I have been in this
> field for >10 years, and studied law @ masters level on software
> licensing. You are reading the GPL wrong, and you're not aware of
> widespread industry practices around it.
> 
> Anyone who is curious about this gitbox thing, and interested in
> *facts* instead of fiction, could advance our knowledge with a simple
> procedure:
> 
>  - Download the "free" version (or payfor the paid version!). It's a
> zipfile, no need to hurt any Macs.
> 
>  - See if it includes the src or a link to download the src -- it will
> probably be in a corner of the documentation or license. Maybe there's
> an offer to provide the src in a different way, but a download link is
> the usual trick.
> 
>  - Does the link work? Can you effectively get the src?
> 
>  - Does the src match the binaries you got?

Excellent answer; it would be great to know in detail what would be an
effective (and efficient, if possible) procedure for validating GPL
compliance. Something like a cartoon guide to the GPL for developers
and/or users. I don't even know if my own GPL'ed projects are within the
letter of the law here.

-- 
Victor

Re: Gitbox

[Martin Langhoff]

On Thu, Apr 14, 2011 at 10:15 AM, Victor Engmark
<victor.engmark@terreactive.ch> wrote:
> Excellent answer; it would be great to know in detail what would be an
> effective (and efficient, if possible) procedure for validating GPL
> compliance.

Glad you find it useful. There are some interesting FAQs in FSF's
website. And http://gpl-violations.org/faq/sourcecode-faq.html

> Something like a cartoon guide

Both why the lucky stiff and Hunter Thompson would be candidates if
they were still around.


m
-- 
 martin.langhoff@gmail.com
 martin@laptop.org -- Software Architect - OLPC
 - ask interesting questions
 - don't get distracted with shiny stuff  - working code first
 - http://wiki.laptop.org/go/User:Martinlanghoff

Re: Gitbox

[Tim Smith]

When GPLv2 talks about works "based on" another work, they mean works that are "derivative works" under copyright law. This is mentioned explicitly in section 0.

GPLv2 also uses the term "derived from", which it does not define and is not a term of copyright law.

Copyright law gives the copyright owner certain exclusive rights. No one else is allowed to do those things without permission from the copyright owner. The GPLv2 (and most other free software licenses) are designed to give you that permission. They serve strictly to remove restrictions, not to add them.

Contrast to a typical commercial software EULA, which might grant to you some permissions, but also will try to stop you from doing things that you are normally allowed to do. For instance, copyright law places no limit on the number of backup copies you can make of software you purchase. All that it requires (at least in the US) is that if you transfer ownership of your copy of the software to someone else you either destroy all your backups, or transfer them with the software. The typical EULA will make you agree to only keep one backup copy.

The FSF is quite clear that they intend the GPLv2 to not be a EULA. It only extends rights to you, it does not take any away. Accordingly, when trying to decide what an undefined term like "derived from" means in the context of GPLv2, you should look to copyright law. This makes it clear that they are using "derived" from to mean the same thing that copyright law means by a "derivative work".

Gitbox is not a derivative work of Git. The only thing Gitbox is doing that requires permission from the Git copyright owners is distributing an unmodified copy of Git with Gitbox. As long as Gitbox obeys the GPL by making the corresponding Git source available in a way allowed under GPLv2, they have permission. End of story.

They tried to clear up some of this confusing in GPLv3:

> To “modify” a work means to copy from or adapt all or part of the work in a fashion requiring copyright permission, other than the making of an exact copy. The resulting work is called a “modified version” of the earlier work or a work “based on” the earlier work.



-- 
--Tim Smith