Re: [Savannah-hackers-public] help needed with savannah

[Alex Fernandez]

Hi again,

I am not sure if others find the discussion relevant or interesting;
it takes me a long time to discuss these issues and sadly I don't have
the time, so I will have to withdraw from making further comments in
this thread. Feel free to elaborate on your side though, and others
are also welcome to advance the discussion.

On Mon, Jul 18, 2011 at 12:16 AM, Mario Castelan Castro
<address@hidden> wrote:
>> But I also think that we could serve the Savannah project better if we
>> improved our means (including our tooling).
>
> It seems than we disagree in our concept of improved tooling.  A python
> script to automatically check license headers don't fits in my idea of
> an improved tooling for instance, that's the source of this verbal
> friction.

I don't know what else fits the description of "improved tooling".
Just improving the admin pages can be seen as a small optimization of
our tasks, but it doesn't lift the heavy burden of Savannah
administration. At least not for me.

>> An example of submission blocked because of the developer's opinions:
>>   https://savannah.gnu.org/task/?10695
>> Your message:
>>   "- From the above I conclude atpic.com don't cares about freedom nor real
>> sharing because it encourages proprietary licensing and discourages
>> usage of works whose Copyright is held by third parties."
>>
>> IMHO, it is not our mission to evaluate the developer's attitude
>> towards freedom or sharing; we need to assess if the submitted
>> software respects the four freedoms *with regards to software*.
>
> It's not the case than I'm evaluating the developer's attitude altough
> sometimes I make *comments* on what's ours (Our philosophy).  What I'm
> evaluating is compliance with
> https://savannah.gnu.org/register/requirements.

[The correct URL is https://savannah.gnu.org/register/requirements.php.]

> Software dedicated as a client to a proprietary work archive site don't
> promotes but usage of those propreitary works, that was the issue with
> this project.

Nowhere in the requirements does it state that a "proprietary work
archive" should not be promoted, it only speaks about free formats. I
think you may be misconstruing the requirements to fit anything you
don't like. Note that the GNU web pages may be seen as a "proprietary
work archive" since they are an archive of works which don't allow
modifications; we are promoting usage of such an archive and it's
perfectly fine.

So the net effect is that a perfectly valid submission was blocked
because of an attitude problem.

>> A submission blocked because one of the libraries imported used the
>> term "open source": [...]
>
> This is not true.  Your usage of "blocked" is misleading.  I don't
> "block" project, I either approve, reject them or ask the maintainer to
> make changes to comply with our requirements, this is a case of the
> later.

We can redefine words if you want, but the net effect is a blocked
submission. This is only if we expect submissions to move forward,
which admittedly they don't do on their own.

> Please read carefully the task you're commenting about.  In no place I
> did forbid the maintainer to use xajax or whatever other library he
> wishes.  The verbatin quote is "Could you please replace (Preferable) or
> at least append this reference to open source by one to free software?."
> Because (Also pointed in the same message) *the library was included in
> the tarball*.  There was no need to remove anything and in no place I
> stated such a thing.  Yet the maintainer decision was to remove the
> library, rather than to replace a string in the uploaded tarball.

Modifying library dependencies is not good practice, and it should not
be requested in any case. Now, you might have asked the developer to
push the change upstream, but again that is out of scope for Savannah
administration. I would prefer if we would center in the four freedoms
and the hosting requirements for projects, without extending their
reach into libraries unless it is really needed. Not in this case.

>> Philosophical issues are fine, but the GNU project is concerned with
>> software and the four freedoms (to inspect, to modify, to copy and to
>> distribute modified versions). Many other GNU artifacts have licenses
>> which restrict freedom, and in fact the GPL is a prime example of
>> restricting people's freedom -- in order to attain the four freedoms
>> that the FSF defends.
>
> I see no "prime example" of freedom restriction in the GNU GPL.
> Copyleft is a restriction of power, not freedom.  This is explained in
> http://www.gnu.org/philosophy/freedom-or-power.html.

Nowhere in the article you link does it mention copyleft, since it
does not speak about copyleft. Copyleft is an intentional restriction
of freedom (compared to any BSD or Apache license) for the sake of
promoting "share and share alike" in software. In copyleft, the
recipient has some boundaries as to what they can do with the software
they receive, in exchange for the four freedoms. I am sure you will be
able to find links yourself and I don't want to turn this discussion
into a link war.

>> Not to speak about the GFDL: here the developer has a few limitations
>> with what they can do with the documentation.  The FSF thinks that the
>> compromise between freedom and collaboration is acceptable.
>
> Could you please elaborate?.

The GFDL restricts the freedoms of the recipients in that they cannot
omit certain parts of documents licensed under it (invariant sections
IIRC).

>> The key point here is that the philosophical foundations are
>> debatable, while the four freedoms are not. Either the developers is
>> fine with you modifying the code and distributing the result, or they
>> are not. Legal standing is more debatable, but from our developer
>> point of view the issue should be clear.
>
> It's not clear to me if you're suggesting than we should take a
> mechanical attitude of approval/rejection without further comments.

Nope, we should take a machine-assisted approach to approval and rejection.

> GNU Savannah mission is to advance free software, hosting it is only a
> part of the work.

And the other parts are? Because Savannah is a hosting site; if we
fail at hosting I don't see how we can advance free software.

>> [...] The GNU project does not promote modification of those documents
>> it views as pure opinion. Even in the GNU project itself: the license
>> for GNU web pages does not allow modifications. Here the GNU project
>> puts some other principles in front of the right to modify some
>> documents.
>
> The GNU project supports the right to modify *software*.
>
> The GNU project don't allows modifications of its *opinion articles*.
>
> It does not follows (Non sequitur) than the GNU Project puts some other
> principles in front of anothers.

Huh? Just try digging a little deeper.

>> My point is that checking for ideological purity is hard to do. But
>> checking if some piece of software respects the four freedoms and some
>> other bits of GNU terminology is easier to do automatically, making
>> our task as admins much easier and more productive. An automatic check
>> will never be 100% exact, but I prefer to let slip 10 mentions of
>> "open source" because they are hyphenated than losing one valuable
>> contributor.
>
> Isn't this a false dichotomy?.  I see no reason to infer a contributor
> kept (Not lost) will result in a project labeled as "open source".

Whatever. In the end the truth of the matter is that evaluations are
not being done and projects are being lost in the ether, because they
require too much time.

> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Hope I have cleared your inquires on my evaluation of those projects.

Not really, but we can leave it at that. I don't think this discussion
will turn into anything productive any time soon.

Alex.