[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Do Debian's users care about the AGPL?



On Wed, 3 Sep 2008 14:40:39 -0700
"Steve Lamb" <grey@dmiyu.org> wrote:

> On Wed, Sep 03, 2008 at 05:03:27PM -0400, Daniel Dickinson wrote:
> > Which is the thing.  GPL guarantees freedom the users of the
> > software. The AGPL says that the user is the one writing the
> > documents with the software is the user not the one running the
> > code.  I agree with the AGPL on that.
> 
>     Those users have not lost their freedom.  That's the distiction
> people have lost.  Lets take a simple example, Google's web-based
> spreadsheet.  Who is using the software, you or Google?  Answer:
> Google.  They provide you *access* to that software but you're not
> the user.  Just as if you came into my house, sat down at my computer
> and then thought just because you are poking at software I've written
> and allowing you to access gives you any rights to the source.  If I
> distributed that software to you then, yes, you would have rights to
> it. 

I think this is where public/private distinction comes into play.
Someone coming into your house to use your laptop is using your
software in a private context, not a public one, and thus is more like
'within a company'.  Such a distinction makes sense, just as it does in
real life public/private distinction.

I think you might disagree about what's public and what's private
though.  

The difference with the Google scenario is that Google is providing
machines to run the code on so that you can use the software (as I see
it the service is running the code, not the software itself).

> > I don't follow.  If you mean ASP turned to Open Source because they
> > get free (as in beer) software that don't have to share, then I'd
> > argue that is an example of a popular loophole, no a robust example
> > of the benefits of libre software.
> 
>     No, they turned to it because they got free (as in beer) tools
> for which they could create works which they could *choose* to keep
> in house if needed.

But if they are not modifying code and distributing it to users (in the
sense of user I mean) they are still not required to distribute the
modified works, it's only when they are distributing the works that it
matters.

I guess for me 'software freedom' means freedom for the end-user, where
as you interpret user as one who's machine the code runs on.

> > Jeopardizing ensuring rights of user to control their own software
> > is restricting freedom?  I don't think so.  Of course we disagree
> > on who the user is, so I mean a different thing by that then you.
> 
>     Exactly so.  The implications are far more profound than you
> think when we take your definition of "user".
> 
> > I disagree with you interpretation of the license in this regard,
> > and I probably disagree with your definition of freedom as well.
> > The definition of freedom as shown in the GPL and AGPL are what
> > counts here, and opinion of which definition of user is accurate.
> 
>     Really?  So the fact that you're provided access to a custom
> application means you're a user and thus must have rights to the
> source code?  You do realize how much software you're provided access
> to but aren't the user?  Let me think off the top of my head of
> software which customers of mine have had access to which fall under
> your overly broad interpretation of user in the quest for the holy
> grail of restricting other people's uses to your narrow definition.

My narrow definition.  I do believe you're the one who is specify that
user means the one running the code.  Period.
 
>     Room reseveration software.  ATM software.  Slot machine software
> (Btw, you do realize most slots nowadays are computers, right?).
> Patron management software (better known as point systems).
> Point-of-sale terminals and, by extension, their servers.  In-room
> entertainment software.  Did I miss any portion of a casino that the
> customer doesn't touch?

Except of course I think these are cases where the user should, if
released under a license that defines user as I do, be granted access
to the code.  The whole point of software freedom is that the user has
the ability to improve and modify the software they use.

> 
>     It boils down to this.  The casino (and hotel) is using that
> software to provide a service to the customer.  The customer is not
> the user of the software even though they do "use" it.  Same thing
> for ASP.  That is why the S is *S*ERVICE.  If you close this
> "loophole" for ASPs then you are, by fiat, also demanding that any
> time you use an ATM machine you are using the software thus have a
> right to the source code of said software.  If you don't see how that
> could cause some serious problems for the adoption of OS tools to be
> used to build those custom applications then you really don't
> understand what's going on nor the motivactions of why people chose
> the solutions they do outside of your narrow mindset.

I'm not suggesting OS tools cause release of code unless they're part
of a derivative work.  Just because use use gcc to build code doesn't
mean that my code is open source, so this looks like strawman to me.

If I write a webapp that I make AGPL it's because the whole point of
the apps is to be used by end-users, with a service provider (whose
service running the code, not providing, say, a word processor) in the
middle.  For me the service provider is just a vehicle to the user, for
you the service provider is the user.

I can see how in some situations one would apply and not the other, but
I don't think a blanket, the one running the code is the user, is
correct.  I think this is again the whole public/private divide.

The stuff that is done internally for internal purpose is different
than the stuff that is done for public consumption.  The code for a
web-based word processor is public, the host operating system is not.
It's a difference in target.  The word processor is what the user is
running, through the provider, the operating system could, presumably,
be any *nix and isn't relevant to the user.

Anyway, back the AGPL and Debian.  I think the AGPL should be in main
because I don't think it is non-free in terms of restricting use of
non-distributor users (distributor users are restricted in the GPL too,
it's the definition of who is a distributor that is different; for
the AGPL you're distributing through network use as well as for
direct on-machine use).

I do think that with your narrow definition of user that the AGPL is
probably non-free.

> > [snip off-topic political commentary, some of which is right and
> > some which isn't]
> 
>     Tsk, it's all correct.

heh, we'll not get into that :-)


-- 
And that's my crabbing done for the day.  Got it out of the way early, 
now I have the rest of the afternoon to sniff fragrant tea-roses or 
strangle cute bunnies or something.   -- Michael Devore
GnuPG Key Fingerprint 86 F5 81 A5 D4 2E 1F 1C      http://gnupg.org
The C Shore: http://www.wightman.ca/~cshore

Attachment: signature.asc
Description: PGP signature


Reply to: