ASP loophole - where is the line
Mark Rafn
loophole" is where to draw the line. I'm currently of the opinion that
distribution is a good line and any other is fuzzy, but I'd kind of like
to be convinced otherwise.
Here's the continuum I see:
a) Joe opens a business "Joe's Typesetting Service". His customer brings
in handrwitten pages, Joe uses GPL software (some of which he's modified
to fit his work patterns) to enter, typeset, and print the documents.
The printed documents are given to the customer.
b) His customer brings in floppies with text files, and recieves printed
documents.
c) Customer prefers Joe to give him a floppy with Postscript output
on it instead of printed documents.
d) Joe automates this conversion such that he can put a floppy into a
drive and the textfile on the floppy will be replaced by a postscript
file. This saves him a lot of time.
e) To save further time, he allows his customer to put the floppy into
the drive rather than interrupting Joe.
f) The customer e-mails Joe a textfile and joe e-mails the output.
g) Joe adds to his program the ability to e-mail the resulting postscript
file automatically.
h) Joe sets up a webpage which uses a mailto: form to send him the file.
i) Joe sets up procmail to run his program when the customer sends
e-mail.
j) His webpage is still up, now much more responsive.
k) Joe converts his process to run as a CGI, accepting form posts and
returning the Postscript immediately
l) Joe adds an XML-RPC interface to his page.
The challenges:
In which cases should Joe be forced to give his program to his customer?
Without mention of a specific protocol, please try to find a definition of
"user" for a theoretical closure of the ASP loophole which would require
Joe to give his program (including source, he wants to follow the GPL) to
his customer.
--
Mark Rafn da...@dagon.net <http://www.dagon.net/>
--
To UNSUBSCRIBE, email to debian-leg...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Branden Robinson
Great work, Mark. I think the reason that no one has replied is because
those who have already made their minds up are having trouble thinking
of reasons why their preference shouldn't be moved up or down on the
scale. :)
I don't have a good answer to your challenge at present, but it is a
very nice exposition of the sort of problem I was wrestling with when I
was trying to come up with the "DFCL". (I guess haven't completely
given up hope on that yet, though. :) )
--
G. Branden Robinson | To stay young requires unceasing
Debian GNU/Linux | cultivation of the ability to
bra...@debian.org | unlearn old falsehoods.
http://people.debian.org/~branden/ | -- Robert Heinlein
Anthony Towns
> The challenges:
> In which cases should Joe be forced to give his program to his customer?
IMO none of them: distributing software to customers is too significant
a technical burden over and above presenting the output of software
to customers.
> Without mention of a specific protocol, please try to find a definition of
> "user" for a theoretical closure of the ASP loophole which would require
> Joe to give his program (including source, he wants to follow the GPL) to
> his customer.
Distributing the source just once, to a single person or group (eg, the
upstream author/development community, Debian, the FSF, a government
organisation like the patent office) and giving them rights to freely
redistribute the source effectively solves the ASP loophole too, without
imposing the technical burden that distributing to (potentially) all of
your customers would.
Cheers,
aj
--
Anthony Towns <a...@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
``Dear Anthony Towns: [...] Congratulations --
you are now certified as a Red Hat Certified Engineer!''
Henning Makholm
> On Thu, Mar 13, 2003 at 05:08:10PM -0800, Mark Rafn wrote:
> > The challenges:
> > In which cases should Joe be forced to give his program to his customer?
> IMO none of them: distributing software to customers is too significant
> a technical burden over and above presenting the output of software
> to customers.
> Distributing the source just once, to a single person or group (eg, the
> upstream author/development community, Debian, the FSF, a government
> organisation like the patent office) and giving them rights to freely
> redistribute the source effectively solves the ASP loophole too,
That's sidestepping the issue. Now: In which of Mark's cases should
Joe be forced to give his program to the upstrem author, Debian, the
FSF, or the government?
--
Henning Makholm "That's okay. I'm hoping to convince the
millions of open-minded people like Hrunkner Unnerby."
Jakob Bohm
> loophole" is where to draw the line. I'm currently of the opinion that
> distribution is a good line and any other is fuzzy, but I'd kind of like
> to be convinced otherwise.
>
Here are some more cases to the end of the spectrum w/o forced
distribution.
a-6) Joe uses no GPL software at all, but his ISP is using a GPL
server or router in the path between Joe and his customer.
a-5) Joe uses no GPL software on his computer (he may be running
Windows etc.), but his firewall computer is running GPL software
and was modified by Joe.
a-4) Joe did not modify the typesetting software, but he did
modify some other GPL code supporting the task, such as the
kernel or libc (ignoring the Linux kernel license exception and
LGPL linking exceptions for this argument, imagine the forced
distribution clause to go there too).
a-3) Joes only modification is to compile the software (a form
of modification under the GPL definitions).
a-2) Joes only modification is to pass arguments to
./Configure, thereby effectively changing some hardcoded paths.
a-1) Joes only modification is to change a few text strings and
hardcoded paths.
>
> a) Joe opens a business "Joe's Typesetting Service". His customer brings
> in handrwitten pages, Joe uses GPL software (some of which he's modified
> to fit his work patterns) to enter, typeset, and print the documents.
> The printed documents are given to the customer.
>
> b) His customer brings in floppies with text files, and recieves printed
> documents.
>
> c) Customer prefers Joe to give him a floppy with Postscript output
> on it instead of printed documents.
>
> d) Joe automates this conversion such that he can put a floppy into a
> drive and the textfile on the floppy will be replaced by a postscript
> file. This saves him a lot of time.
>
> e) To save further time, he allows his customer to put the floppy into
> the drive rather than interrupting Joe.
>
> f) The customer e-mails Joe a textfile and joe e-mails the output.
>
> g) Joe adds to his program the ability to e-mail the resulting postscript
> file automatically.
>
> h) Joe sets up a webpage which uses a mailto: form to send him the file.
>
> i) Joe sets up procmail to run his program when the customer sends
> e-mail.
>
> j) His webpage is still up, now much more responsive.
>
> k) Joe converts his process to run as a CGI, accepting form posts and
> returning the Postscript immediately
>
> l) Joe adds an XML-RPC interface to his page.
>
> The challenges:
>
> In which cases should Joe be forced to give his program to his customer?
>
> Without mention of a specific protocol, please try to find a definition of
> "user" for a theoretical closure of the ASP loophole which would require
> Joe to give his program (including source, he wants to follow the GPL) to
> his customer.
> --
> Mark Rafn da...@dagon.net <http://www.dagon.net/>
>
And a different set of cases which don't fall at a particular
point in the above continuum:
x1) Joe is non-commercial, he typesets for charity.
x2) Joe is just offering the service of replying to e-mails
people send him (the continuum goes from manually answering
verbal in-person questions to Joe being www.google.com).
x3) Joe is not even doing that, he is just running a GPL identd
on his machine, and has hardcoded some of the responses to hide
his name from people like doubleclick.com. Should he be obliged
to distribute the patch that hardcodes removal of his name to
the very people he is hiding the name from (since those people
are the primary users of the service modification to not provide
the name)?
There are even more extreme examples occurring in normal usage of
Debian systems.
As for a ground for saying that forced distribution can be
against Debian rules, how about clauses 1 and 4 of the social
contract itself?
--
This message is hastily written, please ignore any unpleasant wordings,
do not consider it a binding commitment, even if its phrasing may
indicate so. Its contents may be deliberately or accidentally untrue.
Trademarks and other things belong to their owners, if any.
Anthony Towns
> Scripsit Anthony Towns <a...@azure.humbug.org.au>
> > On Thu, Mar 13, 2003 at 05:08:10PM -0800, Mark Rafn wrote:
> > > The challenges:
> > > In which cases should Joe be forced to give his program to his customer?
> > a technical burden over and above presenting the output of software
> > to customers.
> > Distributing the source just once, to a single person or group (eg, the
> > upstream author/development community, Debian, the FSF, a government
> > organisation like the patent office) and giving them rights to freely
> > redistribute the source effectively solves the ASP loophole too,
> That's sidestepping the issue. Now: In which of Mark's cases should
> Joe be forced to give his program to the upstrem author, Debian, the
> FSF, or the government?
I don't see any problems with him being required to present it on request
to some particular person/group in any of the cases, as long as the
requests aren't unduly unreasonable.
Branden Robinson
> On Sun, Mar 16, 2003 at 03:48:04PM +0100, Henning Makholm wrote:
> > That's sidestepping the issue. Now: In which of Mark's cases should
> > Joe be forced to give his program to the upstrem author, Debian, the
> > FSF, or the government?
>
> I don't see any problems with him being required to present it on request
> to some particular person/group in any of the cases, as long as the
> requests aren't unduly unreasonable.
What does that mean? What's your definition of "unduly unreasonable"?
Would you care to share some "pet tests" with us that would help us to
better understand what you think of as "unduly unresonable", versus the
preferable (I suppose) "duly unresonable"?
--
G. Branden Robinson | Don't use nuclear weapons to
Debian GNU/Linux | troubleshoot faults.
bra...@debian.org | -- US Air Force Instruction 91-111
http://people.debian.org/~branden/ |