[A bit rambling] Open source licensing being questioned by anti-copyright types
ClassCastException
come to suspect that open source licensing is itself a source of trouble.
Notably, Clojure's license is incompatible with the GPL, under which
large chunks of other open source software is licensed. It seemed to me
that a measure intended to free up software so anyone could develop it
and contribute to it, as long as their contributions in turn became
available to others, had somehow gotten itself tangled in knots that
actually hindered this purpose.
So I decided to do a little reading on copyright in general. Why does it
even exist? The nominal purpose, it turns out, is to "promote the
progress of science and the useful arts" by providing a way for the
creators of any popular or important work to ensure remuneration,
basically. Which smells suspiciously like a grant of monopoly -- which,
barring the notion of "fair use", it basically is. Furthermore there are
a LOT of blogs out there expressing serious criticism of copyright,
pointing out that fair use is not in most cases a workable defense even
when it should be, and that copyright has been twisted away from its
original purpose by corporations seeking to extend and tighten their
control over lucrative media and software properties.
It thus seems that copyright was twisted away from its original purpose,
to which it might have been poorly suited to begin with, and open source
licenses try to twist it back toward that purpose. Double twist. Is it
any wonder it's getting tangled in knots?
Here's something even weirder though: my blog-surfing led me eventually
to http://www.againstmonopoly.org/index.php?perm=593056000000003021 where
there is mention of open source licenses being not without problems. And
I saw something familiar out the corner of my eye: [insult deleted] in
the "most recent comments" thing at the right. Curious, I clicked on "My
Growing Library of Banned Books" and wouldn't you know it, it is indeed
Twisted, flaming some poor pro-copyright person who showed up at the blog
to criticise it. Talk about sticking your head in the lion's den,
especially when one of the lions is Twisted!
Upshot: I think I'll use BSD-type licenses for now. They're compatible
with almost anything, license-wise, including the GPL and Clojure's
license, and have a decent level of respect in the open source world. I
don't think I can go far wrong if I use the two-clause BSD license on my
code.
If anyone knows differently, or has any other insights on the licensing/
how-to-make-money/copyrights-are-they-good-or-evil issue as they apply to
Java/JVM developers, I wouldn't mind knowing what they have to say.
Please do try not to turn this into a useless anti-Twisted flamefest
though; this newsgroup has gone a good long time without one of those and
I for one like it just fine that way.
Mike Schilling
> Upshot: I think I'll use BSD-type licenses for now. They're compatible
> with almost anything, license-wise, including the GPL and Clojure's
> license, and have a decent level of respect in the open source world.
> I don't think I can go far wrong if I use the two-clause BSD license
> on my code.
The main difference between BSD and GPL is that BSD doesn't preclude use in
for-profit software. If that's your intent, it's a good choice.
Nigel Wade
The GPL makes no such exclusion. I presume that you've been reading the
FUD rather than the GPL.
This comes direct from the GPL FAQ:
"If I use a piece of software that has been obtained under the GNU GPL,
am I allowed to modify the original code into a new program, then
distribute and sell that new program commercially?
You are allowed to sell copies of the modified program commercially,
but only under the terms of the GNU GPL. Thus, for instance, you must
make the source code available to the users of the program as described
in the GPL, and they must be allowed to redistribute and modify it as
described in the GPL.
These requirements are the condition for including the GPL-covered
code you received in a program of your own. "
--
Nigel Wade
Thomas Pornin
> The main difference between BSD and GPL is that BSD doesn't preclude
> use in for-profit software.
That's only indirectly true. What the GPL does is that when you take GPL
code and reuse it in your own application, and distribute the result,
then whoever receives that application shall have access to your source
code as well under the GPL terms (not only the source code of the GPL
code you have included, but everything you have added to it, too). The
BSD license does not do that: it does not force you to give your source
code, and (that's important) it does not make _your_ application a
BSD-licensed piece of code.
Nominally, neither GPL nor BSD precludes on-profit software. The GPL
just makes it harder to make money out of the sells, because any of your
customers automatically gets entitled to receive the source code, and to
redistribute the software (modified or unmodified) under the GPL terms.
In particular, that customer may, legally, give copies of the software
(and source code) for free.
So the GPL is, in practice (not in theory), incompatible with for-profit
mass-market software. With more specialized software, GPL is not adverse
to money-making: if you make a big, expensive piece of code that you
plan to sell to a dozen big companies, you can use the GPL, because none
of your customers would be keen on redistributing the code under the GPL
(that would be giving the code for free to their competitors).
Also, theory has it that big money resides in support and maintenance
contracts. The software is then some kind of commercial argument, and
you can give it for free since users automatically become potential
customers for your support and maintenance activity. There again, it
does not work well with mass-market software.
There _is_ some mass-market software where profit and GPL should work
well together. These are video games. Theoretically, you could give away
the game engine, and sell the artistic content (which is more than 95%
of the game size, counted in bytes). The GPL has some convoluted wording
with regards to linking; it was initially designed for C. The GPL is
known to be a bit flaky (legally speaking) when applied to code written
in languages with dynamic loading (in particular Java). But it is still
clear that it is legal to distribute a GPL engine which loads
independently copyrighted pictures, music, video, 3D models...
(Please note that I have written "should work well". Theory and practice
are identical... in theory.)
--Thomas Pornin
David Lamb
I'd really like to understand the various FOSS licence issues better, so
please forgive the question if it's a diversion:
Does this quote mean that, although one can legally sell modified code,
there is little incentive for anyone to pay for it, since they can
obtain, use, modify, and redistribute for free?
Lew
>> This comes direct from the GPL FAQ:
>>
>> "If I use a piece of software that has been obtained under the GNU GPL,
>> am I allowed to modify the original code into a new program, then
>> distribute and sell that new program commercially?
>>
>> You are allowed to sell copies of the modified program commercially,
>> but only under the terms of the GNU GPL. Thus, for instance, you must
>> make the source code available to the users of the program as described
>> in the GPL, and they must be allowed to redistribute and modify it as
>> described in the GPL.
>>
>> These requirements are the condition for including the GPL-covered
>> code you received in a program of your own. "
David Lamb wrote:
> Does this quote mean that, although one can legally sell modified code,
> there is little incentive for anyone to pay for it, since they can
> obtain, use, modify, and redistribute for free?
The quote does not mean that.
You might choose to conclude that based on other data, but it's not what the
quote means.
The tendency to blame the license for the conclusions that one draws about it
causes misunderstanding and misuse or abuse of open-source licenses and FUD
among potential users of the software they cover.
The GPL as cited only says you have to provide the source code. It does not
preclude incentives to pay for it. For example, source code is really only
useful if you have a developer handy. Users that don't have one handy will
have to pay to have one handy in order to use the source code. They can pay a
third party to provide the GPLed code and make it work, or they can hire an
employee to do so. Either way, they're paying for the software.
And why would they redistribute it? It's open source; there's no onus on them
to distribute what any user can obtain on their own. So that provides no
disincentive to purchase the software.
Sure, they can obtain it for free, and theoretically even use it for free.
But what if they want more than the minimum functionality? Someone still has
to apply expertise to the use of the software, expertise that must be
purchased regardless of licensing fees.
Modify for free is no disincentive either - who wants to modify the code
rather than just use it without trouble?
So the only disincentives are the cost to obtain the software, which is always
rolled into the purchase cost even in non-open-source software, and - and -
that's it. The cost to use the software is roughly the same, and most
customers have no interest in the source code at all.
Being the one to provide what does interest them is where you make money. GPL
does not stand in the way of that at all.
--
Lew
Joshua Cranmer
> The main difference between BSD and GPL is that BSD doesn't preclude use in
> for-profit software. If that's your intent, it's a good choice.
A better way of phrasing it is that GPL code begets only GPL code,
whereas BSD code can be used more widely. GPL is actually rather crappy
for many libraries, since you can't use GPL code in many major open
source projects.
--
Beware of bugs in the above code; I have only proved it correct, not
tried it. -- Donald E. Knuth
Arne Vajhøj
The GPL license does not preclude that.
But the terms of the GPL license combined with cost minimizing
users makes certain business models not working.
Arne
Arne Vajhøj
For a traditional software sale model "here is a CD with the software
and a valid license key - if you want support then you have to
pay extra", then YES.
There are several companies selling GPL software and making money
using different business models.
Arne
Arne Vajhøj
> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd
> come to suspect that open source licensing is itself a source of trouble.
> Notably, Clojure's license is incompatible with the GPL, under which
> large chunks of other open source software is licensed.
And so what?
This means that you can not modify the Clojure compiler using
GPL code.
Very few libraries are GPL. LGPL and GPL with linking exception
was invented for libraries.
And it has no impact on people writing apps in Clojure.
> So I decided to do a little reading on copyright in general. Why does it
> even exist? The nominal purpose, it turns out, is to "promote the
> progress of science and the useful arts" by providing a way for the
> creators of any popular or important work to ensure remuneration,
> basically. Which smells suspiciously like a grant of monopoly -- which,
> barring the notion of "fair use", it basically is.
If everybody could copy software exactly as they wanted then
I am pretty sure that the software industry would be in a very
poor shape.
> Furthermore there are
> a LOT of blogs out there expressing serious criticism of copyright,
> pointing out that fair use is not in most cases a workable defense even
> when it should be, and that copyright has been twisted away from its
> original purpose by corporations seeking to extend and tighten their
> control over lucrative media and software properties.
Sounds like blogs from teenagers that wants to be able to download
everything for free and have parents to pay the bills.
> It thus seems that copyright was twisted away from its original purpose,
> to which it might have been poorly suited to begin with, and open source
> licenses try to twist it back toward that purpose.
Not really.
Open source use copyright the exact same way as closed source. The
only way.
Open source has different license terms than closed source, but that
does not change the copyright as such.
> Upshot: I think I'll use BSD-type licenses for now. They're compatible
> with almost anything, license-wise, including the GPL and Clojure's
> license, and have a decent level of respect in the open source world. I
> don't think I can go far wrong if I use the two-clause BSD license on my
> code.
When you have copyright for the code, then you decide on the
license.
I think that you can generally divide in:
- strong copyleft (GPL)
- weak copyleft (LGPL etc.)
- permissive (Apache, BSD etc.)
Due to Apache's strong involvement in Java technologies, then I
think there are more Java code under Apache than BSD license.
Arne
ClassCastException
> Does this quote mean that, although one can legally sell modified code,
> there is little incentive for anyone to pay for it, since they can
> obtain, use, modify, and redistribute for free?
Red Hat makes quite a bit of money selling copies of GPL'd software on
physical media.
This might have something to do with the fact that Aquafina makes quite a
bit of money bottling and selling stuff that pretty much all of their
potential customers can get out of a faucet for free.
It might also have something to do with the fact that the entertainment
industry is not, contrary to popular belief, losing revenues to piracy.
Declining sales of things like DVDs and recorded music have a complex web
of causes, in which the effects of online piracy are not reliably
different from zero according to the statistical studies.
It *is* possible to compete with free, generally by segmenting the market
and taking the high end somehow. Image, branding, convenience, or some
other thing can differentiate you from the free competition. For example,
bottled water comes complete with a new, clean bottle to carry it in and
can be an image thing. Red Hat sold on a CD is as simple to get and
install as buy in store, stick in disk drive, run, similar to other store-
bought software, without having to do anything special like download gigs
of .iso possibly over a shoddy line and then figure out how to turn
a .iso file into a working bootable install disc and etc., so there it's
convenience. If you have the money, buying a DVD is probably more
convenient than trying to find a good, non-foreign-language rip on
BitTorrent and buying a Blu-Ray is almost certainly more convenient than
trying to find a good, non-foreign-language HIGH-DEF rip on BT. CDs are
pretty much toast but a 99 cent iTunes download is more convenient for
anyone who owns a credit card than an attempt to find and download a good
rip on LimeWire.
The moral of the story: the GPL absolutely is NOT incompatible with
profiting from selling software, NOT EVEN if you restrict your business
model to selling copies. When you broaden that to include selling support
and ancillary merchandising of various kinds, it's even more possible.
Someone suggested game content, but that's also ultimately copyable --
unless it resides server-side. An MMORPG could be entirely GPL'd,
including the server code, and even without any copyright enforcement on
the game content, and still be easily profitable. First, you run servers
time and bandwidth on which are scarce and charge for this. Competitors
can duplicate your service and, with some effort, duplicate your content
more-or-less, but any such competitor can be kept ahead of by developing
fresh content (necessary anyway to hang onto long-time players). On top
of that if your game gets popular enough you can make a sideline selling
plush toys, action figures, or whatnot of game characters, potentially.
If these are not trademarked or otherwise "protected" you'll have
competition there too, but there's plenty of competition selling, say,
chicken nuggets or chairs, but plenty of money to be made in the chicken
nugget and chair industries too, and your game's popularity will grow the
market for this merchandise and, with it, your slice of the merchandising
pie for those characters, even if that slice is not the whole pie.
Indeed, almost anything you or your competitors do will grow the overall
market and cause all of you to sell more of everything.
A strange game -- the only losing move is not to play.
ClassCastException
> On 27-05-2010 22:58, ClassCastException wrote:
>> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd
>> come to suspect that open source licensing is itself a source of
>> trouble. Notably, Clojure's license is incompatible with the GPL, under
>> which large chunks of other open source software is licensed.
>
> And so what?
>
> This means that you can not modify the Clojure compiler using GPL code.
>
> Very few libraries are GPL. LGPL and GPL with linking exception was
> invented for libraries.
>
> And it has no impact on people writing apps in Clojure.
Actually, it does; if you GPL Clojure code and distribute it it will
violate at least one license because the GPL'd code will link against
(and thus try to force the "viral" GPL upon) code that comes with Clojure
and has a GPL-incompatible license.
You can add a linking exception to the GPL on your Clojure code that
permits linking against Clojure libraries and the like without requiring
the GPL attach to those, but this has two problems:
1. If existing, say, Java code is GPL, and you want to use it in your
Clojure project, you still can't, since the GPL without linking
exception will "contaminate" your own code and then the base Clojure
code that's not GPLable.
2. This, and similar situations, will lead to the proliferation of
hundreds of almost-GPLs with different and incompatible linking
exceptions.
In hindsight, the linking requirement in the GPL, which was made in the
days of C, was a big mistake. But it's one it's apparently way too late
to fix.
The GPL v3 apparently tries. It has something in it about an automatic
linking exception for the code's programming language's "base libraries",
but someone said this was vague enough or otherwise had loopholes that
made it impossible to confidently apply to Clojure's GPL-incompatible
libraries.
>> So I decided to do a little reading on copyright in general. Why does
>> it even exist? The nominal purpose, it turns out, is to "promote the
>> progress of science and the useful arts" by providing a way for the
>> creators of any popular or important work to ensure remuneration,
>> basically. Which smells suspiciously like a grant of monopoly -- which,
>> barring the notion of "fair use", it basically is.
>
> If everybody could copy software exactly as they wanted then I am pretty
> sure that the software industry would be in a very poor shape.
I actually doubt this; I think it would work rather differently from now
in some ways, but that people would have found a way to make it work and
to profit in it. Plenty of businesses profit from open source software in
various ways, including by selling support or simply by funding
development of open source software that they use in-house and get
productivity gains from, and, by funding it, get more influence to have
features they'd find useful added and the bugs that particularly harm
their productivity prioritized.
> Sounds like blogs from teenagers that wants to be able to download
> everything for free and have parents to pay the bills.
I have my doubts whether Against Monopoly (run by a pair of degreed
economists) and Techdirt (run by a successful dot-com entrepreneur) are
"from teenagers that want to be able to download everything for free and
have parents to pay the bills". :)
>> It thus seems that copyright was twisted away from its original
>> purpose, to which it might have been poorly suited to begin with, and
>> open source licenses try to twist it back toward that purpose.
>
> Not really.
>
> Open source use copyright the exact same way as closed source. The only
> way.
>
> Open source has different license terms than closed source, but that
> does not change the copyright as such.
I'm sorry you don't seem to get what I'm driving at.
Copyright was born from the theory that letting authors of "writings and
discoveries" close off access and control the use of their work, e.g. to
set up a tollbooth, would promote progress.
At least in the case of software, this turned out to be wrong (in some
opinions), and the GPL's "copyleft" was specifically designed to force
copyright to do the reverse: force access open as widely as possible, by
requiring publication of the source code and disallowing monopoly. The
theory this time being that maximizing access and minimizing any one
vendor's control over program code would promote progress.
Judging by the stellar progress made in improving Linux since its
inception, copyleft is at least as viable as traditional exclusive
copyright in promoting progress in software.
However, both have proven capable of getting in the way in various
(separate sets of) situations.
Tom Anderson
> On Fri, 28 May 2010 20:31:14 -0400, Arne Vajh?j wrote:
>
>> On 27-05-2010 22:58, ClassCastException wrote:
>>> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd
>>> come to suspect that open source licensing is itself a source of
>>> trouble. Notably, Clojure's license is incompatible with the GPL, under
>>> which large chunks of other open source software is licensed.
>>
>> And so what?
>>
>> This means that you can not modify the Clojure compiler using GPL code.
>>
>> Very few libraries are GPL. LGPL and GPL with linking exception was
>> invented for libraries.
>>
>> And it has no impact on people writing apps in Clojure.
>
> Actually, it does; if you GPL Clojure code and distribute it it will
> violate at least one license because the GPL'd code will link against
> (and thus try to force the "viral" GPL upon) code that comes with Clojure
> and has a GPL-incompatible license.
Wrong. The GPL does not forbid you from linking against any libraries at
all. It (effectively) forbids you from *redistributing* the *products* of
linking against libraries that are not GPL'd. Do you want to distribute
binaries that include both your GPL'd Clojure code and the Clojure
runtime? No. So you're fine.
This is an important point: *none* of the open-source licenses restrict
what you do with code on your own machine, or inside your own company. The
*only* apply to *redistribution* of the code.
tom
--
This isn't right. This isn't even wrong.
David Lamb
> Copyright was born from the theory that letting authors of "writings and
> discoveries" close off access and control the use of their work, e.g. to
> set up a tollbooth, would promote progress.
Well, that theory *used* to be true; Dickens apparently got no royalties
from "pirated" US editions of his works. But that was a long time ago
when the world was very different.
ClassCastException
> On Sat, 29 May 2010, ClassCastException wrote:
>
>> On Fri, 28 May 2010 20:31:14 -0400, Arne Vajh?j wrote:
>>
>>> On 27-05-2010 22:58, ClassCastException wrote:
>>>> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd
>>>> come to suspect that open source licensing is itself a source of
>>>> trouble. Notably, Clojure's license is incompatible with the GPL,
>>>> under which large chunks of other open source software is licensed.
>>>
>>> And so what?
>>>
>>> This means that you can not modify the Clojure compiler using GPL
>>> code.
>>>
>>> Very few libraries are GPL. LGPL and GPL with linking exception was
>>> invented for libraries.
>>>
>>> And it has no impact on people writing apps in Clojure.
>>
>> Actually, it does; if you GPL Clojure code and distribute it it will
>> violate at least one license because the GPL'd code will link against
>> (and thus try to force the "viral" GPL upon) code that comes with
>> Clojure and has a GPL-incompatible license.
>
> Wrong. The GPL does not forbid you from linking against any libraries at
> all. It (effectively) forbids you from *redistributing* the *products*
> of linking against libraries that are not GPL'd. Do you want to
> distribute binaries that include both your GPL'd Clojure code and the
> Clojure runtime? No.
Yes. Any packaging of an application for end-user use (that an end-user
is realistically going to be able to install easily) is going to include
the runtimes and thus violate one license or both. The effect is for
GPL'd Clojure code to effectively be stamped "hacker use only" and be
inaccessible for normal, end-user use.
Use it privately? Sure. Share with other Clojure hackers? Ditto. Make a
killer game with GPL'd Clojure code and post it somewhere, or write and
publish something to beat Photoshop *and* the GIMP, or whatever? Uh-uh,
sorry, no can do.
ClassCastException
Interestingly, Dickens actually profited from this. Hence why I said that
that was the *theory* behind copyright; there is some evidence that that
theory was false, or at best semi-true. In actual practice, the existence
of art of many kinds for thousands of years before anyone thought of the
concept of a copyright is perhaps the strongest evidence that it wasn't
ever actually needed. A case for it being needed for some very expensive
modern art forms such as blockbuster movies might be made. Open source
development models, including with licenses like BSD that permit closed-
source derivative works, point to its *not* being necessary for software
innovation to occur (without copyright software development would
probably be very similar to how it actually is in BSD-like-license-
dominated ecosystems, such as those surrounding BSD Unix itself and
Apache; there could be some closed-source derivative works but this
potential doesn't seem to have killed BSD or Apache).
Perhaps even more interesting is that a similar case can be made for
getting rid of patents on most everything except pharmaceuticals, where
they seem both most needed (drug development costs make $200 million
blockbuster movie productions look cheap at times) and most controversial
(artificially inflated drug prices can actually *kill*, unlike
artificially inflated movie prices or artificially inflated software
prices).
Cleverer minds than mine continue to argue, cogently at times and quite
acrimoniously at times, both sides of the debate, in the comments of both
the blogs I named elsewhere in this thread. (And, at one of the blogs,
there's also Twisted ... apparently taking a particularly radical anti-
copyright anti-patent position ... for whatever that's worth.)
Meanwhile open source continues to win practical victories. The recent
school laptop scandal, for example: the initial scandal just cries out
for owner-override rights for hardware owners, so against any species of
so-called "trusted computing" such as has been pushed at times by
Microsoft as a way of limiting "piracy" (and such as has quietly been
actually implemented in game consoles, DVD/BluRay drives/players, and
Apple's iPod/iPhone/iPad portables); the later revelation of serious
security flaws in the "remote administration" software that was snuck
onto those machines points to the dangers of using closed source remote
administration, operating system, or security software of any kind: at
best you may have severe security flaws hidden from peer review but
eventually findable by a determined black hat; at worst the nightmare
scenario from the movie "The Net" where some popular closed-source
firewall software or similarly turns out to have a deliberate back door
and its makers, once it becomes very widely deployed, start using it in a
nefarious world-domination plot or to pull off massive heists or
whatever. (Imagine a deliberate back door in Windows Vista and a
"Praetorian" conspiracy within Microsoft. Shudder. Install Linux. Sigh
with relief. Recall that many government and other trusted-by-the-public
computer systems and networks are riddled with Microsoft software.
Shudder. Note that Apple has even tighter control over the iFoo
ecosystem. Shudder. Ditch your iPhone for an Android phone. Sigh with
relief. Watch as a controversy breaks out over Amazon's retroactive
unpurchasing of Kindle customers' copies of Nineteen Eighty-Four by
Amazon and how all their annotations unexisted because Amazon decided
they were doubleplusungood, or something like that. Shudder. Ditch Kindle
for ... er, wait, dammit!, the iPad is just as evil. Groan and use the
Android phone with its inferior reading display and wait for a googlePad
or similar device to hit the shelves. Pirate what isn't available at
Project Gutenberg, because officially-sanctioned ebooks aren't available
for Android platforms, at least not yet; maybe buy a paperback copy of
each pirated book so the authors get some money. Or perhaps just lug
actual paper books around and forget e-reading for now. Sigh with relief;
now your books can't be unpublished by Amazon, Apple, or anybody else
short of jackbooted thugs breaking down your door and prying them from
your cold, dead fingers, and you trust your 12-gauge named Old Betsy
together with an electric fence, basic property rights, and the 1st, 2nd,
and 4th Amendments to keep any wannabe Guy Montagues at bay. Remember how
most of the AI-takeover apocalypses also have a similar element --
defense contractors buying from US Robotics in I, Robot, defense
contractors buying from Cyberdyne in Terminator, etc., and these having
intentional backdoors in them that let Viki/Skynet/Colossus/whatever take
over. Shudder.)
David Lamb
> Also, once they have the
> software, how do you make them pay you? The GPL already
> entitles them to legally own it without any payment to you.
Well, contract law would still apply. "I'll hand over this disk and
you'll pay me" with suitable legalize boilerplate ought to mean they
have to pay you. I suppose one could start an argument about the seller
not having the money to sue.
ClassCastException
> Thomas Pornin <por...@bolet.org> writes:
>>So the GPL is, in practice (not in theory), incompatible with for-profit
>
> If you put a box in a shop with some software for $329, people who
> think that it addresses their needs will buy it. It might contain a
> sheet of paper with the GPL and directory with the source code. Most
> consumers will not read either of them. There are bought many goods in
> our world that one also might get for free by some other means.
Perhaps. Though at that price point they might feel ripped off if they
later learned they could have gotten it for free elsewhere.
On the other hand I can see someone spending even $40 on a CD that's nice
and easy to install and considering that okay even if they find they
could have had it for free as a downloadable .iso file to figure out what
to do with to install it. People will willingly pay, sometimes quite a
lot, for convenience. Hence fast food, more expensive than home cooking
and often not as good yet very very profitable.
> This would entitle /you/ to redistribute it, and your customers might
> not like this. Also, once they have the software, how do you make them
> pay you? The GPL already entitles them to legally own it without any
> payment to you.
Obviously you'd structure the transaction as an up-front sale: swap the
money for a disc.
>>Also, theory has it that big money resides in support and maintenance
>>contracts. The software is then some kind of commercial argument, and
>>you can give it for free since users automatically become potential
>>customers for your support and maintenance activity.
>
> Or for the support and maintenance of any other skilled party ...
Still gives you an incentive to develop the code:
1. The more valuable you make the software, the larger the support pie,
and the larger the support pie, the larger your slice of it,
regardless of what percentage of the whole pie that slice might be.
2. The more of it you've developed, the better you know the codebase
relative to any other skilled party, and the better you know the
codebase relative to any other skilled party, the better you can make
your support offering compared to anyone else's. This makes your slice
of the pie larger.
So, you can potentially get yourself a larger slice of a larger pie by
participating in developing the open-source software, compared to if you
let other people do all the development work on it. Developing it builds
in-house expertise with that software, and it's the time of your experts
that you're selling; making those experts better experts makes that time
more valuable, and so does making the thing they're experts on more
popular.
Make that time more valuable and you can attract more customers so you
can sell more of it, charge more for it, or whatever.
Arne Vajhøj
> On Fri, 28 May 2010 16:09:36 -0400, David Lamb wrote:
>> Does this quote mean that, although one can legally sell modified code,
>> there is little incentive for anyone to pay for it, since they can
>> obtain, use, modify, and redistribute for free?
>
> Red Hat makes quite a bit of money selling copies of GPL'd software on
> physical media.
>
> This might have something to do with the fact that Aquafina makes quite a
> bit of money bottling and selling stuff that pretty much all of their
> potential customers can get out of a faucet for free.
Not a good comparison.
When you buy RHEL from Redhat instead of downloading Centos, then
you get support from Redhat.
> It might also have something to do with the fact that the entertainment
> industry is not, contrary to popular belief, losing revenues to piracy.
> Declining sales of things like DVDs and recorded music have a complex web
> of causes, in which the effects of online piracy are not reliably
> different from zero according to the statistical studies.
That "fact" is very disputed.
> The moral of the story: the GPL absolutely is NOT incompatible with
> profiting from selling software, NOT EVEN if you restrict your business
> model to selling copies.
Companies making monet on GPL usually do it by selling support
or by dual license (GPL and commercial).
The GPL license does not prohibit it, but the terms of GPL
plus basic economics do that you need something other than
selling copies to prosper.
Arne
Arne Vajhøj
>
> people who think that it addresses their needs will buy it.
> It might contain a sheet of paper with the GPL and directory
> with the source code. Most consumers will not read either of
> them. There are bought many goods in our world that one
> also might get for free by some other means.
The word will get around.
Arne
Arne Vajhøj
> On Sat, 29 May 2010 21:54:24 +0000, Stefan Ram wrote:
>> Thomas Pornin<por...@bolet.org> writes:
>>> So the GPL is, in practice (not in theory), incompatible with for-profit
>>
>> If you put a box in a shop with some software for $329, people who
>> think that it addresses their needs will buy it. It might contain a
>> sheet of paper with the GPL and directory with the source code. Most
>> consumers will not read either of them. There are bought many goods in
>> our world that one also might get for free by some other means.
>
> Perhaps. Though at that price point they might feel ripped off if they
> later learned they could have gotten it for free elsewhere.
>
> On the other hand I can see someone spending even $40 on a CD that's nice
> and easy to install and considering that okay even if they find they
> could have had it for free as a downloadable .iso file to figure out what
> to do with to install it. People will willingly pay, sometimes quite a
> lot, for convenience.
Maybe, but you will not get rich by selling $40 software.
Arne
Arne Vajhøj
> On Fri, 28 May 2010 20:31:14 -0400, Arne Vajhøj wrote:
>> On 27-05-2010 22:58, ClassCastException wrote:
>>> In dealing with Java, OpenJDK, and Clojure stuff in recent months I'd
>>> come to suspect that open source licensing is itself a source of
>>> trouble. Notably, Clojure's license is incompatible with the GPL, under
>>> which large chunks of other open source software is licensed.
>>
>> And so what?
>>
>> This means that you can not modify the Clojure compiler using GPL code.
>>
>> Very few libraries are GPL. LGPL and GPL with linking exception was
>> invented for libraries.
>>
>> And it has no impact on people writing apps in Clojure.
>
> Actually, it does; if you GPL Clojure code and distribute it it will
> violate at least one license because the GPL'd code will link against
> (and thus try to force the "viral" GPL upon) code that comes with Clojure
> and has a GPL-incompatible license.
If you want to make it work, then you can make it work.
Use a linking exception.
Use GPL V3.
Distribute the source and ask the users to get Clojure themselves
and build.
Make a very narrow definition of how GPL should be interpreted. The
point is that the two copyright holders you and Clojure will not
bring you to court.
If you want to see a problem, then you can definitely find problems.
> You can add a linking exception to the GPL on your Clojure code that
> permits linking against Clojure libraries and the like without requiring
> the GPL attach to those, but this has two problems:
>
> 1. If existing, say, Java code is GPL, and you want to use it in your
> Clojure project, you still can't, since the GPL without linking
> exception will "contaminate" your own code and then the base Clojure
> code that's not GPLable.
> 2. This, and similar situations, will lead to the proliferation of
> hundreds of almost-GPLs with different and incompatible linking
> exceptions.
Not much Java code intended for inclusion in other programs are GPL.
And "similar" is not specific.
> The GPL v3 apparently tries. It has something in it about an automatic
> linking exception for the code's programming language's "base libraries",
> but someone said this was vague enough or otherwise had loopholes that
> made it impossible to confidently apply to Clojure's GPL-incompatible
> libraries.
"someone", "vague enough" and "otherwise had loopholes" is just empty
words.
I would not be concerned over those.
>>> So I decided to do a little reading on copyright in general. Why does
>>> it even exist? The nominal purpose, it turns out, is to "promote the
>>> progress of science and the useful arts" by providing a way for the
>>> creators of any popular or important work to ensure remuneration,
>>> basically. Which smells suspiciously like a grant of monopoly -- which,
>>> barring the notion of "fair use", it basically is.
>>
>> If everybody could copy software exactly as they wanted then I am pretty
>> sure that the software industry would be in a very poor shape.
>
> I actually doubt this; I think it would work rather differently from now
> in some ways, but that people would have found a way to make it work and
> to profit in it. Plenty of businesses profit from open source software in
> various ways, including by selling support or simply by funding
> development of open source software that they use in-house and get
> productivity gains from, and, by funding it, get more influence to have
> features they'd find useful added and the bugs that particularly harm
> their productivity prioritized.
Since open source requires copyright then no copyright means
no open source.
So the existence of open source does not prove that copyright is
not necessary.
>> Sounds like blogs from teenagers that wants to be able to download
>> everything for free and have parents to pay the bills.
>
> I have my doubts whether Against Monopoly (run by a pair of degreed
> economists) and Techdirt (run by a successful dot-com entrepreneur) are
> "from teenagers that want to be able to download everything for free and
> have parents to pay the bills". :)
Given that you do not provide links or names or anything, then
the teenager guess is as good as any.
>>> It thus seems that copyright was twisted away from its original
>>> purpose, to which it might have been poorly suited to begin with, and
>>> open source licenses try to twist it back toward that purpose.
>>
>> Not really.
>>
>> Open source use copyright the exact same way as closed source. The only
>> way.
>>
>> Open source has different license terms than closed source, but that
>> does not change the copyright as such.
>
> I'm sorry you don't seem to get what I'm driving at.
>
> Copyright was born from the theory that letting authors of "writings and
> discoveries" close off access and control the use of their work, e.g. to
> set up a tollbooth, would promote progress.
>
> At least in the case of software, this turned out to be wrong (in some
> opinions), and the GPL's "copyleft" was specifically designed to force
> copyright to do the reverse: force access open as widely as possible, by
> requiring publication of the source code and disallowing monopoly. The
> theory this time being that maximizing access and minimizing any one
> vendor's control over program code would promote progress.
>
> Judging by the stellar progress made in improving Linux since its
> inception, copyleft is at least as viable as traditional exclusive
> copyright in promoting progress in software.
>
> However, both have proven capable of getting in the way in various
> (separate sets of) situations.
There may be many different ways to use copyright.
But copyright works identical for all - open source or closed
source.
In fact open source are some of the fiercest enforcer of copyright,
because:
* they need it even more when they hand out source code
* a large fraction of open source people do not believe in
patents, so they only have copyright
Arne
Arne Vajhøj
> The GPL as cited only says you have to provide the source code.
Actually it does a lot more.
Open source means a lot more than access to the source code.
> It does
> not preclude incentives to pay for it.
Correct.
> For example, source code is
> really only useful if you have a developer handy. Users that don't have
> one handy will have to pay to have one handy in order to use the source
> code. They can pay a third party to provide the GPLed code and make it
> work, or they can hire an employee to do so. Either way, they're paying
> for the software.
For custom software the difference between open source and closed
source are not that important.
But for consumer software it is bigger, because they will never
want it customized.
> And why would they redistribute it? It's open source; there's no onus on
> them to distribute what any user can obtain on their own. So that
> provides no disincentive to purchase the software.
Bullshit.
If the open source is distributed for free, then there are no
incentive for users to redistribute.
But if it is distributed for a cost (above pocket change), then
there sure is an incentive.
Let us say that your favorite Linux distro was sold for $1000
and you bought a copy. I am pretty sure that you would be willing
to provide your brother and your best friend with a copy. And
they would also be willing to provide it to a few people.
> Sure, they can obtain it for free, and theoretically even use it for
> free. But what if they want more than the minimum functionality? Someone
> still has to apply expertise to the use of the software, expertise that
> must be purchased regardless of licensing fees.
This is again custom software not consumer software.
Arne
Mike Schilling
> On Thu, 27 May 2010 23:21:16 -0700, Mike Schilling wrote:
>
>> ClassCastException wrote:
>>> Upshot: I think I'll use BSD-type licenses for now. They're
>>> compatible with almost anything, license-wise, including the GPL
>>> and Clojure's license, and have a decent level of respect in the
>>> open source world. I don't think I can go far wrong if I use the
>>> two-clause BSD license on my code.
>>
>> The main difference between BSD and GPL is that BSD doesn't preclude
>> use in for-profit software. If that's your intent, it's a good
>> choice.
>
> The GPL makes no such exclusion. I presume that you've been reading
> the FUD rather than the GPL.
>
> This comes direct from the GPL FAQ:
>
> "If I use a piece of software that has been obtained under the GNU
> GPL, am I allowed to modify the original code into a new program, then
> distribute and sell that new program commercially?
>
> You are allowed to sell copies of the modified program
> commercially, but only under the terms of the GNU GPL.
Right. The GPL is viral, resulting in restictions on the code you add to
it.. BSD is not.
Bent C Dalager
>
> Since open source requires copyright then no copyright means
> no open source.
If open source "requires" copyright then it is only because in the
current legal regime copyright is mandatory.
It does help, of course, that copyright is such a huge and
indiscriminate sledgehammer you can use it to kill just about
/anything/, including copyright. Open source will tend to use the best
tool for the job even if that tool is copyright.
> In fact open source are some of the fiercest enforcer of copyright,
> because:
> * they need it even more when they hand out source code
> * a large fraction of open source people do not believe in
> patents, so they only have copyright
This doesn't make much sense. Open source doesn't use copyright to
protect its source from being copied and so they do not "have
copyright" in the sense that copyright is protecting their business
model. To the extent copyright is used as a tool at all in open source
licenses it tend to be used to force the source to remain open but
this then is an ideological goal, not a means with which to protect
one's business in any way.
Neither patents nor copyright are needed for software development to
thrive so it seems to me your conclusion is based on false premises.
Cheers,
Bent D
--
Bent Dalager - b...@pvv.org - http://www.pvv.org/~bcd
powered by emacs
Lew
>> In fact open source are some of the fiercest enforcer of copyright,
>> because:
>> * they need it even more when they hand out source code
>> * a large fraction of open source people do not believe in
>> patents, so they only have copyright
Bent C Dalager wrote:
> This doesn't make much sense. Open source doesn't use copyright to
> protect its source from being copied and so they do not "have
False. They do use copyright to give them legally assertable authority to
determine the rules under which source can be copied, and they therefore use
copyright to protect source from being copied outside the boundaries of the
license.
> copyright" in the sense that copyright is protecting their business
> model. To the extent copyright is used as a tool at all in open source
> licenses it tend to be used to force the source to remain open but
> this then is an ideological goal, not a means with which to protect
> one's business in any way.
>
> Neither patents nor copyright are needed for software development to
> thrive so it seems to me your conclusion is based on false premises.
Arne never proffered the premise that patents or copyright are needed for
software development, so that does not apply. He only said they need it more
when they give away source code, clearly because of the need for all the legal
ammunition at your disposal when you give things away freely.
What Arne did suggest was that, in fact, open-source licenses rely on
copyright for their enforceability, and that copyright holders do, in fact,
rely on that legal theory to enforce their rights, and they are, in fact,
fierce enforcers of their copyright rights. All these are evident in the
world as a matter of record.
Whether it's true or relevant that a "large fraction of open source people do
not believe in patents, so they only have copyright [to protect their
interests]" is debatable, and you'd be justified to challenge Arne on both the
evidence of the assertion and the applicability to the use of copyright law to
enforce open-source licenses. The statement itself is not very far-fetched,
so I'm willing to accept that many "open-source people" favor copyright
protection over patent protection. They probably have very well thought-out
reasaons for this preference, no doubt well founded in legal and tactical
concerns.
I believe Arne could also make a very strong case for relevance. His stated
premises seem at least plausible, certainly not worthy of dismissal out of
hand, his conclusions are strongly explained and his asserted facts are mostly
in evidence.
There's a difference between an assertion that copyright law is used to
enforce ownership rights, which is quite clearly so, and an assertion that
there is no better way to enforce those rights. Perhaps you mean to assert
that there is such a better way, which is an interesting proposition in itself
depending on how one proposes to go about it.
--
Lew
Arne Vajhøj
> On 2010-05-30, Arne Vajhøj<ar...@vajhoej.dk> wrote:
>> Since open source requires copyright then no copyright means
>> no open source.
>
> If open source "requires" copyright then it is only because in the
> current legal regime copyright is mandatory.
Copyright is not mandatory.
It is perfectly legal to put code in public domain.
Open source chose not to do so.
Because they want to have some license conditions.
No copyright => no license conditions => no open source.
> It does help, of course, that copyright is such a huge and
> indiscriminate sledgehammer you can use it to kill just about
> /anything/, including copyright. Open source will tend to use the best
> tool for the job even if that tool is copyright.
Open source does not kill copyright. They just use copyright for
a different purpose than MS, IBM, Oracle etc. (or should I say for
those companies close source offerings - they do have open source
offerings as well).
>> In fact open source are some of the fiercest enforcer of copyright,
>> because:
>> * they need it even more when they hand out source code
>> * a large fraction of open source people do not believe in
>> patents, so they only have copyright
>
> This doesn't make much sense. Open source doesn't use copyright to
> protect its source from being copied and so they do not "have
> copyright" in the sense that copyright is protecting their business
> model. To the extent copyright is used as a tool at all in open source
> licenses it tend to be used to force the source to remain open but
> this then is an ideological goal, not a means with which to protect
> one's business in any way.
Copyright is a tool.
That tool is used for different purposes:
- traditional closed source - to make money
- open source - give end users certain rights and limit
some responsibilities
- Roedy - make sure that the code is not used for military purposes
etc.
Copyright may have been invented for the first purpose, but copyright
is copyright no matter what purpose it is used for.
> Neither patents nor copyright are needed for software development to
> thrive so it seems to me your conclusion is based on false premises.
Neither traditional closed source nor open source would be possible
without copyright.
I am not aware of any company making a living from public domain
software.
Based on that I am very skeptical about software industry in a world
with no copyright.
I don't think anyone would miss software patents if they disappeared.
That is anyone in software. The lawyers would definitely miss it.
Arne
Roedy Green
<zjkg3...@gmail.invalid> wrote, quoted or indirectly quoted someone
who said :
>So I decided to do a little reading on copyright in general. Why does it
>even exist? The nominal purpose, it turns out, is to "promote the
>progress of science and the useful arts" by providing a way for the
>creators of any popular or important work to ensure remuneration,
>basically. Which smells suspiciously like a grant of monopoly
With copyright there is no problem with someone creating a similar
work -- unlike patent, so for example nobody can corner the market on
"how to program in Java" books.
It is required or it would be almost impossible to make money writing
books or programs. Custom programs would survive, but no one-size-fits
all.
--
Roedy Green Canadian Mind Products
http://mindprod.com
Have you ever noticed that any computer search in the movies, is always linear, with, for example, candidate fingerprints flashing up on the screen one after another? The public is still under the delusion that electronic files are microscopic filing cabinets made out of tiny wires or magnetic patches inside the computer. Most lay people are surprised that it is easy for a computer to file things simultaneously by a dozen different schemes, and that they can have any report printed in any number of different sorted orders. With physical files, they are limited to one ordering/access.
Arne Vajhøj
> Whether it's true or relevant that a "large fraction of open source
> people do not believe in patents, so they only have copyright [to
> protect their interests]" is debatable,
Obviously it is.
But there are lots of examples of it.
A couple of very high profile open source personalities:
http://www.gnu.org/philosophy/stallman-mec-india.html
http://news.techworld.com/applications/3059/torvalds-joins-in-anti-patent-attack/
Arne
Bent C Dalager
> On 30-05-2010 08:45, Bent C Dalager wrote:
>> On 2010-05-30, Arne Vajhøj<ar...@vajhoej.dk> wrote:
>>> Since open source requires copyright then no copyright means
>>> no open source.
>>
>> If open source "requires" copyright then it is only because in the
>> current legal regime copyright is mandatory.
>
> Copyright is not mandatory.
>
> It is perfectly legal to put code in public domain.
How do you propose to do this? The Berne Convention clearly says that
all works that can be covered by copyright are covered by default. The
only mention of the public domain in the convention is in saying that
expired works enter into the public domain. There is no provision for
explicitly entering a work into the public domain before its expiry
date.
Furthermore, my own national laws stipulate that a subset of the
privileges offered by copyright are inalienable, that is you could not
sign them away even if every fibre of your being trembled with desire
to do so: The law won't let you.
I do not see that there is any way for an author to put his work into
the public domain. He might write "this work is in the public domain"
on the front page of course but it is not clear that this holds any
legal weight at all and indeed my own country says you still keep
certain inalienable copyright privileges over it.
The only apparent way to mitigate the impact of copyright over one's
own work is to write very generous license terms for it. As you point
out, this means that your work is still protected under copyright
however generous you are. This is as it must be under the current
legal regime.
> Open source chose not to do so.
>
> Because they want to have some license conditions.
I believe it is rather that they realize they /must/ write a license
because the law effectively requires it of them and then they tend to
tag on a pet provision or two after the "use as you wish" bit of the
open source license.
> No copyright => no license conditions => no open source.
Why is it that software which has the most generous license term
possible under the law (that is, no conditions), cannot be "open
source"? Surely if it were /possible/ to release source code into the
public domain then this would be just as open as an open source
license is?
>> It does help, of course, that copyright is such a huge and
>> indiscriminate sledgehammer you can use it to kill just about
>> /anything/, including copyright. Open source will tend to use the best
>> tool for the job even if that tool is copyright.
>
> Open source does not kill copyright. They just use copyright for
> a different purpose than MS, IBM, Oracle etc. (or should I say for
> those companies close source offerings - they do have open source
> offerings as well).
Open source developers certainly have a varied set of motivations for
their chose of license; some would prefer to have their license
destroy copyright while others believe copyright is essential for them
to earn a living.
>> Neither patents nor copyright are needed for software development to
>> thrive so it seems to me your conclusion is based on false premises.
>
> Neither traditional closed source nor open source would be possible
> without copyright.
But of course they would: closed source would be protected as a trade
secret, which is effectively what is happening already anyway, and any
source that anyone actually distributed would be open source by
default.
> I am not aware of any company making a living from public domain
> software.
>
> Based on that I am very skeptical about software industry in a world
> with no copyright.
While the estimates vary, around 80% of the software industry consists
of various in-house projects: Customer tracking, inventory management,
project reporting, niche expert systems, etc. This 80% of the industry
would benefit greatly from the absence of copyright since they don't
need it anyway: their software is only applicable to, distributed
within, and used by their own organisation. The absence of copyright
would make their job much easier - and less expensive - by not forcing
them to reinvent the same wheel that hundreds of others have already
designed to death time and time again before them.
Shelfware might or might not be impacted by the absence of copyright,
I suspect that for the most part that would redesign their business
model slightly and keep on earning money.
> I don't think anyone would miss software patents if they disappeared.
> That is anyone in software. The lawyers would definitely miss it.
That is certainly one thing we can agree about :-)
ClassCastException
> On 29-05-2010 03:56, ClassCastException wrote:
>> This might have something to do with the fact that Aquafina makes quite
>> a bit of money bottling and selling stuff that pretty much all of their
>> potential customers can get out of a faucet for free.
>
> Not a good comparison.
Of course it is. It's the *classic* example of successfully competing
with free!
>> It might also have something to do with the fact that the entertainment
>> industry is not, contrary to popular belief, losing revenues to piracy.
>> Declining sales of things like DVDs and recorded music have a complex
>> web of causes, in which the effects of online piracy are not reliably
>> different from zero according to the statistical studies.
>
> That "fact" is very disputed.
By idiots. The studies' results are what they are, even if many refuse to
accept them.
Anyone who clings to a theory despite a growing weight of contrary
evidence has ceased to practice science and has become a pseudoscientific
quack at best and a religious nut at worst.
>> The moral of the story: the GPL absolutely is NOT incompatible with
>> profiting from selling software, NOT EVEN if you restrict your business
>> model to selling copies.
>
> Companies making monet on GPL usually do it by selling support
Correction: usually have at least *one* revenue stream be selling
support. It need *not* be the *only* one.
> The GPL license does not prohibit it, but the terms of GPL plus basic
> economics do that you need something other than selling copies to
> prosper.
If that were true, then Aquafina would have to sell more than just
bottled water to prosper.
ClassCastException
> On 29-05-2010 04:17, ClassCastException wrote:
>> On Fri, 28 May 2010 20:31:14 -0400, Arne Vajhøj wrote:
>>> If everybody could copy software exactly as they wanted then I am
>>> pretty sure that the software industry would be in a very poor shape.
>>
>> I actually doubt this; I think it would work rather differently from
>> now in some ways, but that people would have found a way to make it
>> work and to profit in it. Plenty of businesses profit from open source
>> software in various ways, including by selling support or simply by
>> funding development of open source software that they use in-house and
>> get productivity gains from, and, by funding it, get more influence to
>> have features they'd find useful added and the bugs that particularly
>> harm their productivity prioritized.
>
> Since open source requires copyright then no copyright means no open
> source.
>
> So the existence of open source does not prove that copyright is not
> necessary.
I don't think so. Open source with the BSD or Apache license has so
little restriction on distribution that if all software code was public
domain, the BSD license and Apache license using development/business
models would pretty much be unaffected.
If copyright disappeared tomorrow, the whole software industry would
eventually reorganize along the lines of such projects.
Note that the world's premiere web server software is one of those
projects. So I doubt we would want for software or software development.
>>> Sounds like blogs from teenagers that wants to be able to download
>>> everything for free and have parents to pay the bills.
>>
>> I have my doubts whether Against Monopoly (run by a pair of degreed
>> economists) and Techdirt (run by a successful dot-com entrepreneur) are
>> "from teenagers that want to be able to download everything for free
>> and have parents to pay the bills". :)
>
> Given that you do not provide links or names or anything, then the
> teenager guess is as good as any.
I noticed a rather abrupt, borderline-rude tone in several of your recent
posts to this thread. And now this. It boggles the mind. I gave two
specific names, a quick Google of either leads to links, and I posted a
link to one of them at the very start of this thread, and yet you accuse
me of providing neither -- *while quoting the names*?!
What the hell is the *matter* with you? I get the distinct feeling that
you might be arguing purely for the sake of arguing. Or that any
questioning of copyright sends you into a violent, negative, emotional
response so powerful as to render you incoherent and reduce your capacity
for either reason or restraint (and thus diplomacy).
If the latter, you have some issues that need working out. It's not like
it's a matter of life or death or something!
Now, since you seem to need some real hand-holding here, I'll post the
links:
http://www.againstmonopoly.org/
http://www.techdirt.com/
HTH.
> There may be many different ways to use copyright.
>
> But copyright works identical for all - open source or closed source.
>
> In fact open source are some of the fiercest enforcer of copyright,
> because:
> * they need it even more when they hand out source code
> * a large fraction of open source people do not believe in
> patents, so they only have copyright
Most open source licenses do not actually enforce anything crucial to the
business model, so to the ability to profit from the software
development. Arguably even the GPL doesn't, but particularly most of the
non-"copyleft" ones just ask for attribution and maybe one or two other
minor things. Trademark law and laws against misrepresentation and fraud
could probably be used to cover such needs, and they can probably be done
without.
ClassCastException
> [copyright] is required or it would be almost impossible to make money
> writing books or programs. Custom programs would survive, but no one-
> size-fits-all.
That is the popular misconception, yes, but there's mounting evidence
that it simply isn't true.
For one thing, authors made money long before there was copyright.
There's only been copyright for about 300 years, but people have been
writing (and being paid to write) for thousands.
Lots of software development, including "one-size-fits-all", occurs under
the aegis of very liberal licenses like BSD and Apache. This development
includes a fair bit that's for-profit and not one dime of their revenue
came from copyright enforcement. Still more such development is non-
profit in character. The Linux kernel itself was written as a hobby, for
example. Copyright was not needed to incentivize this.
The question about copyright now, realistically, is not whether it is
*needed* or *nothing* will get created, but what amount will maximize
creation and access to the results. It's becoming clear that that amount
is much lower than what we currently have -- perhaps two or so years with
broad fair-use exemptions, enough for books, movies, and new software
versions to make the bulk of the money they will ever make -- and quite
possibly zero.
Thomas Pornin
> It is perfectly legal to put code in public domain.
It depends on the country. Under French law, the only legal way to put
code in public domain is to die, then wait for 70 years.
--Thomas Pornin
Arne Vajhøj
I don't know french law.
Some software are put in public domain.
Example:
http://www.sqlite.org/copyright.html
Arne
David Lamb
> Anyone who clings to a theory despite a growing weight of contrary
> evidence has ceased to practice science and has become a pseudoscientific
> quack at best and a religious nut at worst.
I know nothing about the particular studies in question, but a non-quack
could object to the methodology used to carry out the study or the
assumptions on which it was based. Not really an objection; just
feeling contrarian today.
David Lamb
> I do not see that there is any way for an author to put his work into
> the public domain. He might write "this work is in the public domain"
> on the front page of course but it is not clear that this holds any
> legal weight at all and indeed my own country says you still keep
> certain inalienable copyright privileges over it.
What country is that?
It has been a long time since I looked at anything even semi-official,
but once upon a time, under the national laws of the USA and Canada, the
originator could certainly declare "this work is in the public domain."
I don't think we used to have any inalienable privileges, and would be
delighted to find out we now do.
Bent C Dalager
> On 31/05/2010 5:55 AM, Bent C Dalager wrote:
>> I do not see that there is any way for an author to put his work into
>> the public domain. He might write "this work is in the public domain"
>> on the front page of course but it is not clear that this holds any
>> legal weight at all and indeed my own country says you still keep
>> certain inalienable copyright privileges over it.
>
> What country is that?
Norway. The English (non-authoritative) version of the law is here:
http://www.regjeringen.no/upload/KKD/Medier/Acts%20and%20regulations/Aandsverkloven_engelsk_versjon_nov2008.pdf
in particular para 3:
(...)
"The author may not waive his rights under the first and second
paragraphs, unless the use of the work in question is limited in
nature and extent."
"(...) This right may not be waived by the author."
also para 38c:
(...)
"The right under the first paragraph can be neither waived not
transferred. (...)"
and para 39e:
(...)
"Any agreement that to the detriment of the author significantly
deviates from the provisions of the second, sixth and seventh
paragraph cannot be applied. The author may not waive his right
pursuant to the fifth paragraph."
(...)
There may be other examples, these are the ones I could easily find by
searching on "waive" in the document. Not all of the above may apply
to software, I didn't check. Some specifics of software are treated in
paras 39g,h,i.
Arne Vajhøj
There are slightly different rules for copyright around the world,
but the fact is that public domain software is almost universal
de facto recognized.
>> Open source chose not to do so.
>>
>> Because they want to have some license conditions.
>
> I believe it is rather that they realize they /must/ write a license
> because the law effectively requires it of them and then they tend to
> tag on a pet provision or two after the "use as you wish" bit of the
> open source license.
They could have chose public domain.
Some code writer do choose that.
Some choose an open source license.
>> No copyright => no license conditions => no open source.
>
> Why is it that software which has the most generous license term
> possible under the law (that is, no conditions), cannot be "open
> source"? Surely if it were /possible/ to release source code into the
> public domain then this would be just as open as an open source
> license is?
Because the open source license gives some rights.
>>> It does help, of course, that copyright is such a huge and
>>> indiscriminate sledgehammer you can use it to kill just about
>>> /anything/, including copyright. Open source will tend to use the best
>>> tool for the job even if that tool is copyright.
>>
>> Open source does not kill copyright. They just use copyright for
>> a different purpose than MS, IBM, Oracle etc. (or should I say for
>> those companies close source offerings - they do have open source
>> offerings as well).
>
> Open source developers certainly have a varied set of motivations for
> their chose of license; some would prefer to have their license
> destroy copyright while others believe copyright is essential for them
> to earn a living.
>
>>> Neither patents nor copyright are needed for software development to
>>> thrive so it seems to me your conclusion is based on false premises.
>>
>> Neither traditional closed source nor open source would be possible
>> without copyright.
>
> But of course they would: closed source would be protected as a trade
> secret, which is effectively what is happening already anyway, and any
> source that anyone actually distributed would be open source by
> default.
Would it be a material change to protect closed source by
trade secret instead of copyright?
>> I am not aware of any company making a living from public domain
>> software.
>>
>> Based on that I am very skeptical about software industry in a world
>> with no copyright.
>
> While the estimates vary, around 80% of the software industry consists
> of various in-house projects: Customer tracking, inventory management,
> project reporting, niche expert systems, etc. This 80% of the industry
> would benefit greatly from the absence of copyright since they don't
> need it anyway: their software is only applicable to, distributed
> within, and used by their own organisation. The absence of copyright
> would make their job much easier - and less expensive - by not forcing
> them to reinvent the same wheel that hundreds of others have already
> designed to death time and time again before them.
The problem is that even though a lot of software is only used
inhouse, then most companies do not want to reject the possibility
to at some later point in time sell the software to external
customers.
Otherwise corporations would rarely have a problem with GPL!
Arne
Arne Vajhøj
FireFox and Thunderbird uses SQLite.
The authors of SQLite has released it as public domain.
If such a release is not valid in Norway, then all users
of FF and TB in Norway are breaking copyright.
I think a lot of IT people would be willing to testify about
"industry practice", "common sense", "the intention of the law"
etc. if that came to a trial.
Arne
Arne Vajhøj
> On Sun, 30 May 2010 19:59:36 -0700, Roedy Green wrote:
>> [copyright] is required or it would be almost impossible to make money
>> writing books or programs. Custom programs would survive, but no one-
>> size-fits-all.
>
> That is the popular misconception, yes, but there's mounting evidence
> that it simply isn't true.
>
> For one thing, authors made money long before there was copyright.
> There's only been copyright for about 300 years, but people have been
> writing (and being paid to write) for thousands.
Back when copying meant writing the same by hand, then the cost
of copying created a natural barrier.
> Lots of software development, including "one-size-fits-all", occurs under
> the aegis of very liberal licenses like BSD and Apache. This development
> includes a fair bit that's for-profit and not one dime of their revenue
> came from copyright enforcement. Still more such development is non-
> profit in character. The Linux kernel itself was written as a hobby, for
> example. Copyright was not needed to incentivize this.
But a lot of the people working on these projects are working for
companies that makes a huge part and in some cases the majority
of their money on software licenses.
> The question about copyright now, realistically, is not whether it is
> *needed* or *nothing* will get created, but what amount will maximize
> creation and access to the results. It's becoming clear that that amount
> is much lower than what we currently have -- perhaps two or so years with
> broad fair-use exemptions, enough for books, movies, and new software
> versions to make the bulk of the money they will ever make -- and quite
> possibly zero.
It is not particular clear.
This is just statements without any facts.
Arne
Arne Vajhøj
> On Sat, 29 May 2010 21:47:50 -0400, Arne Vajhøj wrote:
>
>> On 29-05-2010 03:56, ClassCastException wrote:
>>> This might have something to do with the fact that Aquafina makes quite
>>> a bit of money bottling and selling stuff that pretty much all of their
>>> potential customers can get out of a faucet for free.
>>
>> Not a good comparison.
>
> Of course it is. It's the *classic* example of successfully competing
> with free!
>
>>> It might also have something to do with the fact that the entertainment
>>> industry is not, contrary to popular belief, losing revenues to piracy.
>>> Declining sales of things like DVDs and recorded music have a complex
>>> web of causes, in which the effects of online piracy are not reliably
>>> different from zero according to the statistical studies.
>>
>> That "fact" is very disputed.
>
> By idiots. The studies' results are what they are, even if many refuse to
> accept them.
>
> Anyone who clings to a theory despite a growing weight of contrary
> evidence has ceased to practice science and has become a pseudoscientific
> quack at best and a religious nut at worst.
Bullshit.
All the serious economic science expect there to be some effect
but consider it very difficult to estimate it.
For a summary of where science stands read:
http://www.gao.gov/new.items/d10423.pdf
>>> The moral of the story: the GPL absolutely is NOT incompatible with
>>> profiting from selling software, NOT EVEN if you restrict your business
>>> model to selling copies.
>>
>> Companies making monet on GPL usually do it by selling support
>
> Correction: usually have at least *one* revenue stream be selling
> support. It need *not* be the *only* one.
No. But it is by far the largest for most of the wellknown open
source companies.
>> The GPL license does not prohibit it, but the terms of GPL plus basic
>> economics do that you need something other than selling copies to
>> prosper.
>
> If that were true, then Aquafina would have to sell more than just
> bottled water to prosper.
You can probably sell GPL software for the same price as a bottle
of bottled water.
But that will not make the software industry prosper.
Arne
Arne Vajhøj
> On Sat, 29 May 2010 22:10:56 -0400, Arne Vajhøj wrote:
>> On 29-05-2010 04:17, ClassCastException wrote:
>>> On Fri, 28 May 2010 20:31:14 -0400, Arne Vajhøj wrote:
>>>> If everybody could copy software exactly as they wanted then I am
>>>> pretty sure that the software industry would be in a very poor shape.
>>>
>>> I actually doubt this; I think it would work rather differently from
>>> now in some ways, but that people would have found a way to make it
>>> work and to profit in it. Plenty of businesses profit from open source
>>> software in various ways, including by selling support or simply by
>>> funding development of open source software that they use in-house and
>>> get productivity gains from, and, by funding it, get more influence to
>>> have features they'd find useful added and the bugs that particularly
>>> harm their productivity prioritized.
>>
>> Since open source requires copyright then no copyright means no open
>> source.
>>
>> So the existence of open source does not prove that copyright is not
>> necessary.
>
> I don't think so. Open source with the BSD or Apache license has so
> little restriction on distribution that if all software code was public
> domain, the BSD license and Apache license using development/business
> models would pretty much be unaffected.
They seem to think otherwise, because they chose their licenses
over public domain.
> If copyright disappeared tomorrow, the whole software industry would
> eventually reorganize along the lines of such projects.
Unlikely.
Apache is getting a lot of money from IBM.
IBM making less money => less resources to Apache, Eclipse etc..
>>>> Sounds like blogs from teenagers that wants to be able to download
>>>> everything for free and have parents to pay the bills.
>>>
>>> I have my doubts whether Against Monopoly (run by a pair of degreed
>>> economists) and Techdirt (run by a successful dot-com entrepreneur) are
>>> "from teenagers that want to be able to download everything for free
>>> and have parents to pay the bills". :)
>>
>> Given that you do not provide links or names or anything, then the
>> teenager guess is as good as any.
> Now, since you seem to need some real hand-holding here, I'll post the
The top story right now is a speech by a biochemist and
something about T-shirts respectively.
Not exactly convincing as the place for serious Economic
science.
>> There may be many different ways to use copyright.
>>
>> But copyright works identical for all - open source or closed source.
>>
>> In fact open source are some of the fiercest enforcer of copyright,
>> because:
>> * they need it even more when they hand out source code
>> * a large fraction of open source people do not believe in
>> patents, so they only have copyright
>
> Most open source licenses do not actually enforce anything crucial to the
> business model, so to the ability to profit from the software
> development. Arguably even the GPL doesn't, but particularly most of the
> non-"copyleft" ones just ask for attribution and maybe one or two other
> minor things. Trademark law and laws against misrepresentation and fraud
> could probably be used to cover such needs, and they can probably be done
> without.
Apparently they believe otherwise, because they chose their
licenses.
Arne
Arved Sandstrom
You can make a reasonable argument that anyone who has put a statement
into their documentation disavowing copyright and explicitly releasing
it into the public domain (or so they believe) has in fact retained
copyright but effectively told users that they can do with the software
whatever they wish. This latter approach is one of the recommendations,
either that or using one of the more permissive licenses, for
accomplishing the same ends as placing something into the public domain
(which latter is difficult or impossible to do, and may be
counter-productive).
I'm no software lawyer but my reading on this subject tells me that
trying to disclaim copyright and put software into the public domain is
problematic. I recommend perusing
http://linuxmafia.com/faq/Licensing_and_Law/public-domain.html for
staters, but there is a lot of easily located discussion on this
subject...none of which purports to say that placing software into the
public domain is easy to do.
My own opinion on Unlicenses and the SQLite public domain declaration:
you can put the statements in to your heart's content. It just probably
won't have legal effect. Even the SQLite team recognizes that when they
include this as a reason for obtaining an SQLite license: "You are using
SQLite in a jurisdiction that does not recognize the right of an author
to dedicate their work to the public domain."
To sum up, my reading is that if an author or authors have stated that
they wish to disclaim copyright and place software into the public
domain, odds are very good that they haven't legally done so. But I
expect a reasonable legal case could be made that in such cases, while
the copyright is still held by the author(s), that they have in effect
granted users the permission to do whatever they like with the software.
AHS
--
The most amazing achievement of the computer software industry is its
continuing cancellation of the steady and staggering gains made by the
computer hardware industry.
-- Henry Petroski
Bent C Dalager
>
> FireFox and Thunderbird uses SQLite.
>
> The authors of SQLite has released it as public domain.
>
> If such a release is not valid in Norway, then all users
> of FF and TB in Norway are breaking copyright.
Even though the copyright holders have claimed the work to be in the
public domain I doubt it actually /is/ because there is no provision
in the law for that particular option. Of course, if SQLite is itself
from a nation other than Norway then things can become complicated and
it will be difficult to tell.
In practical terms I expect that this will be considered by the courts
to be a maximally permissive license, which still leaves a few aspects
of copyright in place - and is different from public domain.
Cheers,
Bent D.
Bent C Dalager
>>
>> The only apparent way to mitigate the impact of copyright over one's
>> own work is to write very generous license terms for it. As you point
>> out, this means that your work is still protected under copyright
>> however generous you are. This is as it must be under the current
>> legal regime.
>
> There are slightly different rules for copyright around the world,
> but the fact is that public domain software is almost universal
> de facto recognized.
I don't see that this is the case. What I do see is that there exist a
number of people who would /like/ public domain software to be
achievable, and there are some who act as if it is. You might say that
these people, at least, have de facto recognized public domain softare
but what is the benefit of this? Copyright is governed by copyright
law and I cannot see that the law recognizes the public domain in any
useful fashion when it comes to software (primarily because there is
no significant body of software from before 1930-ish).
>> I believe it is rather that they realize they /must/ write a license
>> because the law effectively requires it of them and then they tend to
>> tag on a pet provision or two after the "use as you wish" bit of the
>> open source license.
>
> They could have chose public domain.
Well, that is basically the point under debate: It's not clear that
one "can" choose public domain in this context. The only thing one can
choose is how to license one's copyrighted work. If you, as an author,
"choose" public domain then I believe that the best case outcome is
that the judge gives you some leeway for not having any legal
education and maps your nonsensical (in the legal sense) choise over
into a very permissive license governed by copyright.
>> Why is it that software which has the most generous license term
>> possible under the law (that is, no conditions), cannot be "open
>> source"? Surely if it were /possible/ to release source code into the
>> public domain then this would be just as open as an open source
>> license is?
>
> Because the open source license gives some rights.
What rights are these, and who benefits from them?
>> But of course they would: closed source would be protected as a trade
>> secret, which is effectively what is happening already anyway, and any
>> source that anyone actually distributed would be open source by
>> default.
>
> Would it be a material change to protect closed source by
> trade secret instead of copyright?
The material change would be that any source code you could get your
hands on would be yours to use as you wish, unless of course you
obtained it through industrial espionage.
>> While the estimates vary, around 80% of the software industry consists
>> of various in-house projects: Customer tracking, inventory management,
>> project reporting, niche expert systems, etc. This 80% of the industry
>> would benefit greatly from the absence of copyright since they don't
>> need it anyway: their software is only applicable to, distributed
>> within, and used by their own organisation. The absence of copyright
>> would make their job much easier - and less expensive - by not forcing
>> them to reinvent the same wheel that hundreds of others have already
>> designed to death time and time again before them.
>
> The problem is that even though a lot of software is only used
> inhouse, then most companies do not want to reject the possibility
> to at some later point in time sell the software to external
> customers.
To the extent that this possibility appears as an asset at all, if it
is given a realistic value it must be very very small indeed. From
what I have seen, in-house software projects tend to be completely
unsuitable for distribution to any other than the primary client and
would take gargantuan effort to change into a sellable product.
I expect the cost savings from being able to use any and all other
software without having to worry about licensing costs and liabilities
would by far dwarf whatever miniscule economic opportunity exists in
one's own in-house projects.
> Otherwise corporations would rarely have a problem with GPL!
I expect the GPL resistance has very much to do with some fundamental
misunderstandings about how the GPL actually works. Surprisingly many
people seem to think that if they develop a GPL application then they
are automatically required to distribute it to the world and provide
support for it to anyone who asks. An accountant who actually believes
this will quickly point out that it effectively amounts to unlimited
liability and advise strongly against going anywhere near such a
license.
Lew
> In practical terms I expect that this will be considered by the courts
> to be a maximally permissive license, which still leaves a few aspects
> of copyright in place - and is different from public domain.
From the user's perspective, a difference that makes no difference is no
difference. They can safely use the product and leave the hair-splitting to
the lawyers.
--
Lew
Bent C Dalager
Users cannot safely use the product if they end up violating any of
the inalienable rights, wrongly thinking that the work is in the
public domain.
Authors also cannot safely just write "public domain" and think it
will do what they think it will.
As an example of the latter, an author might be interested in maximum
exposure of his work and so desire to put it into the public domain in
order to make its wide dissemination legally uncomplicated. This would
seem an obvious strategy for an unknown author trying to make a name
for himself. As it is, however, the work retains its copyright and so
any institution that might not think twice about distributing public
domain works would need to evaluate the legal implications of doing so
with the copyrighted one. If the work simply says "this work is in the
public domain" that is likely to complicate matters even further since
it is unclear what the legal standing of this is while an actual
copyright license would be considerably more clear. In sum this would
tend to restrict the distribution of the work, which directly
contravenes the author's intentions in this case.
Cheers,
Bent D
Jeff Higgins
> In dealing with Java,
Throughout this rambling discussion there has been no citation of case,
regulatory, or statutory law from any jurisdiction.
Arne Vajhøj
> into their documentation disavowing copyright and explicitly releasing
> it into the public domain (or so they believe) has in fact retained
> copyright but effectively told users that they can do with the software
> whatever they wish. This latter approach is one of the recommendations,
> either that or using one of the more permissive licenses, for
> accomplishing the same ends as placing something into the public domain
> (which latter is difficult or impossible to do, and may be
> counter-productive).
>
> I'm no software lawyer but my reading on this subject tells me that
> trying to disclaim copyright and put software into the public domain is
> problematic. I recommend perusing
> http://linuxmafia.com/faq/Licensing_and_Law/public-domain.html for
> staters, but there is a lot of easily located discussion on this
> subject...none of which purports to say that placing software into the
> public domain is easy to do.
They lawyers may find it very problematic.
But there does not seem to be any problems with software
where the authors put code (or try to put code) in public
domain.
> My own opinion on Unlicenses and the SQLite public domain declaration:
> you can put the statements in to your heart's content. It just probably
> won't have legal effect. Even the SQLite team recognizes that when they
> include this as a reason for obtaining an SQLite license: "You are using
> SQLite in a jurisdiction that does not recognize the right of an author
> to dedicate their work to the public domain."
>
> To sum up, my reading is that if an author or authors have stated that
> they wish to disclaim copyright and place software into the public
> domain, odds are very good that they haven't legally done so. But I
> expect a reasonable legal case could be made that in such cases, while
> the copyright is still held by the author(s), that they have in effect
> granted users the permission to do whatever they like with the software.
Could very well be that would be the judges approach.
If there are no legal possibilities for public domain in that
jurisdiction, then it seems as what best fits what both the author
and the users want.
Arne
Arne Vajhøj
> On 2010-06-21, Arne Vajhøj<ar...@vajhoej.dk> wrote:
>>
>> FireFox and Thunderbird uses SQLite.
>>
>> The authors of SQLite has released it as public domain.
>>
>> If such a release is not valid in Norway, then all users
>> of FF and TB in Norway are breaking copyright.
>
> Even though the copyright holders have claimed the work to be in the
> public domain I doubt it actually /is/ because there is no provision
> in the law for that particular option. Of course, if SQLite is itself
> from a nation other than Norway then things can become complicated and
> it will be difficult to tell.
>
> In practical terms I expect that this will be considered by the courts
> to be a maximally permissive license, which still leaves a few aspects
> of copyright in place - and is different from public domain.
As I replied to Arved:
Arne Vajhøj
That is a theory.
But does SQLite distribution seem to have been restricted
due to being declared public domain?
Arne
Arne Vajhøj
> On 2010-06-21, Arne Vajh�j<ar...@vajhoej.dk> wrote:
>> On 31-05-2010 05:55, Bent C Dalager wrote:
>>> The only apparent way to mitigate the impact of copyright over one's
>>> own work is to write very generous license terms for it. As you point
>>> out, this means that your work is still protected under copyright
>>> however generous you are. This is as it must be under the current
>>> legal regime.
>>
>> There are slightly different rules for copyright around the world,
>> but the fact is that public domain software is almost universal
>> de facto recognized.
>
> I don't see that this is the case. What I do see is that there exist a
> number of people who would /like/ public domain software to be
> achievable, and there are some who act as if it is. You might say that
> these people, at least, have de facto recognized public domain softare
> but what is the benefit of this? Copyright is governed by copyright
> law and I cannot see that the law recognizes the public domain in any
> useful fashion when it comes to software (primarily because there is
> no significant body of software from before 1930-ish).
Public domain software is very recognized.
Software by US government employees have always been public domain
and recognized as such.
>>> I believe it is rather that they realize they /must/ write a license
>>> because the law effectively requires it of them and then they tend to
>>> tag on a pet provision or two after the "use as you wish" bit of the
>>> open source license.
>>
>> They could have chose public domain.
>
> Well, that is basically the point under debate: It's not clear that
> one "can" choose public domain in this context.
Some people are choosing public domain.
These people could certainly have done the same.
> The only thing one can
> choose is how to license one's copyrighted work. If you, as an author,
> "choose" public domain then I believe that the best case outcome is
> that the judge gives you some leeway for not having any legal
> education and maps your nonsensical (in the legal sense) choise over
> into a very permissive license governed by copyright.
Most of the legal issues are centered about the fact that the
law in many countries does not mention such a concept.
But unless the law explicitly prohibits disclaiming copyright
(which is also does in some countries), then I would expect
it to be possible - the fact that something is not explicit
approved in the law should not mean that it is prohibited.
>>> Why is it that software which has the most generous license term
>>> possible under the law (that is, no conditions), cannot be "open
>>> source"? Surely if it were /possible/ to release source code into the
>>> public domain then this would be just as open as an open source
>>> license is?
>>
>> Because the open source license gives some rights.
>
> What rights are these, and who benefits from them?
It varies between licenses.
But the general purpose is to give the users certain rights.
(in a few cases also certain obligations)
>>> But of course they would: closed source would be protected as a trade
>>> secret, which is effectively what is happening already anyway, and any
>>> source that anyone actually distributed would be open source by
>>> default.
>>
>> Would it be a material change to protect closed source by
>> trade secret instead of copyright?
>
> The material change would be that any source code you could get your
> hands on would be yours to use as you wish, unless of course you
> obtained it through industrial espionage.
If it is sufficient that you did not do the industrial espionage
yourself to be allowed to use the code, then I think industrial
espionage would be a very profitable business.
If it need to be clean all the way, then it would be very
challenging to use software, because without any license
information how do you know where the code came from.
>>> While the estimates vary, around 80% of the software industry consists
>>> of various in-house projects: Customer tracking, inventory management,
>>> project reporting, niche expert systems, etc. This 80% of the industry
>>> would benefit greatly from the absence of copyright since they don't
>>> need it anyway: their software is only applicable to, distributed
>>> within, and used by their own organisation. The absence of copyright
>>> would make their job much easier - and less expensive - by not forcing
>>> them to reinvent the same wheel that hundreds of others have already
>>> designed to death time and time again before them.
>>
>> The problem is that even though a lot of software is only used
>> inhouse, then most companies do not want to reject the possibility
>> to at some later point in time sell the software to external
>> customers.
>
> To the extent that this possibility appears as an asset at all, if it
> is given a realistic value it must be very very small indeed. From
> what I have seen, in-house software projects tend to be completely
> unsuitable for distribution to any other than the primary client and
> would take gargantuan effort to change into a sellable product.
Often it is.
But the companies does not think so.
It is like driving - some survey showed that 90% of people believe that
they drive better than average.
>> Otherwise corporations would rarely have a problem with GPL!
>
> I expect the GPL resistance has very much to do with some fundamental
> misunderstandings about how the GPL actually works. Surprisingly many
> people seem to think that if they develop a GPL application then they
> are automatically required to distribute it to the world and provide
> support for it to anyone who asks. An accountant who actually believes
> this will quickly point out that it effectively amounts to unlimited
> liability and advise strongly against going anywhere near such a
> license.
I think you are underestimating the corporate lawyers.
Arne
Bent C Dalager
>
> That is a theory.
>
> But does SQLite distribution seem to have been restricted
> due to being declared public domain?
This is impossible to say of course.
In an attempt to unravel the latest series of posts: my point is that
open source projects that choose a copyright license rather than
"public domain" may very well have done so because they are familiar
with, and care about, the legal situation surrounding those choices.
They may not particularly /want/ to own a copyright but they realize
that as a point of legal reality they are forced to do so, and they
make the best out of a bad situation by choosing some very permissive
license. Therefore, the decision to have a license is not in itself an
indication that a project team feels it needs or even desires
copyright protection of their work.
Others may choose to pretend that the public domain is in fact a valid
choice for a work newer than 75 years (or so) but an examination of
the legal climate surrounding the issue would seem to indicate that
this is basically wishful thinking.
Andreas Leitgeb
> They lawyers may find it very problematic.
IANAL, but sometimes playing devil's advocate:
A company could write some useful code, and donate it to
public domain. Later, when others have started using it,
the company could be taken over, and the new owners may
point out, that that previous act of pd-izing was illegal
from the beginning and furtheron demand license-fees ...
Bent C Dalager
>>
>> I don't see that this is the case. What I do see is that there exist a
>> number of people who would /like/ public domain software to be
>> achievable, and there are some who act as if it is. You might say that
>> these people, at least, have de facto recognized public domain softare
>> but what is the benefit of this? Copyright is governed by copyright
>> law and I cannot see that the law recognizes the public domain in any
>> useful fashion when it comes to software (primarily because there is
>> no significant body of software from before 1930-ish).
>
> Public domain software is very recognized.
If you mean that the existence of the public domain is recognized,
then certainly. What does not appear to be recognized is the power of
an author to place his own works into the public domain, other than by
dying then waiting 75 years (or whatever).
> Software by US government employees have always been public domain
> and recognized as such.
Yes, I believe there is specific provision for this in US law. Can you
point to similar provisions for the public domain for privately
developed software?
>>
>> Well, that is basically the point under debate: It's not clear that
>> one "can" choose public domain in this context.
>
> Some people are choosing public domain.
>
> These people could certainly have done the same.
By all appearances, choosing the public domain makes about as much
sense as choosing to put your software in blue. That is, it makes no
sense at all. Which, of course, doesn't actually in itself /prevent/
people from saying "I place my software in blue".
> Most of the legal issues are centered about the fact that the
> law in many countries does not mention such a concept.
>
> But unless the law explicitly prohibits disclaiming copyright
> (which is also does in some countries), then I would expect
> it to be possible - the fact that something is not explicit
> approved in the law should not mean that it is prohibited.
A central point in modern copyright law is that all works are
copyrighted by default. In order to opt out of such a scheme there
would need to be provision in the law for an author to do so. There
wouldn't appear to be any.
>>>> Why is it that software which has the most generous license term
>>>> possible under the law (that is, no conditions), cannot be "open
>>>> source"? Surely if it were /possible/ to release source code into the
>>>> public domain then this would be just as open as an open source
>>>> license is?
>>>
>>> Because the open source license gives some rights.
>>
>> What rights are these, and who benefits from them?
>
> It varies between licenses.
>
> But the general purpose is to give the users certain rights.
Can you give an example of one such right that would not also
automatically be available were the software to be in the public
domain?
>> The material change would be that any source code you could get your
>> hands on would be yours to use as you wish, unless of course you
>> obtained it through industrial espionage.
>
> If it is sufficient that you did not do the industrial espionage
> yourself to be allowed to use the code, then I think industrial
> espionage would be a very profitable business.
I do believe that it is, yes.
Trade secret laws vary widely between jurisdictions so the extent to
which a third party not involved in the actual espionage is affected
will vary.
> If it need to be clean all the way, then it would be very
> challenging to use software, because without any license
> information how do you know where the code came from.
This is little different from the current situation: if code has no
license or copyright information you do not know its legal status and
so you cannot use it.
Chances are that under a pure trade secrets regime you would have an
easier time of it since unless you yourself were party to espionage
you will be seen to have acted in good faith. Conversely, under the
current default copyright regime you should know that all code is
always copyrighted except under rather special circumstances and so it
is much harder to claim good faith.
>> To the extent that this possibility appears as an asset at all, if it
>> is given a realistic value it must be very very small indeed. From
>> what I have seen, in-house software projects tend to be completely
>> unsuitable for distribution to any other than the primary client and
>> would take gargantuan effort to change into a sellable product.
>
> Often it is.
>
> But the companies does not think so.
If this is the case, then removing copyright would be doing them a
favour.
>> I expect the GPL resistance has very much to do with some fundamental
>> misunderstandings about how the GPL actually works. Surprisingly many
>> people seem to think that if they develop a GPL application then they
>> are automatically required to distribute it to the world and provide
>> support for it to anyone who asks. An accountant who actually believes
>> this will quickly point out that it effectively amounts to unlimited
>> liability and advise strongly against going anywhere near such a
>> license.
>
> I think you are underestimating the corporate lawyers.
In my experience lawyers often do not get involved in matters of
specific open source licenses: The idea gets killed long before a
lawyer is involved.
In the cases were lawyers actually /are/ involved on a per license
basis I would expect GPL to see much higher adoption as many of the
popular myths are likely to be dispelled by an educated reading of the
license.
One question it appears many corporate lawyers are asked is along the
lines of "can we use open source licensed software" to which a lawyer
is basically forced to answer "no" because in general you cannot: you
need to examine each individual open source license on its own merits
and cannot make a blanket statement about the set of all possible such
licenses except to reject the idea. Hence, perhaps, the "no open
source" policy in many companies.
Lew
>> They lawyers may find it very problematic.
Andreas Leitgeb wrote:
> IANAL, but sometimes playing devil's advocate:
>
> A company could write some useful code, and donate it to
> public domain. Later, when others have started using it,
> the company could be taken over, and the new owners may
> point out, that that previous act of pd-izing was illegal
> from the beginning and furtheron demand license-fees ...
They would lose. Adverse possession.
--
Lew
Andreas Leitgeb
Your word in ${user.preferred.deity}'s ear. :-)
John B. Matthews
Andreas Leitgeb <a...@gamma.logic.tuwien.ac.at> wrote:
Just in: "Terrible News: Court Says It's Okay To Remove Content From The
Public Domain And Put It Back Under Copyright."
<http://www.techdirt.com/articles/20100621/2320049908.shtml>
Discussion: "Court Takes Away Some of the Public Domain."
<http://yro.slashdot.org/story/10/06/22/1724254/>
--
John B. Matthews
trashgod at gmail dot com
<http://sites.google.com/site/drjohnbmatthews>
ClassCastException
> On 31-05-2010 06:37, ClassCastException wrote:
>> Anyone who clings to a theory despite a growing weight of contrary
>> evidence has ceased to practice science and has become a
>> pseudoscientific quack at best and a religious nut at worst.
>
> Bullshit.
So, climate change deniers, creationists, etc. are not pseudoscientific
quacks and religious nuts?
That's news to me.
> All the serious economic science expect there to be some effect but
> consider it very difficult to estimate it.
There have been studies, by e.g. Boldrin and Levine. The measured effects
are negative, i.e. patents actually make things worse.
>>>> The moral of the story: the GPL absolutely is NOT incompatible with
>>>> profiting from selling software, NOT EVEN if you restrict your
>>>> business model to selling copies.
>>>
>>> Companies making monet on GPL usually do it by selling support
>>
>> Correction: usually have at least *one* revenue stream be selling
>> support. It need *not* be the *only* one.
>
> No. But it is by far the largest for most of the wellknown open source
> companies.
Define "wellknown open source companies". If it's only companies like Red
Hat whose sole product is software, you're probably right; widen it to
include companies like IBM that are involved heavily with open source but
do other things as well ...
>>> The GPL license does not prohibit it, but the terms of GPL plus basic
>>> economics do that you need something other than selling copies to
>>> prosper.
>>
>> If that were true, then Aquafina would have to sell more than just
>> bottled water to prosper.
>
> You can probably sell GPL software for the same price as a bottle of
> bottled water.
>
> But that will not make the software industry prosper.
Define "prosper"? Thinner margins doesn't equate to losses. So if by
"prosper" you mean companies like Microsoft laughing all the way to the
bank and pocketing billions, making people like Bill Gates the richest on
the planet, then no, probably not. If by "prosper" you mean "positive
growth, however small", then it's another story.
In practice, though, I would expect selling support to become more
lucrative than selling copies.
ClassCastException
> On 31-05-2010 06:46, ClassCastException wrote:
>> On Sat, 29 May 2010 22:10:56 -0400, Arne Vajhøj wrote:
>>> So the existence of open source does not prove that copyright is not
>>> necessary.
>>
>> I don't think so. Open source with the BSD or Apache license has so
>> little restriction on distribution that if all software code was public
>> domain, the BSD license and Apache license using development/business
>> models would pretty much be unaffected.
>
> They seem to think otherwise, because they chose their licenses over
> public domain.
As Bent pointed out elsethread, it's not apparently a sure bet to
dedicate a work to the public domain. Legally astute open source
developers may choose super-permissive licenses as a legally-sure-bet
alternative that carries most of the same benefits.
>> If copyright disappeared tomorrow, the whole software industry would
>> eventually reorganize along the lines of such projects.
>
> Unlikely.
>
> Apache is getting a lot of money from IBM.
>
> IBM making less money => less resources to Apache, Eclipse etc..
Why would IBM be making less money? Selling copies of Apache and Eclipse
is not IBM's primary revenue stream. The last time I checked, selling
hardware was, or was one of them.
>>> Given that you do not provide links or names or anything, then the
>>> teenager guess is as good as any.
>
>> Now, since you seem to need some real hand-holding here, I'll post the
>> links:
>>
>> http://www.againstmonopoly.org/
>> http://www.techdirt.com/
>
> The top story right now is a speech by a biochemist and something about
> T-shirts respectively.
Wait a few hours.
>> Most open source licenses do not actually enforce anything crucial to
>> the business model, so to the ability to profit from the software
>> development. Arguably even the GPL doesn't, but particularly most of
>> the non-"copyleft" ones just ask for attribution and maybe one or two
>> other minor things. Trademark law and laws against misrepresentation
>> and fraud could probably be used to cover such needs, and they can
>> probably be done without.
>
> Apparently they believe otherwise, because they chose their licenses.
As Bent pointed out elsethread, it's not apparently a sure bet to
dedicate a work to the public domain. Legally astute open source
developers may choose super-permissive licenses as a legally-sure-bet
alternative that carries most of the same benefits.
ClassCastException
> On 31-05-2010 06:52, ClassCastException wrote:
>> On Sun, 30 May 2010 19:59:36 -0700, Roedy Green wrote:
>>> [copyright] is required or it would be almost impossible to make money
>>> writing books or programs. Custom programs would survive, but no one-
>>> size-fits-all.
>>
>> That is the popular misconception, yes, but there's mounting evidence
>> that it simply isn't true.
>>
>> For one thing, authors made money long before there was copyright.
>> There's only been copyright for about 300 years, but people have been
>> writing (and being paid to write) for thousands.
>
> Back when copying meant writing the same by hand, then the cost of
> copying created a natural barrier.
Irrelevant. There've been somewhat-mechanized means of duplicating
written documents for 400+ years.
>> Lots of software development, including "one-size-fits-all", occurs
>> under the aegis of very liberal licenses like BSD and Apache. This
>> development includes a fair bit that's for-profit and not one dime of
>> their revenue came from copyright enforcement. Still more such
>> development is non- profit in character. The Linux kernel itself was
>> written as a hobby, for example. Copyright was not needed to
>> incentivize this.
>
> But a lot of the people working on these projects are working for
> companies that makes a huge part and in some cases the majority of their
> money on software licenses.
90% or more of programmers are paid to develop in-house code, not
software that will be sold to the general public. *None* get copyright
royalties. And the companies can find other business models, the
existence of which has certainly been proved.
>> The question about copyright now, realistically, is not whether it is
>> *needed* or *nothing* will get created, but what amount will maximize
>> creation and access to the results. It's becoming clear that that
>> amount is much lower than what we currently have -- perhaps two or so
>> years with broad fair-use exemptions, enough for books, movies, and new
>> software versions to make the bulk of the money they will ever make --
>> and quite possibly zero.
>
> It is not particular clear.
Sure it is. Commercially sold works make the bulk of their revenues in
the first year or two. The hardcover book. The blockbuster movie. Suppose
one of each is released today. Six months from now, the movie appears on
PPV and DVD; you can watch it for a fraction the theater ticket price via
PPV or rental. Six months after that, the movie appears on non-PPV pay TV
ad-free movie superchannels and the paperback book comes out. The
hardcover finds its way to the bargain bin with the price lowered to
compete with the paperback. The DVDs are already also appearing in
bargain bins. A year after that, the hardcovers are selling for two or
three bucks in the bargain bins that still have copies and the movie
shows up on free broadcast TV (albeit now interspersed with ads).
Revenue curves likewise show this. Avatar made its first billion within a
few *weeks*.
Really, the only argument you can make is that with a two-year copyright
term people would wait for it to become free rather than pay, but people
pay to see movies in the theater and buy hardback books all the time,
rather than wait for the paperback or for the movie to be on cheap DVD
rental or free over-the-air TV, so apparently that argument would be
wrong.
Furthermore, cinema tickets are actually scarce. Get rid of copyright and
their price would remain well above zero, and put in place revenue-
sharing arrangements (theater will help fund filmmaking in exchange for
being the first on their block to screen it, or whatever) the movie
industry could conceivably make rather a large fraction of its current
box-office revenues in a no-copyright world (and meanwhile its expenses
go through the floor -- no need to license music, etc.; the original Star
Wars proved you could make a blockbuster with at-the-time inexpensive no-
name talent, so the expenditures on million-dollar names and faces can
also go; and end-to-end digital production is another massive cost-
cutting option). Printed books are attractive even given e-readers; zero
out copyright and people will still buy books for the foreseeable future.
As the "pirating" of Dickens in America a century or more ago
demonstrated, an author *can* make a profit selling printed books without
a distribution monopoly. The works that would be harmed the most are
*bad* works, as in any situation where it's easy to "try before you buy".
The inability to sucker an opening weekend audience of millions into
paying to see a bad film does not strike me as a tragedy, though, were it
to become the case. It should in fact serve as an impetus to improve
quality, if it ceases to be possible to actually turn a profit on some
percentage of badly-made films.
ClassCastException
No one jurisdiction covers all software development (or even all Java
software development); there have been mentions of the present laws of
Norway and the United States, though.
Lew
> Just in: "Terrible News: Court Says It's Okay To Remove Content From The
> Public Domain And Put It Back Under Copyright."
>
> <http://www.techdirt.com/articles/20100621/2320049908.shtml>
>
> Discussion: "Court Takes Away Some of the Public Domain."
>
> <http://yro.slashdot.org/story/10/06/22/1724254/>
That article discusses Congress pulling something out of the public domain,
and doesn't address the points raised in this thread about whether the
copyright holder could have put the stuff in the public domain in the first
place, nor whether the copyright holder could have done it without the weight
of the legislature.
--
Lew
John B. Matthews
wrote:
To me it suggested how readily a motivated legislature and judiciary
might act to vitiate the applicability of adverse possession, which you
cited.
<http://groups.google.com/group/comp.lang.java.programmer/msg/8060d9d24cc32fcd>
From the recently overturned opinion: "In the United States, that body
of law includes the bedrock principle that works in the public domain
remain in the public domain."
<http://www.techdirt.com/articles/20090403/1619494384.shtml>
David
this thread.
If you want to make money from GPL software then all you have to do is
start selling it. Just like propriatary software.
If you redistribute GPL covered binaries you are expected to attribute
this and point people to the source (you don't have to host it).
If you are 'mere aggregation' then that's all you have to do. For
example - you can redistribute Apache Tomcat in a binary bundle if all
your software consists of is a set of JSP/HTML files in a WAR file.
If you modify an existing product, then you should host the source
code and make it available in some way. Again, that all you have to
do. You can even cover reasonable costs in doing this. As has already
been pointed out in this thread, the source code is usually worth very
little - it is the support and expertise a company supplies that
provided the draw.
Finally, try to ignore the hysteria. The GPL is not viral, it isn't
dangerous and it doesn't violate copyright. It is a very simple
license based on a clear understanding of copyright law. There are
numerous products using it (and similar licenses) that make plenty of
money.
David
Arved Sandstrom
no legal effect.
Bent C Dalager
> Andreas Leitgeb wrote:
>> Arne Vajhøj <ar...@vajhoej.dk> wrote:
>>> They lawyers may find it very problematic.
>>
>> IANAL, but sometimes playing devil's advocate:
>>
>> A company could write some useful code, and donate it to
>> public domain. Later, when others have started using it,
>> the company could be taken over, and the new owners may
>> point out, that that previous act of pd-izing was illegal
>> from the beginning and furtheron demand license-fees ...
>>
> Not illegal - there's no law (I think) against doing something which has
> no legal effect.
If the intent was communicated clearly enough then there may be a case
of promissory estoppel or similar. That is, the court may hold that
the original promise must be honoured in spite of it not being part of
a contractual obligation. In this way, making a promise will at times
have a very real legal effect even if the sort of promise given isn't
really backed by any very specific law.
But again you are relying on the good will of the court, which is a
capricious beast at best. The company officers who made the original
promise would probably have been much better served by phrasing it in
a manner that is actually based on an applicable law.
David Lamb
>> Would it be a material change to protect closed source by
>> trade secret instead of copyright?
>
> The material change would be that any source code you could get your
> hands on would be yours to use as you wish, unless of course you
> obtained it through industrial espionage.
What I'm fairly sure of is that a company trying to protect something
via trade secret has to keep that something actually "secret" in a
specific sense: it has to make sure that everyone who gets to see, for
example, the source code for some software, or the method for producing
a peculiar blue shade of stained glass, signs some form of
non-disclosure agreement beforehand.
If you have in hand a copy of the source code that you obtained by some
legal means, without having signed non-disclosure, I'm reasonably sure
what you said is true. However, being given a copy by someone who
obtained it illegally probably doesn't itself count as legitimate source
of the secret. If an employee who signed nondisclosure posts the source
code on some website, and that was allowed to count as a legal way of
others obtaining the trade secret -- that sounds like a hole so big as
to make trade secret a worthless concept.
Arved Sandstrom
I think that if employees sign good NDAs and non-competes that if those
employees then disclose the source code for something on a website that
this doesn't by itself void the trade secret...provided that other
measures to protect the secret are in place. That is, if you were not
shredding confidential paperwork and just left it lying around on desks,
there was no kind of policy concerning removal of paperwork or flash
drives from the premises, you had no decent escorted visitor access, and
your offices were routinely cleaned by custodians who were not vetted,
that disgruntled employee could go ahead and post all the source for
your latest in-house app - which has some trade secrets on it - and it
would possibly be legal for competitors to use that knowledge.
Bent C Dalager
>
> I think that if employees sign good NDAs and non-competes that if those
> employees then disclose the source code for something on a website that
> this doesn't by itself void the trade secret...provided that other
> measures to protect the secret are in place. That is, if you were not
> shredding confidential paperwork and just left it lying around on desks,
> there was no kind of policy concerning removal of paperwork or flash
> drives from the premises, you had no decent escorted visitor access, and
> your offices were routinely cleaned by custodians who were not vetted,
> that disgruntled employee could go ahead and post all the source for
> your latest in-house app - which has some trade secrets on it - and it
> would possibly be legal for competitors to use that knowledge.
It's difficult to say very much specific about trade secrets because
they are handled very differently in different jurisdictions. In
general, however, if a company can document that they have good
mechanisms in place for keeping a trade secret secret then this will
tend to count in their favour when determining whether the secret has
legally become public.
Conversely, if a company calls something a trade secret but then does
little or nothing to actually protect it their secret will tend to
lose legal trade secret status very fast.