Fw: Query about GNU-GPL

[Niall Douglas]

On 3 Apr 2005 at 12:30, Simo Sorce wrote:

> > You lose the money that you would have earned if someone else hadn't
> > stolen your code and sold it instead of you. I agree it's not the
> > same as perhaps you were a lousy seller and couldn't have made the
> > money anyway - but IMHO if you take risk and be step change
> > innovative, you must get money. In an ideal legal framework, you
> > would be *encouraging* people to "steal" your code because that
> > would generate additional revenue for you - but under such a system,
> > it's no longer stealing.
> 
> Niall,
> I'm quite surprised after so much time on these list you still have
> the courage to speak of stealing. You sadly remember me of some stupid
> advertise being added on DVD films that equate movies download to car
> stealing, here in Italy.

Perhaps I am not being clear enough. I am speaking of how people see 
the current system - which isn't the same as the *actual* system. 
People tend to equate second order loss as first order eg; if this 
earthquake kills my children, then I must have offended a god.

People equate the loss of what you might have got if someone hadn't 
pilfered your money making thing (ie; potential profit) with that of 
real profit. Another example is that of gambling, where what drives 
people is the belief that they beat the average and can come out 
ahead.

Now what links the concept of duplication of an idea and stealing of 
an idea is precisely this phenomenon. It exists because the copyright 
and patent law make it so. Change the law, and you change the 
association. As I mentioned in my example, under a different legal 
framework you *want* your idea "stolen" from you as it makes more 
money for you.

Hence why I state that to free software, to truly *free* software, it 
needs the law substantially changing. The GPL does not free software, 
it puts it in different chains - some better than proprietary, some 
equal and some worse. Anything based in copyright and/or software 
patents must do so to some extent.

> I remember you had interesting opinions in
> the past, it seem that time is gone, sadly. Anyway it is also
> surprising to see that you think money is everything in life. Not all
> people want money as reward (some do not even want a reward) for their
> job, your vision seem to be really distort.

There are three things which drive humans, as evidenced by the last 
few thousand years of history:

1. Sex
2. Improving understanding/Gaining knowledge
3. Capacity to affect change

All three are about enabling creativity which is the primary role of 
all biological life. The first actually has an important role in 
creation of software, but it and improving understanding are outside 
the scope of this email though ancillory to it. The third can be 
realised in many ways - ego-boosting, standing for election, peer 
recognition etc. but the most important since Norman times at least 
is increasing ones wealth.

> A _good_ legal framework would let anyone have the reward they like,
> not necessarily money, as long as it doesn't hurt other people.

A *good* legal framework would ensure that material wealth is 
distributed to those who most benefit society. Otherwise, put 
frankly, it goes to already rich people in its entirety and society 
ossifies and dies (as it has countless times in human history). The 
great success of the Occidental Empire has been the invention of rich 
people voluntarily giving creative poorer people material wealth, 
thus drawing more people into it and setting a benchmark for the 
entire of society to aim at. There is only around genetically 20% of 
creative people in the population (can vary according to analysis 
methods) and a thriving society is one which encourages them, 
especially as culture's natural tendency is to reign them in.

> > The GPL equates users with programmers as with source access all
> > users can become programmers. Hence my statement stands - under the
> > GPL, everyone donates their work if others do the same.
> 
> No, there's still a big difference, the copyright owner still have all
> the freedom, even the freedom to change the license. Users don't, they
> have freedom to use, modify and redistribute, but they can't change
> the license. So there's still a big difference. What you are speaking
> of are the BSD-like licenses they equate users and programmers
> completely.

That is an artifact of copyright law, not of the GPL itself. The GPL 
or any license cannot give or remove rights copyright law does not 
permit a license to perform - thus, any extra freedoms the copyright 
owner may have have nothing to do with the GPL.

I still hold that under the GPL, everyone donates their work if 
others do the same. Consider this - what software project do you know 
of which annotates each and every source change by each programmer 
every time into the source itself? I suppose a CVS log would do it, 
but what about admins who commit a patch written by someone else?

The reality is that no one claims copyright on their individual edits 
because it is impractical. They can no more license their own edits 
under a different licence to the GPL than build a working project 
from their own edits alone.

> > Microsoft brought a whole raft of mainframe & laboratory features to
> > the home user in a format they could operate without arcane
> > technical knowledge. If that isn't step-change innovation in a
> > nutshell, I don't know what is.
> 
> Niall, this is pure lie. Please where have you been the past 25 years?
> Microsoft didn't brought anything like "mainframe & laboratory
> features to the home user". Where have you read this lie?

Off the top of my head, who other brought the following features from 
mainframe operating systems to more than 5% of the computer using 
public first?

1: Multithreading (Preemptive multitasking too with the exception of 
the Amiga)
2: Multiple users
3: Memory protection for processes
4: Individual ACL based file security
5: Over 85% binary compatibility with the previous twenty years of 
programs for the same architecture (it may be that Apples beat this)
6: POSIX compliance (yes, NT had this over a decade ago)
7: Smoothed fonts (Acorn RISC-OS had this in the early 90's, then 
Apples I think but after Windows)
8: Remote desktops (yes X11 had this, but again didn't beat 5%. It 
may be that VNC beat Win2k)
9: Asynchronous device i/o
10: Virtual memory (Amigas might have had this, I can't remember)

Some of these will be false depending on when NT made 5% market 
penetration - I don't know when that was, I assumed 1997.

> The innovation that made it possible to bring complicated computing
> machines to the man of the street was made by others. You may claim
> Intel was one of the first that made a chip affordable and powerful
> enough to be put in home computers. You may praise Xerox Park
> laboratories for the creation of the graphical interface. You may say
> thanks to Unix programmers for having made one of the first portable
> and usable shell based system. And you may say Apple made the first
> user friendly system.

Actually I wouldn't claim the Intel point.

> But what did Microsoft?
> - Have they ever designed a piece of hardware their system run on? No.
> - Have they designed a good user interface from scratch? DOS? was that
> a usable interface? It was crappier then unix, copied down from CP/M,
> a lot less powerful and user friendly than others... Windows
> 1.0/2.0/3.0 ? The MACs were infinitely better at that time and came
> out before any version of Windows.

Look at Windows 95. By all rights, that software shouldn't never have 
worked at all. It was crap in every way possible, except that 
amazingly it did work long enough for people to get some stuff done. 
Admitted, it was nasty, but it was a damn sight better than Win3.1 
and DOS. And it contained, if nothing else, an API modelled around a 
real OS (not itself) which got into a massive market share, such that 
we thankfully are no longer stuck with writing for 16 bit x86 rubbish 
as 32 bit binaries then written run on real operating systems now 
(which they were actually written for, not 95 itself).

> So in a perfect legal system as you advocate Microsoft should have
> never growth, they should have died long ago.

I have no problem with MS stealing other people's ideas if MS pays 
that person for their idea. However, with software, it's of greater 
benefit if people could sell their /implementations/ of an idea with 
the idea itself unprotectable. Then you could get people writing 
features for Windows which MS would find cheaper to buy than develop 
itself. This leads to a collaborative development form, much like 
open software. Of course, full source would need to be supplied with 
Windows to make this possible.

[rest chopped as all those points explained by the above]

Cheers,
Niall