Open Source and Forking It

2009-10-21 Thread Peter Alcibiades
bForking It/b

Richard asks on another thread (not yet on Nabble, which quite often seems 
to run late) whether there is any way of being Open Source and not having 
forking.

No, definitely not.  As soon as you have any such restriction, you have 
left Open Source.  

And this is fundamental to the idea of Open Source.  What it seeks to do is 
give people power to do whatever modifications they want to the source 
code, as long as they contribute back what they have done.  This is why 
Ballmer made such rude remarks about it.  This is also why it is only about 
the power to do things with the code.  What it runs on or what the 
requirements for running it are, is immaterial.  

The classic example of recent forking occurred with X windows.  Over a 
period of a couple of months, Xorg forked from Xfree86, and secured 
universal adoption by all Linux distributions.  The team simply left and 
took the code with them.  Recently on a matter closer to home, we have seen 
a fork, or an attempted fork, of PythonCard.  If you look on DistroWatch, 
there are north of 350 different distributions of Linux now.  Well, these 
are all forks.  On the other hand, you notice there is only one Python.  
Basically, projects that keep in touch and are responsive do not get 
forked.  Ones that are of no interest do not get forked.  But if you start 
acting, on a project people really are interested in, like an unresponsive 
commercial developer, you definitely will get the project forked.  As 
Xfree86 found.

It can be open source and run on Windows or Mac OS, as long as you get the 
source under a copyright waiver which lets you make any mods of it you want 
as long as you pass them on.

It cannot be open source if anyone can stop you making mods and 
incorporating them into your project, even were it to run on open hardware, 
open cpu, open boot code, open OS.

It cannot be open source if you have the power to modify some of it but not 
the rest of it.  For instance, OSX is not open source, even were it to run 
on an open source kernel.

Something either is or is not open source.  There is no such thing as being 
partly or somewhat open source.

It is because being open source is a feature of the code, and not of what 
it runs on, and because the essence of it is the power to modify and re-use 
it, that the question of proprietary IDE and language is important.  This 
is not a matter of purism or obsession with detail.  It is an essential 
point, though one on which reasonable people can havbe different opinions.

Take a case where the source is released under the GPL, but to work on it 
you need an IDE which costs several tens of thousands, and is itself 
proprietary?  Can this really be open source?  Reasonable people can 
differ.  You could argue that in effect you have the source and the ability 
and the right to recreate it in Python, so it is.  Or you could argue (if 
the code generated by the IDE is totally obfuscated) that this is such a 
barrier to exploiting those rights that it is no longer open source in 
spirit.

This is why the availability of Media is an important consideration in the 
debate.  The case is much stronger, even if the IDE is proprietary, if it 
is free as in beer and freely distributable.

Peter
___
use-revolution mailing list
use-revolution@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution


Re: Open Source and Forking It

2009-10-21 Thread Richmond Mathewson

Peter Alcibiades wrote:

bForking It/b

Richard asks on another thread (not yet on Nabble, which quite often seems 
to run late) whether there is any way of being Open Source and not having 
forking.


No, definitely not.  As soon as you have any such restriction, you have 
left Open Source.  
  

This is the problem; you cannot have your cake and eat it: either the thing
is entirely unrestricted except that any modified versions have to be made
similarly available and unrestricted, or it is CLOSED.

This is where I start getting cheesed-off:

---
tangentially relevant burble starts here
---
My older son had been reading Francis Bacon and started wondering what
was wrong with gross materialism. I told him there was nothing wrong 
with it,
there is no God, no metaphysical world, and so on. Obviously, as I 
pushed this
argument to its logical extreme, he began to see that the axioms it 
rested on

were flawed. So we then went for a roll through the idea that there is no
external world and no other beings, but ALL is just a projection from 
Alexander's

mind; equally daft when pushed to its extreme.
--

Open Source is just the same as any other mind-set; if it is pushed to its
reductio ad absurdam it can be seen to be flawed. That in itself should
come as no surprise; what is surprising is that many of the party faithful
are not prepared to acknowledge that and make adjustments to take
that into account.
And this is fundamental to the idea of Open Source.  What it seeks to do is 
give people power to do whatever modifications they want to the source 
code, as long as they contribute back what they have done.  This is why 
Ballmer made such rude remarks about it.  
Sounds as if he got forked off.  Presumably the only reason the chap 
made rude
remarks was because, in some way, he felt threatened. Presumably, had he 
seen

some advantage to himself, he would have praised it to the skies.

Poor old Ballmer, as insecure as ever. One wonders why; people still 
carry on

paying the earth for Microsoft Windows, Office and so forth when Linux and
Open Office can do the same job for nothing, and, arguably, better.
This is also why it is only about 
the power to do things with the code.  What it runs on or what the 
requirements for running it are, is immaterial.  
  

Well, maybe, if one equated the RunRev engine with an Operating system,
one could term stacks Open Source ??
The classic example of recent forking occurred with X windows.  Over a 
period of a couple of months, Xorg forked from Xfree86, and secured 
universal adoption by all Linux distributions.  The team simply left and 
took the code with them.  Recently on a matter closer to home, we have seen 
a fork, or an attempted fork, of PythonCard.  If you look on DistroWatch, 
there are north of 350 different distributions of Linux now.  Well, these 
are all forks.  On the other hand, you notice there is only one Python.  
Basically, projects that keep in touch and are responsive do not get 
forked.  Ones that are of no interest do not get forked.  But if you start 
acting, on a project people really are interested in, like an unresponsive 
commercial developer, you definitely will get the project forked.  As 
Xfree86 found.


It can be open source and run on Windows or Mac OS, as long as you get the 
source under a copyright waiver which lets you make any mods of it you want 
as long as you pass them on.


It cannot be open source if anyone can stop you making mods and 
incorporating them into your project, even were it to run on open hardware, 
open cpu, open boot code, open OS.


  

QUITE: so how anything that has been built using RunRev can be Open Source
escapes me. I have released all sorts of silly little stacks into the 
wild with

no licence at all; and for all I know somebody may have used some of my
code without any acknowledgement or anything in their commercial
project (ha, ha, ha; most of my stacks only demonstrate that I can reinvent
the wheel). This is a risk, but as, for the code to work, they are 
dependent on

a closed-source engine, they are not Open Source. Probably, had I felt
those stacks had any real, lasting value, I could have spent donkey's ages
fooling around on the internet finding some licence, which, ultimately,
would have offered me just about as much protection as nothing.
It cannot be open source if you have the power to modify some of it but not 
the rest of it.  For instance, OSX is not open source, even were it to run 
on an open source kernel.


Something either is or is not open source.  There is no such thing as being 
partly or somewhat open source.
  

Well, there is: stacks written with RunRev are, surely, partly?
It is because