Searching \ for '[OT] OSS' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=oss
Search entire site for: 'OSS'.

Exact match. Not showing close matches.
PICList Thread
'[OT] OSS'
2005\12\18@061453 by Gerhard Fiedler

picon face
Xiaofan Chen wrote:

> So far most of the OSS movement is in the software world. I am not so
> sure if people like to see it in the hardware world. At least I am a bit
> scared of a GPLed design of some common circuits or an common IC. Sorry
> again I think I can not phrase properly but hopefully people can
> understand what I am trying to say.

I'm not sure I understand you. Isn't a GPLed circuit pretty much the same
as a "closed source" circuit that is too expensive for a project? The GPL
tacked on to something is a cost. You can decide whether that cost is worth
it. But differently to closed source developments, you at least can look at
it and learn from it.

Not sure that catches it... maybe you come up with an example?

Gerhard

2005\12\18@064009 by Xiaofan Chen

face picon face
On 12/18/05, Gerhard Fiedler <spam_OUTlistsTakeThisOuTspamconnectionbrazil.com> wrote:
>
> I'm not sure I understand you. Isn't a GPLed circuit pretty much the same
> as a "closed source" circuit that is too expensive for a project? The GPL
> tacked on to something is a cost. You can decide whether that cost is worth
> it. But differently to closed source developments, you at least can look at
> it and learn from it.
>
> Not sure that catches it... maybe you come up with an example?
>

I am not a legal expert but let me try with an example. Maybe I am
totally wrong.

Say if a common circuit like a follower is GPLed. We need to use it
in our design, and by use it in the circuit, our design will be consider
derived work so we will need to publish the whole circuit. That is good
if "we" do not want to earn money but not if we do want to survive.

Another example, suppose the 0.18um CMOS gate design is GPLed,
and all the 0.18um CMOS IC will be again considered as derived work.
Guess that is good to everyone except those who work for semiconductor
companies. ;-)

Just another seeming crazy example, suppose helloworld C program
is GPLed, what will happen to all the C programs?

Regards,
Xiaofan

2005\12\18@075914 by Wouter van Ooijen

face picon face
> Say if a common circuit like a follower is GPLed. We need to use it
> in our design, and by use it in the circuit, our design will
> be consider
> derived work so we will need to publish the whole circuit.
> That is good
> if "we" do not want to earn money but not if we do want to survive.
>
> Another example, suppose the 0.18um CMOS gate design is GPLed,
> and all the 0.18um CMOS IC will be again considered as derived work.
> Guess that is good to everyone except those who work for semiconductor
> companies. ;-)
>
> Just another seeming crazy example, suppose helloworld C program
> is GPLed, what will happen to all the C programs?

You seem to miss one important point: GPL is just a specific form of
licensing of a copyrighted work. If something can not be copyrighted, it
can not be GPL-ed. So you must first ask yourself: can those things be
copyrighted? (I think no, or at lesast not effectively - copyright is
closely tied to a specific way of 'stating' somthing, so it would not be
effective angainst most independently-derived 'derivates'.) If not, it
can't be GPL-ed. Next you must realise that *if* it can be copyrighted,
it would probably *be* copyrighted - to be released on royalty payment
only. So if you are against GPL on the grounds you mention you are in
fact opposing copyright law. Not that I think you are, you seem to have
a much to broad view of what copyright does.

Note that a follower (I assume you mean a gain-1 non-inverting opamp
circuit?) could be *patented* if it was sufficiently new. But patent law
is a different story, and I am not aware of any GPL-like things in the
patent world (Yes, I do know GPL has certain claims on patents - but it
still derives its validity from copyright only).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\18@113545 by Danny Sauer

flavicon
face
Xiaofan wrote regarding 'Re: [OT] OSS' on Sun, Dec 18 at 05:44:
> I am not a legal expert but let me try with an example. Maybe I am
> totally wrong.
>
> Say if a common circuit like a follower is GPLed. We need to use it
> in our design, and by use it in the circuit, our design will be consider
> derived work so we will need to publish the whole circuit. That is good
> if "we" do not want to earn money but not if we do want to survive.
>
> Another example, suppose the 0.18um CMOS gate design is GPLed,
> and all the 0.18um CMOS IC will be again considered as derived work.
> Guess that is good to everyone except those who work for semiconductor
> companies. ;-)
>
> Just another seeming crazy example, suppose helloworld C program
> is GPLed, what will happen to all the C programs?

Those are examples of things which are already available in the public
domain, and are obvious enough in implementation that they could be
recreated in a "clean room" environment without having to see the
original work.  If someone could recreate the functionality without
seeing the original design, then their new work would not be covered
under the original license.  That's sort of how Wine works - they have
the specs for what goes in and what comes out, but they haven't
actually seen the Windows source.  So they recreate the functionality
and are not encumbered by whatever license would be required to view
MS's copyrighted code.

For hardware, it might be be better to use a modified BSD license [1]
or the Apache license [2].  Both basically allow redistribution of
derived works as long as any source distribution includes the original
license statements (where the licensed code is used only) and gives
credit to the original author.  They do not require that the
derivative work be licensed under the same license, or that
modifications be made public.

[1] http://www.xfree86.org/3.3.6/COPYRIGHT2.html#5
[2] http://www.apache.org/licenses/LICENSE-2.0

I generally release stuff under the GPL with contact information - if
someone wants to redistribute modified versions of things I write,
I'll happily license it to them under different terms - but basically
that's because I'd just like to know if someone's using my code.  I'm
not in it for the money - I just want anyone who can benefit from my
work to be able to do so.  Someone developing for money would likely
have a different view. :)

--Danny

2005\12\18@114145 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 06:12, Gerhard Fiedler a écrit :
> I'm not sure I understand you. Isn't a GPLed circuit pretty much the same
> as a "closed source" circuit that is too expensive for a project? The GPL
> tacked on to something is a cost. You can decide whether that cost is
> worth it. But differently to closed source developments, you at least can
> look at it and learn from it.

You can do more than look and learn: you can modify and keep for yourself, or modify and redistribute your modifications. Cost is not an issue with GPL; you can ask for money or not, depending on your contracting agreements.

> Not sure that catches it... maybe you come up with an example?

http://opencores.org/ is one pretty big example.
--
Marc

2005\12\18@115308 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 06:40, Xiaofan Chen a écrit :
> I am not a legal expert but let me try with an example. Maybe I am
> totally wrong.
>
> Say if a common circuit like a follower is GPLed. We need to use it
> in our design, and by use it in the circuit, our design will be consider
> derived work so we will need to publish the whole circuit. That is good
> if "we" do not want to earn money but not if we do want to survive.
>
> Another example, suppose the 0.18um CMOS gate design is GPLed,
> and all the 0.18um CMOS IC will be again considered as derived work.
> Guess that is good to everyone except those who work for semiconductor
> companies. ;-)
>
> Just another seeming crazy example, suppose helloworld C program
> is GPLed, what will happen to all the C programs?

Nobody can claim the ownership of the helloworld concept using a license.
A copyrighted software is not a patent.

But M$, IBM, Apple or some other evil company might eventually patent it.
For more on this hot topic:
http://www.nosoftwarepatents.com/
--
Marc

2005\12\18@120339 by Nate Duehr
face
flavicon
face
William Chops Westfield wrote:

> Perhaps the OSS movement is a reaction to this; back in the days when
> universities used mainframes, it was pretty common for each university
> to have their own hacks in the OS (and frequently their own hardware as
> well.)  As mainframes went away and were replaced by micros running
> (usually) windows, that option went away.  Till linux/etc came along.

AFAIK, the first truly "open" license for OS's that was both free as in
free-beer and Free as in Freedom was the BSD license, not Linux.  So
that should say "until BSD came along".

Actually, there were probably others, but BSD would have been the first
one that reall "took off".

There are a number of things that are involved in software licensing.
As I think through it, I think these are the important ones:

- Copyright
- Patents
- License

For example, one of the projects I help out with uses software that is
truly Public Domain, as licensed by the author, and the author retains
no Copyright and there are no known Patents that affect the software.
This is probably the most "Free" code I work with regularly.  PD means
you can literally do anything with it -- the author retains NO rights,
and you can even take their code and make a completely closed product
from it if you like.

Next up, there are licenses like BSD that specifically state you can
take the open source "gift" and turn it into a commercial closed product
if you retain the Copyright and attributes to the University of
California at Berkeley.  BSD licensed code is a good starting point for
commercial applications where you don't want to be encumbered by license
issues, and is considered one of the more open licenses.

Another example of a semi-open license is the Perl scripting language.
The Perl license allows the use of the Perl scripting language for just
about anything, but retains the rights to the name "Perl" and is known
as the "Artistic" license.  The author retains the rights to the code
called "Perl"... take his code and do anything else you like with it,
just don't call it Perl.  Perl is actually "double-licensed" under
either the "Artistic" or GPL licenses.  It's up to the next person who
uses it as to which they wish to use for their modifications, and Larry
Wall is open to either.  It's well-explained on the perl.org website.

And then there's the venerable GPL.  GPL is "viral" in that it requires
the next developer to release their source code if they create anything
for public consumption using the original GPL code.  Obviously the GPL
is more "controversial" than other licenses, since it uses the license
to push a social agenda -- sharing code.

All of the above licenses have no restrictions on charging money for
binaries, pre-compiled systems, or service on these systems.
Additionally, none of them have any restrictions on requiring software
use keys or other DRM type systems -- but in the case of the licenses
that require the code be open, anyone could see how you are doing the
DRM and/or release a version with the DRM removed... so it probably
wouldn't be useful.  However, if your end-users respect your need to
make a living, they may just ignore the fact that they could bypass your
system and might continue to support you -- you just never know.
Additionally if your DRM code is closed and just run by your GPL
application, well -- there'd be people that would complain, but that's
probably a reasonable way to handle it too... of course, again, your
interface to your DRM application would be visible for the world to see
and/or bypass if your support wasn't up to par or someone just decided
to be rude.

Finally there's also licenses like the LGPL (the so-called "Lesser" GPL)
for things like libraries that may need to be linked into closed code.
I haven't had any opportunity to build anything that needs/uses the
LGPL, but it's out there as another option.

As far as Patents go, my personal opinion is that patent law is quite
out of control, at least here in the U.S. where the USPTO (U.S. Patent
and Trade Office) is far too willing to hand out patents for things that
really aren't new or innovative, and is undergoing a troubled time where
some issues need to be addressed, and that due to the nature of the
expense inherent in the U.S. legal system, tends to be more a tool for
large corporations than it is anything other than a nice "feather in
your cap" for the individual developer... but it's there to attempt to
protect everyone equally.  (I think it tends to lead more toward the
"How much justice can you afford?" motto, in real-world practice.)

I guess that's the point of this post -- to educate and share that there
are a MYRIAD of options for both licenses and copyright (and copyright
law changes from country to country, making things REALLY interesting).
Additionally, depending on the country you're in, patents might
override all of the above.  But the important thing to know as a
developer -- you NEED to pay attention to the business aspects of
software licensing and Copyright.  Release your code just once under the
wrong license, and you may find yourself watching someone else benefit
greatly from your work.  If that was your intent (GPL fanatics), you may
not mind... as long as the person using your code shares their changes
and "passes it on".  If you're really looking to "give away" your code
to the world completely without encumberances, the BSD or true Public
Domain licenses may be the way to go.  And many developers blindly slap
any old license (if any at all) on their code and end up regretting it
later.  It's worth paying attention to these things, and deciding early
when/if you'll use a license.

Okay, there's my not-so-informative post for those in-the-know, and
hopefully helpful to any new developers out there reading along...

Nate, .....nateKILLspamspam@spam@natetech.com

2005\12\18@120511 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 07:59, Wouter van Ooijen a écrit :
> Note that a follower (I assume you mean a gain-1 non-inverting opamp
> circuit?) could be *patented* if it was sufficiently new. But patent law
> is a different story, and I am not aware of any GPL-like things in the
> patent world (Yes, I do know GPL has certain claims on patents - but it
> still derives its validity from copyright only).

Excerpt from http://www.gnu.org/copyleft/gpl.html :
"Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all."

For example:
http://www.advogato.org/article/89.html
http://www.levien.com/patents.html

There's a strong link between patents and copyright. Local patent laws are causing big headaches. The next verison of the GPL will try to address this problem: www.eweek.com/article2/0,1759,1730102,00.asp
--
Marc

2005\12\18@122842 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 12:03, Nate Duehr a écrit :
> For example, one of the projects I help out with uses software that is
> truly Public Domain, as licensed by the author, and the author retains
> no Copyright and there are no known Patents that affect the software.
> This is probably the most "Free" code I work with regularly.  PD means
> you can literally do anything with it -- the author retains NO rights,
> and you can even take their code and make a completely closed product
> from it if you like.

By "close", do you mean "proprietary"?
You can't turn PD work into something proprietary...
http://en.wikipedia.org/wiki/Public_domain :
"The public domain comprises the body of knowledge and innovation (especially creative works such as writing, art, music, and inventions) in relation to which no person or other legal entity can establish or maintain proprietary interests."

> And then there's the venerable GPL.  GPL is "viral" in that it requires
> the next developer to release their source code if they create anything
> for public consumption using the original GPL code.  Obviously the GPL
> is more "controversial" than other licenses, since it uses the license
> to push a social agenda -- sharing code.

Stating that the GPL is "viral" is sending a bad message:
Excerpt from the "Criticism" section of
http://en.wikipedia.org/wiki/GNU_General_Public_License :
"However, the GPL does not force copyright owners to do anything with code they own. Copyright owners of works licensed with the GPL are free to negotiate alternate terms with authors of derived works. Dual-licensing is becoming more common, where software licensed under the GPL can be made available under a commercial license for a fee, allowing others to create derived works without licensing them under the GPL."

--
Marc

2005\12\18@123230 by Nate Duehr

face
flavicon
face
Xiaofan Chen wrote:

> So far most of the OSS movement is in the software world. I am not so
> sure if people like to see it in the hardware world. At least I am a bit
> scared of a GPLed design of some common circuits or an common
> IC. Sorry again I think I can not phrase properly but hopefully people
> can understand what I am trying to say.

See http://www.opencores.org - an example of a good use of open-source licenses
in hardware.

Another good use is in emerging technologies that need a push, like
Software Defined Radio needed a few years ago:

http://www.gnu.org/software/gnuradio/

Nate, natespamKILLspamnatetech.com

2005\12\18@140114 by Wouter van Ooijen

face picon face
> PD means you can literally do anything with it
> -- the author retains NO rights

That is not strictly true. Anything readable that is published is
covered by a copyright. The author can grant free use of his copyrighted
material to everyone, but he still holds the copyright (otherwise he
would not be able to grant any rights to everyone!)

> You can't turn PD work into something proprietary...
AFAIK you can. You create a derived work, which is automatically
copyrighted by you. If you owe nothing to the original author you
affectively have the full copyright, including the right to keep

> http://en.wikipedia.org/wiki/Public_domain :

I don't think that article states what you stated. And anyway, a wiki is
not the ultimate truth. It is just something written my someone.

> Stating that the GPL is "viral" is sending a bad message:

So what? It is still true.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\18@140114 by Wouter van Ooijen

face picon face
> See http://www.opencores.org - an example of a good use of
> open-source licenses in hardware.

Where is the license they use? The only thing I could find quickly is a
GPLed FAQ :(

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\18@140115 by Wouter van Ooijen

face picon face
> There's a strong link between patents and copyright.

Which link? I am not aware of any specific link.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\18@153628 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Stating that the GPL is "viral" is sending a bad message:
>
> So what? It is still true.

First of all, I think the expression "viral" in this context is not a
factual statement (which could be true or false), it is an analogy. I don't
think an analogy can be "true". It can be good or bad, more or less
appropriate, but not "true".

Second, I don't think "virus" is a good analogy. I at least can't come up
with a good relationship between the major characteristics of GPL code and
a virus.

The most outstanding characteristic of a virus -- and the one that is
probably for most people the strongest association -- is that it inflicts
damage without the damaged entity having a choice. I don't see that in GPL
code at all. There's always the choice to use it or not to use it. Calling
the GPL license a virus in a code that's protected by it (the cost is that
derived works have to be GPL'ed) is not different from calling the
dependency of Windows software on the Windows OS a virus (the cost is that
the user has to use and therefore buy or have bought a Windows license).
Both are not appropriate analogies, IMO, as in both cases there is a
conscious choice involved -- that's not usually the case with viruses
(whether biological or software).

Gerhard

2005\12\18@155655 by Wouter van Ooijen

face picon face
> Second, I don't think "virus" is a good analogy. I at least
> can't come up with a good relationship between the major
> characteristics of GPL code and a virus.

Seen from a company perspective (not from an individual programmer or
program manager) it definitely has viral properties.

> Calling
> the GPL license a virus in a code that's protected by it (the
> cost is that
> derived works have to be GPL'ed) is not different from calling the
> dependency of Windows software on the Windows OS a virus

I would not object to that either.

> Both are not appropriate analogies, IMO, as in both cases there is a
> conscious choice involved -- that's not usually the case with viruses
> (whether biological or software).

Ok, thats a vieuwpoint, but I think the viral (or maybe better:
contaminating) properties of GPL are an important issue. But don't take
me wrong: I admire the concept! IMHO the GPL is definitely the greatest
creation of RS, I rate it even above GCC.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\18@172001 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 14:01, Wouter van Ooijen a écrit :
> > You can't turn PD work into something proprietary...
>
> AFAIK you can. You create a derived work, which is automatically
> copyrighted by you. If you owe nothing to the original author you
> affectively have the full copyright, including the right to keep

A devired work is something different. It's not a PD work. All derived work must comply with the copyright of the original work. The PD concept applies to the expiration of copyrighted works. PD is the absence of copyright. So, even if new editions of old books are derived works and are copyrighted, their content are not copyrighted, at least not the parts that are in the PD; I can even scan the content of new editions, redistribute it for free, or encrypt it and ask for millions (hoping that some fools will pay you). Even in extremely closed forms, a work in the PD is not proprietary.

> > http://en.wikipedia.org/wiki/Public_domain :
>
> I don't think that article states what you stated. And anyway, a wiki is
> not the ultimate truth. It is just something written my someone.

All good sources of knowledge are reviewed by peers, which is certainly the case of Wikipedia. I'm pretty sure than enough competent people reviewed this article about Public domain. Even "Nature", a respected scientific publication, considers that Wikipedia is as good as the Britannica encyclopedia: www.nature.com/news/2005/051212/full/438900a.html
I was once sceptic about the publication process of Wikipedia, because an opened wiki can easily be "defaced", but they decided to "semiclose" it:
http://en.wikipedia.org/wiki/Wikipedia:Semi-protection_policy.

--
Marc

2005\12\18@180844 by Gerhard Fiedler

picon face
Marc Lavallée wrote:

>> I'm not sure I understand you. Isn't a GPLed circuit pretty much the same
>> as a "closed source" circuit that is too expensive for a project? The GPL
>> tacked on to something is a cost. You can decide whether that cost is
>> worth it. But differently to closed source developments, you at least can
>> look at it and learn from it.

> Cost is not an issue with GPL; you can ask for money or not, depending on
> your contracting agreements.

It seems you didn't understand what I meant with "cost". The GPL imposes a
cost of using GPL code, and the cost imposed is that published derivative
works are also open source and GPL.

> You can do more than look and learn: you can modify and keep for
> yourself, or modify and redistribute your modifications.
Only if the GPL is acceptable for you. If it isn't, look and learn is
pretty much all you can do :)

Gerhard

2005\12\18@182433 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

> but think about why someone would make their code publicly available
> without any kind of paid royalty.  They have already decided not to try
> to get paid for that code.  There are therefore two main reasons left
> that I can see, some fame and recognition for themselves, and a desire
> to help the progress of software in general.

Here's where you seem to fall short. How about open source software? I'm
pretty sure many developers publish their code as GPL because they want to
see more open source code, and they are willing to contribute to that goal
-- under the condition that whoever uses their code contributes also.

> I think the real reason for the public source requirements of the GPL
> comes from an attitude that it's evil for someone to make a profit.
> [...] So in a nutshell, the GPL is the preachy result of an arrogant and
> unrealistic world view.

Or the result from liking open source better. Just as simple as that.
Nobody has to use GPL software in anything she does; there's always Windows
CE... :)

One could just as easily say that the ones who want free software that's
not GPL'ed want a free lunch and bitch if they don't get it :)

Gerhard

2005\12\18@183048 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Second, I don't think "virus" is a good analogy. I at least can't come
>> up with a good relationship between the major characteristics of GPL
>> code and a virus.
>
> Seen from a company perspective (not from an individual programmer or
> program manager) it definitely has viral properties.

A company -- just like an individual -- can make an informed and conscious
decision to use or not to use GPL code. (And they definitely should make
that decision consciously, and not only download code from the net, use it
and the afterwards, when they don't want to publish their code, complain
that it's GPL code...)  

That doesn't sound "viral" to me. I haven't heard of anybody consciously
downloading a virus to infect her network.

>> Calling the GPL license a virus in a code that's protected by it (the
>> cost is that derived works have to be GPL'ed) is not different from
>> calling the dependency of Windows software on the Windows OS a virus
>
> I would not object to that either.

Here it starts not making sense to me. This means (if I understand you
correctly) that pretty much every software is a virus, in that it creates
dependencies, that may create other dependencies... I wouldn't object to
the viewpoint, but it seems then useless, as it's no distinction anymore.
If I apply this, every software has a "viral" quality, not only GPL
software, other open source software, or closed source software -- simply
/every/ software. And not only software...

Gerhard

2005\12\18@185933 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 18:06, Gerhard Fiedler a écrit :
> It seems you didn't understand what I meant with "cost". The GPL imposes
> a cost of using GPL code, and the cost imposed is that published
> derivative works are also open source and GPL.

In economic terms, publishing free software is largely compensated in the form of a large "free as in speech" code base. What you call s "cost" is actually an investment.

> > You can do more than look and learn: you can modify and keep for
> > yourself, or modify and redistribute your modifications.
>
> Only if the GPL is acceptable for you. If it isn't, look and learn is
> pretty much all you can do :)

As long as "look and learn" doesn't mean "copy/paste", or stealing. With the GPL (and other fine free software licenses, not only the GPL) it's definitely possible to reimplement the code in some proprietary form, but that's a waste of energy, and it's also asocial. There were attempts to directly redistribute GPL code in binary form, disguised as proprietary software, but it didn't last long (Sigma Design  vs XVid and KISS vs Mplayer are good examples).

SCO also wants us to look at "their" old code, but we better not...
--
Marc

2005\12\18@190244 by William Chops Westfield

face picon face
On Dec 18, 2005, at 3:06 PM, Gerhard Fiedler wrote:
>
>> You can do more than look and learn: you can modify and keep for
>> yourself, or modify and redistribute your modifications.
>
> Only if the GPL is acceptable for you. If it isn't, look and learn is
> pretty much all you can do :)
>
Serious companies find that they can't even do that, as there's a chance
that their minds will become "contaminated" by something they've seen,
causing the software they write that was supposed to be proprietary to
suddenly fall under GPL.  For instance, I know that the compiler writers
at intel are not allowed to get any where NEAR gcc...  I don't know if
the gnu people would actually go so far as to go after such cases, but
it's definitely something corporate lawyers worry about...

The most virus-like aspect of gpl/etc is the way it can suddenly cause
in unexpected results.  For instance, we just said that LGPL only leaves
applications proprietary if libraries are dynamically linked.  That's
swell for windows and linux, but none of my embedded systems have a
facility for doing dynamically linked libraries.  Does that mean that
everything I write with avrgcc that uses the provided C libraries
(perhaps
even the internal libraries that do multi-byte math or floating point)
should be open source?  Ewww!

BillW

2005\12\18@191733 by Marc Lavallée

flavicon
face
Le 18 Décembre 2005 14:01, Wouter van Ooijen a écrit :
> > There's a strong link between patents and copyright.
>
> Which link? I am not aware of any specific link.

Eben Moglen explains it better than I can:
Excerpt from http://emoglen.law.columbia.edu/publications/lu-16.html :

"Copyright law permits authors to control the copying and distribution of their expressions, but copyright law does not permit the monopolization of ideas. Patent law works differently: once someone has been permitted by a patent office to "claim" a technical process or invention, no one in the place where that patent has issued may "practice" of the patent, even if he independently invented, without a license."
--
Marc

2005\12\18@201140 by olin piclist

face picon face
Gerhard Fiedler wrote:
> Here's where you seem to fall short. How about open source software? I'm
> pretty sure many developers publish their code as GPL because they want
> to see more open source code, and they are willing to contribute to
> that goal -- under the condition that whoever uses their code
> contributes also.

But that's exactly the social agenda I was talking about.

******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\18@224855 by Peter

picon face
William "Chops" Westfield <westfw <at> mac.com> writes:

> suddenly fall under GPL.  For instance, I know that the compiler writers
> at intel are not allowed to get any where NEAR gcc...  I don't know if
> the gnu people would actually go so far as to go after such cases, but
> it's definitely something corporate lawyers worry about...

As they should ...

> The most virus-like aspect of gpl/etc is the way it can suddenly cause
> in unexpected results.  For instance, we just said that LGPL only leaves
> applications proprietary if libraries are dynamically linked.  That's
> swell for windows and linux, but none of my embedded systems have a
> facility for doing dynamically linked libraries.  Does that mean that
> everything I write with avrgcc that uses the provided C libraries
> (perhaps
> even the internal libraries that do multi-byte math or floating point)
> should be open source?  Ewww!

It might. And the solution would be if someone (f.ex. Olin) would have a GPLd
version of, say, his floating point libs. Then you would hard link against his
library. Once happy that it works (you do not need to publish anything at this
time, since you are not distributing it - yet), you can contact Olin and work
out a license for his libs. Then you hard link against the (improved, for you,
by  Olin, at a price) library supplied by Olin and everyone is happy. Or gone to

http://www.uclibc.org/FAQ.html and read a little (Q6). The 'linkable object
file' you would have to provide would be in no way different from a dll you
would purchase commercially for another OS. It is not 'source code'.

See how the GPL fear prevented you from ripping code that *might* have been
someone else's ? You apply self censorship because of this. GPL at work! In
practice, I doubt any of the people whose code you might have been tempted to
'quietly incorporate', would have any resources to sue you or your employer.
But under the GPL, you can't, you don't want to rip (just supposing)!

Peter

ok, this thread was long enough. I am getting off it.

2005\12\19@011249 by Tad Anhalt

picon face
William Chops Westfield wrote:
> The most virus-like aspect of gpl/etc is the way it can suddenly cause
> in unexpected results.  For instance, we just said that LGPL only leaves
> applications proprietary if libraries are dynamically linked.  That's
> swell for windows and linux, but none of my embedded systems have a
> facility for doing dynamically linked libraries.  Does that mean that
> everything I write with avrgcc that uses the provided C libraries (perhaps
> even the internal libraries that do multi-byte math or floating point)
> should be open source?  Ewww!

  Check the license that was distributed with avrgcc.  It's been a
while since I had a reason to look into this (and not for avr at any
time), but the "standard" libraries probably have a specific exemption
allowing you to link them however you want.

  In other words, I'd suspect that you could compile a program that
would run and you'd be able to use sprintf, etc.  But you may not be
able to use some of the bundled libraries that would make your life more
convenient.

  A quick search on google turned up this (not directly applicable, but
this is the sort of thing you are looking for):

{Quote hidden}

  You do want to be very careful with compile flags.  If you do fall
under such an exemption, you'll want to double and triple check that you
aren't linking in a library that doesn't provide the same exemption.

  There may also be #defines or compiler flags that need to be set in
order to insure that you don't end up inadvertently including stuff that
is not covered (especially in C++ where a lot of implementation details
actually end up in header files).  The standard (eg: iostream, etc.)
files do seem to have the appropriate verbiage, but don't take my word
for it, check it yourself.

HTH,
Tad Anhalt

2005\12\19@020656 by Wouter van Ooijen

face picon face
> > > You can't turn PD work into something proprietary...
> >
> > AFAIK you can. You create a derived work, which is automatically
> > copyrighted by you. If you owe nothing to the original author you
> > affectively have the full copyright, including the right to keep
>
> A devired work is something different. It's not a PD work.

True. But the effect is still that you can take a PD work at source
level and create a proprietory derivate of it (the derivate effort can
be zero). So you *can* turn a PD work into something proprietory. That
is the main difference between a PD (or almost PD, like BSD) work and a
GPL work.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@020656 by Wouter van Ooijen

face picon face
> Here it starts not making sense to me. This means (if I understand you
> correctly) that pretty much every software is a virus

Not *is* a virus, but *has viral properties*, to so extent. This is just
my opinion, not a mathematical truth. I look at how something behaves at
the large, not how individual dicesions are made. In this respect GPL
and Windows have a high viral quality.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@020656 by Wouter van Ooijen

face picon face
> > > There's a strong link between patents and copyright.
> >
> > Which link? I am not aware of any specific link.
>
> Eben Moglen explains it better than I can:
> Excerpt from http://emoglen.law.columbia.edu/publications/lu-16.html :
>
> [snip]

OK, no link at all except that they are both intellectual property laws.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@034738 by Nate Duehr

face
flavicon
face
On Sunday 18 December 2005 01:33 pm, Gerhard Fiedler wrote:

> Second, I don't think "virus" is a good analogy. I at least can't
> come up with a good relationship between the major characteristics of
> GPL code and a virus.

If a derived work is created from GPL'ed code, the derived work must
also be GPL'ed.  That's the "viral" part.

You can't change (or dual-license) the portion that's someone else's
work prior to yours, so for all intents and purposes, once something is
released under the GPL *ONLY* as a starting point, everything derived
from that work will also be GPL'ed only, unless arrangements are made
with the upstream author to dual-license later derived works.

That's how I read it all, anyway... If it all ever really gets to a
court or two, which will take provable multi-million dollar GPL
licensing abuse and an advocate willing to take someone that cocky to
court, we'll see how the judges feel.

--
Nate Duehr, .....nateKILLspamspam.....natetech.com

2005\12\19@035212 by Nate Duehr

face
flavicon
face
On Sunday 18 December 2005 03:19 pm, Marc Lavallée wrote:
> Le 18 Décembre 2005 14:01, Wouter van Ooijen a écrit :
> > > You can't turn PD work into something proprietary...
> >
> > AFAIK you can. You create a derived work, which is automatically
> > copyrighted by you. If you owe nothing to the original author you
> > affectively have the full copyright, including the right to keep
>
> A devired work is something different. It's not a PD work. All
> derived work must comply with the copyright of the original work. The
> PD concept applies to the expiration of copyrighted works. PD is the
> absence of copyright. So, even if new editions of old books are
> derived works and are copyrighted, their content are not copyrighted,
> at least not the parts that are in the PD; I can even scan the
> content of new editions, redistribute it for free, or encrypt it and
> ask for millions (hoping that some fools will pay you). Even in
> extremely closed forms, a work in the PD is not proprietary.

Okay, let's say I agree with all of your assertions.  (And I do.)

I take a PD codebase, create a work, compile binaries and release my new
program to the world -- if the original author retains no copyright
(you said yourself none exists), and I don't release my source code...
where would the real-world lawsuit come from to challenge the person
who created the derived work?  

There's a difference between the letter of PD law and the practice in
reality.  Even if the derived work is supposedly PD somehow -- who'd
care?  No judge is going to require a person creating a derived work
from something explicitly stated as PD and where the original copyright
owner explicitly gave away all rights, to release their new source or
call their new derived work anything but their own.

--
Nate Duehr, EraseMEnatespam_OUTspamTakeThisOuTnatetech.com

2005\12\19@035828 by Nate Duehr

face
flavicon
face
On Sunday 18 December 2005 12:01 pm, Wouter van Ooijen wrote:
> > See http://www.opencores.org - an example of a good use of
> > open-source licenses in hardware.
>
> Where is the license they use? The only thing I could find quickly is
> a GPLed FAQ :(

Darn, I was going off of their original plan which was to use GPL
licenses or similar.  But now the site isn't responding from here and
you have me curious if they were able to stick with it.

--
Nate Duehr, natespamspam_OUTnatetech.com

2005\12\19@041800 by Wouter van Ooijen

face picon face
> Darn, I was going off of their original plan which was to use GPL
> licenses or similar.  But now the site isn't responding from here and
> you have me curious if they were able to stick with it.

The problem with GPL + hardware is that GPL talks about things like
combining into an application, which is more or less well-defined when
talking about software. But how would you interpret that when deling
with hardware? That is a very important question, because it is the
boundary of the GPL viral/contaminating effect. With software the
boundary is the linked application. But were would the boundary be with
hardware? Chip level? PCB level? Apparatus level?

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@061840 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Here it starts not making sense to me. This means (if I understand you
>> correctly) that pretty much every software is a virus
>
> Not *is* a virus, but *has viral properties*, to so extent. This is just
> my opinion, not a mathematical truth. I look at how something behaves at
> the large, not how individual dicesions are made. In this respect GPL
> and Windows have a high viral quality.

Ok. I haven't expressed myself well. The question still stands (now
slightly rephrased): Using this meaning of "viral quality", is there
(successful, in actual use) software that doesn't have it?

You say that GPL software has it, and that Windows software has it. I think
that covers probably 80% of existing software (just a rough guess).
Extending the argument, any software designed to run on Linux also has the
same "viral quality" (of requiring GPL software to run it, in this case).
That probably brings it up to 90%. Then add all the software that is
written for /any/ specific platform (because, of course, they also have
that "viral quality" of requiring/promoting their specific platform).

So what software doesn't have that "viral quality" you are talking about?

Gerhard

2005\12\19@062717 by Gerhard Fiedler

picon face
Nate Duehr wrote:

>> Second, I don't think "virus" is a good analogy. I at least can't
>> come up with a good relationship between the major characteristics of
>> GPL code and a virus.
>
> If a derived work is created from GPL'ed code, the derived work must
> also be GPL'ed.  

I know that this is the case.

> That's the "viral" part.

What is viral in this? Has anybody using this term ever thought about how a
virus works? What's the semblance? A few viruses are passed from parents to
children, but not all. So passing on the GPL from parent works to derived
"children" works is not really a "viral" quality. "Viral" sounds good, but
only as long as you don't start to think about it.

The prime quality of a virus is that you contract it without being asked to
decide whether you want it or not. That's not how it works with GPL code.

So where's the "viral" analogy exactly?


> You can't change (or dual-license) the portion that's someone else's work
> prior to yours, so for all intents and purposes, once something is
> released under the GPL *ONLY* as a starting point, everything derived
> from that work will also be GPL'ed only, unless arrangements are made
> with the upstream author to dual-license later derived works.

Right, that's my understanding, too. How is this different from other
licensing schemes? I have used non-GPL, "CCL" ("classic commercial license"
:) code that didn't allow me to sell it. Nobody talks about this being
"viral". I have used CCL code that didn't allow me to create derivative
works in the first place. Nobody talks about this being "viral" either.

Gerhard

2005\12\19@063302 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

> OK, no link at all except that they are both intellectual property laws.

Maybe one link is that a copyright that precedes a patent on something used
in the code may not help you if you don't have the money to fight the
claims of the patent holder. In that case, you may have the copyright to
your code, but can't use it because it violates a (later filed) patent.

Gerhard

2005\12\19@064242 by Gerhard Fiedler

picon face
Marc Lavallée wrote:

>> Only if the GPL is acceptable for you. If it isn't, look and learn is
>> pretty much all you can do :)
>
> As long as "look and learn" doesn't mean "copy/paste", or stealing.

How do you get the idea that "look and learn" could mean "copy/paste"? If
you learn by copying and pasting, or maybe copy and paste by looking, I
envy you :)

Gerhard

2005\12\19@065318 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

> Gerhard Fiedler wrote:
>> Here's where you seem to fall short. How about open source software? I'm
>> pretty sure many developers publish their code as GPL because they want
>> to see more open source code, and they are willing to contribute to
>> that goal -- under the condition that whoever uses their code
>> contributes also.
>
> But that's exactly the social agenda I was talking about.

So what's the problem? Everybody has "social agendas" of this type,
including you. When releasing some of your code with your license, you
further a different "social agenda". When not releasing at all some other
code you write, you further again a certain "social agenda". When you sell
your programmers and services, you further still another "social agenda".
When trying to get along with your neighbor, that's also such a "social
agenda". Everybody likes the things he likes, and everybody tries to get
things a bit closer to that.

That's one of the uses of expressions that makes these expressions
worthless, because in this sense pretty much every action that has effects
on others (and what action doesn't?) has such a "social agenda".

Gerhard

2005\12\19@065847 by Wouter van Ooijen

face picon face
> Ok. I haven't expressed myself well. The question still stands (now
> slightly rephrased): Using this meaning of "viral quality", is there
> (successful, in actual use) software that doesn't have it?

There is lots of software that is succesfull but not so dominant as to
create a de-facto lock-in. Many compilers, fall into this category, just
to name something. Adobe PDF is very dominant, but not closed enough to
create lock-in (there are alternatives, both for the reader and the
writer). In the OSS department there is everything that is BSD-licensed.

> You say that GPL software has it, and that Windows software
> has it. I think
> that covers probably 80% of existing software (just a rough guess).
> Extending the argument, any software designed to run on Linux
> also has the
> same "viral quality" (of requiring GPL software to run it, in
> this case).

I would argue with that: although running software that requires Linux
indeed requires Linux I would only say it has viral/contaminating
properties when it both
- locks you in in the sense that after some time you have almost no
alternative
- forces you into something you otherwise would not do (release your own
sources, pay $$$, etc)

> So what software doesn't have that "viral quality" you are
> talking about?

Its of course not a black-and-white issue. I would say for instnace that
without star office the uSoft office package (especially Word) would
have this quality. With Star Office this is history (you are no longer
locked in). The Apache server does not have this quality, because it
does not force you into any actions (you are not required to release
your HTML pages under any license).

Again, don't take me wrong: the fact that I call something
viral/contaminating is not a moral/ethical/value attribution. It is just
that in my observation things have this property to a more or less
effect. The dominance of Windows is (party) caused by this effect. The
fear among some software companies shows that the GPL principle (note:
not all OSS licenses!) also scores high in this category.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@070303 by Wouter van Ooijen

face picon face
> > OK, no link at all except that they are both intellectual
> property laws.
>
> Maybe one link is that a copyright that precedes a patent on
> something used
> in the code may not help you if you don't have the money to fight the
> claims of the patent holder. In that case, you may have the
> copyright to
> your code, but can't use it because it violates a (later
> filed) patent.

But in that sense copyright is in the same league as any other
itellectual property right (like artistic rights, trade mark rights,
etc).

Note that in the case you describe it is the patent holder that has to
sue you, not the other way round. And a proveable copyright (which
requires publication, otherwise you can't have a copyright!) would be a
very easy proof of prior art, so I doubt the patent holder would gain
much by sueing the copyright holder.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@073743 by Wouter van Ooijen

face picon face
>> (OSS agenda)
>> But that's exactly the social agenda I was talking about.
>
> So what's the problem? Everybody has "social agendas" of this type,

Think of Microsoft. They definitely have an agenda (Windows on every
device), and they are as clear about it as OSS hardliners (all software
under GPL). In fact every company has an agenda (let's make as much
profit as we can) and most other institutions too.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@080202 by Xiaofan Chen

face picon face
On 12/19/05, Tad Anhalt <@spam@tjaKILLspamspamnetins.net> wrote:

>   Check the license that was distributed with avrgcc.  It's been a
> while since I had a reason to look into this (and not for avr at any
> time), but the "standard" libraries probably have a specific exemption
> allowing you to link them however you want.

I think you mean avr-libc. It is a nice license --> modified BSD license.
I guess it should be the preferred to distribute embedded MCU
libraries.

Regards,
Xiaofan

>From http://www.nongnu.org/avr-libc/LICENSE.txt:

The contents of avr-libc are licensed with a Modified BSD License.

All of this is supposed to be Free Software, Open Source, DFSG-free,
GPL-compatible, and OK to use in both free and proprietary applications.

See the license information in the individual source files for details.

Additions and corrections to this file are welcome.
....
....
....

2005\12\19@090404 by olin piclist

face picon face
Gerhard Fiedler wrote:
>> But that's exactly the social agenda I was talking about.
>
> So what's the problem?

None except that they pretend they are doing it for other reasons.

> Everybody has "social agendas" of this type,
> including you. When releasing some of your code with your license, you
> further a different "social agenda".

I don't agree with this broad use of the term "social agenda".  I'm not
trying to get others to behave in a certain way nor try to set new norms for
acceptable behavior by releasing my code.  I do it for selfish reasons,
which I'm not trying to hide.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\19@104747 by Wouter van Ooijen

face picon face
> I don't agree with this broad use of the term "social
> agenda".  I'm not
> trying to get others to behave in a certain way
> nor try to set new norms for
> acceptable behavior by releasing my code.  I do it for
> selfish reasons,
> which I'm not trying to hide.

I agree with you that the "I, you, they" in this discussion can get
confusing :)

But who is hiding anything? Stallmans' motives behind the GPL are well
documented, by him and by others. Nearly every OSS license has a text
explaining its intent. So what's the trouble? If they (for some value of
'they') want to strive to a world where all software is GPLed, and they
do it by legal means and they put their efforts where there mouth is,
what's wrong with that? Yes, they have an agenda, but so does every
non-profit oranisation, including churches/religions, political parties,
etc. (I do object to persons saying that all software in the world
should be GPLed and others - not they - should make that happen.)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@152732 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Maybe one link is that a copyright that precedes a patent on something
>> used in the code may not help you if you don't have the money to fight
>> the claims of the patent holder. In that case, you may have the
>> copyright to your code, but can't use it because it violates a (later
>> filed) patent.
>
> But in that sense copyright is in the same league as any other
> itellectual property right (like artistic rights, trade mark rights,
> etc).

Yes, at least there are similarities.

> Note that in the case you describe it is the patent holder that has to
> sue you, not the other way round. And a proveable copyright (which
> requires publication, otherwise you can't have a copyright!) would be a
> very easy proof of prior art, so I doubt the patent holder would gain
> much by sueing the copyright holder.

Not sure. It's usually not black and white, and once legal speak sets in,
technical issues get to sit on the back seats. Since there may be a few
years between what you wrote that is really the basis of the patent and
what you are doing when you are being sued to stop what you are doing,
things may be a bit more complicated.

Say you wrote (and published) in 2000 some code that implements what in
2003 got patented by someone else. In 2005 you have a business based on
code that includes this, in modified and extended form. You get sued, with
the allegation that your current code violates the patent. Which it
possibly does. So you kind of have to counter-sue to fight their patent,
and prove that your earlier copyright constitutes prior art. And the law
suit of the patent holder doesn't necessarily help you in that (it of
course would if they claimed that your work of 2000 violated their patent
of 2003, but it doesn't as they are only claiming that your work of 2005
violates their patent).

You simply may not have the funds to do that. That's then pretty much end
of story.

Gerhard

2005\12\19@154022 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Malicious is any pigopolist who takes PD code, changes the attributions
>> and deletes the copyrights, burns it into a Flash and ships 100,000
>> routers to all over the world based on that.
>
> If it is fully PD ("anything allowed") code, then what's wrong? And what
> does it matter when "attributions" are removed from the source (which
> no-one will see) or not? They do not end up in the compiled version
> (that is what is flashed) anyway.

The code will be seen by people, if only the future maintainers of the
code. If it's PD code that has a copyright notice to it, why would you
remove it? Or why would you say it's ok to remove it? Isn't removing a
copyright notice not already a breach of copyright law?

Gerhard

2005\12\19@161335 by Wouter van Ooijen

face picon face
> Say you wrote (and published) in 2000 some code that
> implements what in
> 2003 got patented by someone else. In 2005 you have a
> business based on
> code that includes this, in modified and extended form. You
> get sued, with
> the allegation that your current code violates the patent. Which it
> possibly does. So you kind of have to counter-sue

no, you defend in the lawsuit that was brought against you

> to fight their patent,
> and prove that your earlier copyright constitutes prior art.

Copyright is not important here, prior art is.

> And the law
> suit of the patent holder doesn't necessarily help you in that (it of
> course would if they claimed that your work of 2000 violated
> their patent of 2003, but it doesn't as they are only claiming that
your
> work of 2005 violates their patent).

Which you counter by showing your publication. If it is indeed about the
same thing as the patent it is very clearly prior art. Whether you have
copyright on it is not important.

> You simply may not have the funds to do that.

To bring that publication to court? If you don't have the funds to drive
to the courthouse there is probably an anti-patent group that will do
that for you.

I certainly won't deny that there are too many occasions where patent
lawsuits are in effect a battle of attrition (is that the right word?).
That is one of my objections to the patent practice (the others are the
lack of novelty and invention  level test before the patent is granted).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\19@184927 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> And the law suit of the patent holder doesn't necessarily help you in
>> that (it of course would if they claimed that your work of 2000
>> violated their patent of 2003, but it doesn't as they are only claiming
>> that your work of 2005 violates their patent).
>
> Which you counter by showing your publication. If it is indeed about the
> same thing as the patent it is very clearly prior art.

I'm not sure it is always so easy. You can show your code to a judge, but
he won't look at it, as he can't read it. You have to bring in an expert
witness. Which you have to pay (after you have already paid the lawyer).
And so on.

> I certainly won't deny that there are too many occasions where patent
> lawsuits are in effect a battle of attrition (is that the right word?).

Battle of legal budgets, you could also say :)

Exactly my point. And this definitely won't be getting easier with software
patents.

Gerhard

2005\12\19@192158 by Juan Garofalo

flavicon
face

Peter Peres said:

>This prevents you from stealing code from others, and others from stealing your
>code. And that's the point.


       Excuse me. You either 'believe' in 'intelectual property' and
patents, or you don't. Tertium non datur.

>You have to understand that people who have made their code available under the
>GPL did not relinquish any rights on it.

       The same contradiction. Either ideas can be owned or....

       If you don't believe in patents and/or IP, then you must completly
drop the license system enforced at the point of a gun. It doesn't matter if
the 'license' to be enforced is GPL LGPL CNN FBI or whatever.

       Otherwise, all this is talk about 'free' software(meaning it's
linked to freedom) is pure hypocrisy and is, IMO, intended to confuse people.


J.



2005\12\20@020530 by Wouter van Ooijen

face picon face
> And this definitely won't be getting easier with software patents.

So fight them. One way is to make public as much code as possible,
preferssably in a structured way, so prior art can be found later.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@020531 by Wouter van Ooijen

face picon face
> If it's PD code that has a copyright notice to it, why would you
> remove it?

That's not the point.

> Or why would you say it's ok to remove it? Isn't removing a
> copyright notice not already a breach of copyright law?

Not (at least not in my country) if the license allowes relay
'everything', which is the ultimate form of PD, in which case I would
say it is OK because the author allowed it.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@061831 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

> Note that I absolutely agree people have the right to set whatever
> conditions they want for software they make available to others, but if
> you use the GPL don't try to pretend it's for some greater good.  

There are few who do that, and none of those seems to have manifested
himself here. At least I didn't read anything in this thread that was about
"some greater good" -- so far. The first one to bring up this concept were
in fact you (ironically).

> If that's what you're truly after, a less restrictive license would be
> more suitable.  

That depends on what "greater good" someone would be after, doesn't it? If
the believed "greater good" is "all open source software" (whether or not
we agree with that), a less restrictive license wouldn't make sense.

> The GPL is about trying to enforce a social agenda.  

This is a strange reading. The GPL essentially says "I put this source out
for everybody to see, use and modify if I can see, use and modify what they
do with it, under the same terms." Sounds like a fair, reciprocal
agreement. It's not free as in "free lunch", but whoever wants a free lunch
should try being born on a different planet next time.

(I also don't see any infectious qualities in GPL. You can have GPL
software and TCL ("Traditional Commercial License") software on the same
computer all day long, and the GPL licensing terms won't "infect" your TCL
software. Whoever calls the GPL "viral" or "infectious" seems to have an
agenda, because the facts don't match these negatively associated analogies
very well -- so their use is probably for their propaganda effect.)

> That's your call, but it doesn't make you a philanthropist or hero in my
> book.  

Again you're talking about those high values (like the "greater good"
earlier); nobody here (besides you) seems to think that publishing software
has anything to do with being or not a philantropist or hero, or with some
greater good. At least nobody wrote anything to this effect (besides you).

> At least the people trying to make profit from software are honest about
> it.

I'm not sure why you think that people writing GPL software don't want to
make a profit from it. The GPL was in part designed to enable people to
publish source code and still be able to make a profit from it. It seems
that many of the people who write GPL software believe that this is
possible, and write GPL software exactly because of this. (There may be
doubts about whether this is possible at all or, if possible, an effective
way to make profit, but that's a different story.)


It seems to me that you have at least as much a "social agenda", trying to
work against a perfectly legitimate and legal form of software licensing on
a mailing list with some 1000+ readers -- since you don't get paid to do
that (or do you?), you must really have an agenda here, or you do it out of
philantropic reasons for the greater good :)

Gerhard

2005\12\20@063419 by Gerhard Fiedler

picon face
Juan Garofalo wrote:

> If you don't believe in patents and/or IP, then you must completly drop
> the license system enforced at the point of a gun. It doesn't matter if
> the 'license' to be enforced is GPL LGPL CNN FBI or whatever.
>
> Otherwise, all this is talk about 'free' software(meaning it's linked to
> freedom) is pure hypocrisy and is, IMO, intended to confuse people.

Don't you think you're going a bit too far here? This sounds like any
property right would be contradictory to your idea of freedom. Of course
your property limits my freedom of movement and my freedom to use
everything out there, but on the other hand the property rights give me
(and you) the freedom of owning something. Every freedom comes at the cost
of another freedom -- there ain't no free lunch, and there is no total
guaranteed freedom. Every guaranteed freedom is a restriction of some sort.

It's perfectly possible to think that the patent system is royally screwed
up and that patents were never meant to apply to software in the first
place, and that therefore software patents (and to a certain degree patents
in general) are not a good thing (anymore) -- all the while still holding
on to the idea of copyright and maybe thinking that copyleft licenses are a
good thing.

There's no hypocrisy in this, just a bunch of different arguments. Patents
are one thing, copyright is another thing, copyleft is still another
thing...  You can think patents are not good but copyright is, you can
think copyright is good but the idea of copyleft isn't, you can think
patents are a good thing but copyleft (or -right) isn't, you can think that
patents in principle are a good thing but that it's impossible to create a
working realization of the idea...

They are quite different concepts, have different realizations and
different rules governing them and have different effects.

Gerhard

2005\12\20@080905 by olin piclist

face picon face
Gerhard Fiedler wrote:
> There are few who do that, and none of those seems to have manifested
> himself here. At least I didn't read anything in this thread that was
> about "some greater good" -- so far. The first one to bring up this
> concept were in fact you (ironically).

I wasn't talking about anyone here specifically, but certainly there is a
lot of retoric out there about how open source is good for the world, blah,
blah, blah, and how wonderful the people are that make their source
available.  I think that's a bit hypocritical when they slap the GPL on it.

> That depends on what "greater good" someone would be after, doesn't it?
> If the believed "greater good" is "all open source software" (whether
> or not we agree with that), a less restrictive license wouldn't make
> sense.

Yes, that kind of goal is what I call a social agenda: all software should
be open, so I'll do what I can to make others open theirs too.  This is
different from a more laudible goal of getting more and better software out
there accessible to the most people.  Someone willing to give up potential
pay for their work to further that goal does deserve extra recognition in my
opinion, whereast the first goal does not.

> (I also don't see any infectious qualities in GPL. You can have GPL
> software and TCL ("Traditional Commercial License") software on the same
> computer all day long, and the GPL licensing terms won't "infect" your
> TCL software.

By infectious I mean that the GPL license propagates itself to derivative
works.

> Again you're talking about those high values (like the "greater good"
> earlier); nobody here (besides you) seems to think that publishing
> software has anything to do with being or not a philantropist or hero,
> or with some greater good. At least nobody wrote anything to this
> effect (besides you).

A lot of the retoric I saw from Stallman and the Open Software Foundation
(or whatever it's called) was heavy on how wonderful this all was and how
great they were all being for doing this, and there was sortof a cult
following.  Maybe they have toned that down in recent years, I don't know.
While this was all an important part of software history and definitely had
an effect on the world, it's hardly the "free" software you might think it
is just from the propaganda.

I guess what really bothers me is the whole attitude of someone using free
software to make money as being somehow evil or dishonest.  If your aim is
to see more and better software accessible to more people, then this is
exactly what you *want* to have happen.  Even here on this thread people
were calling this "stealing" and said the GPL was in response to getting
"ripped off".  If you put software out there and allow anyone to use it for
free, and someone does, that can't possibly be stealing since you allowed
it.

> I'm not sure why you think that people writing GPL software don't want
> to make a profit from it. The GPL was in part designed to enable people
> to publish source code and still be able to make a profit from it.

Exactly.  And I have no problem with that except the mystique of generosity
and greater good that some proponents of the GPL seem to want you to believe
in.

> It seems to me that you have at least as much a "social agenda", trying
> to work against a perfectly legitimate and legal form of software
> licensing

No, you have completely misunderstood me.  I'm not working against the GPL
at all.  I completely support the right of people to put whatever
restrictions they want on their own personal work product.  My issue is with
the sentiment surrounding these actions, and pointing out that software
released under the GPL is very different from "free".


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\20@105546 by Danny Sauer

flavicon
face
Olin wrote regarding 'Re: [OT] OSS' on Tue, Dec 20 at 07:12:
> A lot of the retoric I saw from Stallman and the Open Software Foundation
> (or whatever it's called) was heavy on how wonderful this all was and how
> great they were all being for doing this, and there was sortof a cult
> following.  Maybe they have toned that down in recent years, I don't know.
> While this was all an important part of software history and definitely had
> an effect on the world, it's hardly the "free" software you might think it
> is just from the propaganda.

There's no toning down as far as I can see (from the perspective of
someone heavily involved in the Linux community).  The goal is quite
clearly that all software should be open, and that anyone who
disagrees is somewhere between misguided and pure evil.

Some of my stuff's GPL because I'd prefer that anyone who makes it
better make those changes available to the other people using the same
software, but the majority is public domain because I just like to see
it used without people getting all worried about what license is on
it, sparking week-long discussions about the problems with various
licenses. :)

--Danny

2005\12\20@113543 by Gerhard Fiedler

picon face
Xiaofan Chen wrote:

> 2) Quoted from libhid page http://libhid.alioth.debian.org/
> "We realise that this is a serious impediment. The GPL is a "viral"
> licence, and you will only be able to use libhid in other GPL projects.
> We would like to change the licence, but libhid uses the MGE UPS SYSTEMS
> HID Parser, which is GPL, and thus we cannot. Our solution is to rewrite
> the HID parser. One of these days. We are also in contact with MGE,
> trying to convince them to loosen their licence. If the licencing issues
> are solved, we are likely to re-release libhid under the Artistic
> Licence."

Here we go again... I don't see anything more "viral" in this license than
in most commercial source code licenses. These usually put some severe
restrictions on what you can do with the code. You are usually not allowed
to publish a competing product based on their code, for example. You are
usually not allowed to publish anything you do with their code in source
form. (How's that for a "closed source virus"?) The only "virus" in this
situation is that the GPL code is there, source code and all, and some
people just can't resist the temptation to use it even though they don't
like the license.

Everybody who wants to use this code in question in a commercial (closed
source or otherwise non-GPL) application could get in contact with the HID
Parser authors and the libhid authors and ask them for a license deal. Or
write their own parser (or pay someone to do it), publish it under a
different license and get the libhid authors to use it and release their
code under that less restrictive license. Sounds to me like a completely
normal situation, like with any budgeting questions that involve 3rd party
licenses when working on a commercial application. This is what many don't
seem to get: the whole point of the GPL license is that the authors don't
give up their right to make money off the code (and to get their share if
somebody else does it -- whatever they come up with as being appropriate).

This whole "viral" thing seems like a brainwash attempt to me, by people
who either have a hidden "agenda" :) or want a free lunch or maybe just
don't get it. There's nothing viral in it; you use that code or you don't.
If you don't use it, you're safe, no virus will attack you. If you think
you want to use it, you know up front that you get into GPL territory and
what that means. Not much use complaining afterwards that this isn't what
you want, and that you've been bitten by the "GPL virus" :)  This here is a
perfect example. Some things are not commercially viable without using GPL
code, but I'd say that's the problem of the business model (or the lack
thereof, or the lack of enough substance in the product to warrant the
effort of implementing it), not the GPL license restrictions.

Gerhard

2005\12\20@114853 by Danny Sauer

flavicon
face
I wrote regarding 'Re: [OT] OSS' on Tue, Dec 20 at 09:58:
> There's no toning down as far as I can see (from the perspective of
> someone heavily involved in the Linux community).

Further, there's a relevant interview with RMS from a couple of days
ago (on Slashdot today, which is where I saw it) illustrating the
motives behind the guy who's behind the GPL.

http://www.zmag.org/content/showarticle.cfm?SectionID=13&ItemID=9350

Much like any radical, he has good ideas taken to a bad extreme.

--Danny, who, in protest, refuses to call it "GNU/Linux"

2005\12\20@120215 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

> I wasn't talking about anyone here specifically, but certainly there is a
> lot of retoric out there about how open source is good for the world

Maybe for the world, or maybe not. But for programmers for sure. There are
countless issues in the workings of Microsoft products that are not well
documented and where even very experienced people can only offer guesses
and "statistical" data from past experience rather than certainty about how
something works. Published sources would make that quite different.

I'm coming from a microcontroller, CP/M, MS-DOS, Windows background in
programming, not from Unix. For me, 3rd party sources were by default not
known. People coming from a Unix background apparently had a completely
different experience back in the 70ies and 80ies (and still now). For them,
sources were by default known. I can appreciate that you can work
completely differently if you have access to all sources of the programs
you work with (whether "free lunch" access or access with somehow
restricted rights, that doesn't matter in this context). From what I
understand, much of the "evil" rhetoric WRT closed source comes from that
point of view. And there are a number of arguments why open source is
better from a programming POV. (Leaving all commercial arguments aside, as
these always have a social agenda behind them, a certain economic model
etc.)

Now if you adopt that POV, you may of course want to further the goal of
having as much open source software in your toolkit as possible. For that,
people must have a way to publish their code without giving all rights to
it away -- so that more professional quality code gets released in source
form. For me, that's not a social agenda, that's a programmer's workstyle
agenda.


> My issue is with the sentiment surrounding these actions, and pointing
> out that software released under the GPL is very different from "free".

Here you seem to agree with Stallman. He makes it very clear that the
"free" in "free software" as he uses it is not "free" as in "free beer"
(using his words :). As I understand it, this is the whole point of the
license, not a side effect. GPL software is not and is not supposed to be
free beer software.

Gerhard

2005\12\20@131915 by olin piclist

face picon face
Gerhard Fiedler wrote:
> Here we go again... I don't see anything more "viral" in this license
> than in most commercial source code licenses. These usually put some
> severe restrictions on what you can do with the code. You are usually
> not allowed to publish a competing product based on their code, for
> example.

That's highly restrictive but not viral.  I think the term viral is use here
to point out how the GPL propagates itself.  "Infectious" would probably
have been a better term.  Unfortunately "viral" can mean a lot of other
things, most of them bad, which is where your objection is coming from.
Replace "viral" with "self-propagating" and it will probably make more
sense.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\20@133422 by Peter

picon face

On Mon, 19 Dec 2005, Juan Garofalo wrote:

> Peter Peres said:
>
>> This prevents you from stealing code from others, and others from stealing your
>> code. And that's the point.
>
> Excuse me. You either 'believe' in 'intelectual property' and patents,
> or you don't. Tertium non datur.

I believe that I hate absolute choices ;-) The world is complicated and
it can be subdivided into parts, and parties, by arbitrary criteria in
many ways. If one would not believe in the 'intellectual property' of
patents, as a GPLd code author, then why put it under the GPL which is a
valid license in the spirit of software patents and/or attribution and
licensing. As to tertium, it gets hard to count the number of options
and real life problems caused by our 'incomplete' model of the world in
any language, but especially in Latin. Maybe that's why judgements in
intellectual property cases appear so odd to 'normal' people like me
(ianal).

Just as an example, consider the Msoft:Lindows lawsuit. Msoft
trademarked a dictionary word, Windows, in conjunction with operating
systems (you can't do that, right ? But they did it anyway - although at
the time windowing graphical computing systems were common already).
Lindows is not a dictionary word and yet Msoft was able to sue them
several times over the 'similarity' of Lindows to a dictionary word.
Imho in this case there is a clear double standards issue. How can one
sue someone else for using a non-dictionary word, for its being similar
to a dictionary word, never mind the context ?! Apparently in the
twisted ways of IP law, you can get away with this ?! Barf !!

>> You have to understand that people who have made their code available
>> under the GPL did not relinquish any rights on it.
>
>The same contradiction. Either ideas can be owned or....

What has this got to do with ideas ? Ideas should not be owned,
embodiments should (and are). There should also be a limit to the
breadth of the 'wide scope' of the lawyertalk in the claims of a patent.

After all, a mammoth bone needle patent from 10,000 years ago could very
well be used by some successors of a really old people to sue modern
loom makers for infringement if there would be no limits.

> If you don't believe in patents and/or IP, then you must completly
> drop the license system enforced at the point of a gun. It doesn't
> matter if the 'license' to be enforced is GPL LGPL CNN FBI or
> whatever.

> Otherwise, all this is talk about 'free' software(meaning it's linked
> to freedom) is pure hypocrisy and is, IMO, intended to confuse people.

The GPL has little to do with ideas, it is a license on an embodiment of
an idea, or embodiment of an algorythm, when used on code. When used on
something else, then again it reflects the belief of the author that a
license is necessary (usually in the form of an Artistic License or
Creative Commons license, an not GPL), therefore he surmises that
licenses do work and are necessary, therefore he does believe that
intellectual property exists and must be regulated (up to what point,
remains to be seen).

Meanwhile the GPL and free software are a way to maintain open source
code as is desired by its authors, in despite of the encroaching on this
and other areas by the ip 'industry'.

As to whether it's good or bad, I say it's bad. The most recent example
I could give would be the RIM (Blackberry makers) lawsuit. RIM is a
Canadian maker of the popular Blackberry handheld wireless email
terminals, i.e. an electronic device manufacturer, fairly close to what
many on this list are doing (or hope to be doing). The challenger is
NTP. NTP won against RIM, practically destroying its business in the USA
and now other 'intellectual property' companies in Europe are also suing
RIM. To understand the enormity of the issue one must read what NTP
really is:

http://www.newsday.com/business/printedition/ny-bzoff184555212dec18,0,4971614.story?coll=ny-business-print

(Scroll down to the paragraph that begins with "The fight casts light on a little-known" ...
if impatient - notice that the page is Ajaxed, you cannot highlight and
copy text from it until you do not turn off their Style (Firefox:View->Page Style->No Style))

If IP law means that such a firm can come out of the woods and claim 6%
of revenues once one starts making money, and successfully sue one for
$450 million, then I think IP in the form of patents on ideas is bad in
its present form and that it makes engineering and IT innovation work as
dangerous as walking in a minefield. Note that NTP holds a patent on an
idea that imho even has copious previous art (see military telex coupled
to short wave radio for copious precedents and previous art dating from
pre-WW2 - the exact nature of NTPs ideas are covered elsewhere on the
web).

As to how deep the mud is, here is a a little sample (patent lawyer's
blog):

http://271patent.blogspot.com/2004_06_01_271patent_archive.html

which has a lot of darkness but also some light (look under "PatentCafe
Intoduces Infringement Litigation Suite to Uncover Prior Art"), more
darkness (under "US standard of patent approval 'down sharply'"), darker
("One click, two click, three click, four"), darkness ("Brits Dumping on
the US Patent System"), grayness ("Good Things Come To Those Who Wait")
and on and on.

ianal (and damn glad I ain't)
Peter

2005\12\20@134034 by Marc Lavallée

flavicon
face
Le 19 Décembre 2005 02:06, Wouter van Ooijen a écrit :
> > A devired work is something different. It's not a PD work.
>
> True. But the effect is still that you can take a PD work at source
> level and create a proprietory derivate of it (the derivate effort can
> be zero). So you *can* turn a PD work into something proprietory. That
> is the main difference between a PD (or almost PD, like BSD) work and a
> GPL work.

The BSD is not "almost PD", it's a licence, it's not like expired copyright,
even if it seems more "PDish" to you. We definitely can't turn something in
the PD into something proprietary; it's not because we can use a PD work in
our own productions that we are allowed to claim that this work is ours and
apply any licence we want. A PD work can't have a license anymore, it's
game over. We can only claim what's original (what we create or we own) in
a derivative work created from a PD work.

--
Marc

2005\12\20@140805 by Peter

picon face


On Tue, 20 Dec 2005, Wouter van Ooijen wrote:

>> And this definitely won't be getting easier with software patents.
>
> So fight them. One way is to make public as much code as possible,
> preferssably in a structured way, so prior art can be found later.

And more importantly imho, in a structured work that offers higher
guarantees of non-ripping to the individual without deep pockets, than
the 'automatic copyright' awarded by publication. GPL is one way, there
are others.

Peter

2005\12\20@141908 by olin piclist

face picon face
Danny Sauer wrote:
> Further, there's a relevant interview with RMS from a couple of days
> ago (on Slashdot today, which is where I saw it) illustrating the
> motives behind the guy who's behind the GPL.
>
> http://www.zmag.org/content/showarticle.cfm?SectionID=13&ItemID=9350
>
> Much like any radical, he has good ideas taken to a bad extreme.

I think it's really funny how he goes on and on about how running
proprietary software is giving up your freedom.  It's a barter exchange like
anything else.  You get to use the software in return for something else.
Part of that something else is an agreement that you will not do something
you might want to do, which is pass the software on to others.

By using GPL source code in your software you give up freedom in exactly the
same way Stallman is so against.  You get to use the source code in return
for doing something you might not want to do, like publishing the source
code to the part that you added.

Again, I have no problem with either of these two agreements.  Software
developers should be free to charge whatever "price" they want for their
software, and others are free to accept that price or not.  My beef is when
people call the first one evil and the second one noble.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\20@170125 by Gerhard Fiedler

picon face
Olin Lathrop wrote:

> Gerhard Fiedler wrote:
>> Here we go again... I don't see anything more "viral" in this license
>> than in most commercial source code licenses. These usually put some
>> severe restrictions on what you can do with the code. You are usually
>> not allowed to publish a competing product based on their code, for
>> example.
>
> That's highly restrictive but not viral.  

If you think the GPL propagation is viral, this is just as viral. If I get
a source license for a 3rd party commercial library and use a somewhat
modified version in my product, in almost all cases I can't publish my
product's source code (which would include the 3rd party's modified code)
-- as a result of the 3rd party commercial license "propagating" into my
product. The prohibition to publish the code is as "viral" as the
requirement to publish it -- the only difference is that the "GPL virus" is
likely to help generate more code based on the original work (because this
is possible), whereas the "commercial virus" causes the immediate death of
the chain (because the work can't be published).

I think here the word "viral" can be and is being used for GPL so often
because there actually can be created derivative works (which all then have
the GPL license), whereas in the closed source scenario this is not
possible at all. But I think having a whole bunch of derivative works
available is always better than not having them... "infected" or not.


> I think the term viral is use here to point out how the GPL propagates
> itself.  

It's not really propagating itself. Somebody decides to create GPL code;
only then the resulting code is GPL code. If I decide to write code that's
not GPL, I simply won't use GPL code. I don't see any self-propagation of
the license.

For me, I can only explain that "viral" complaint with people using GPL
software because it's convenient (and because commercial alternatives are
too expensive) and then claiming that the GPL has "itself" propagated into
their code -- basically complaining that GPL software is not free as in
"free beer" or "free lunch". Doesn't make sense to me at all. Everybody who
has GPL problems has at some point used GPL code. They should have thought
about this better.


> Replace "viral" with "self-propagating" and it will probably make more
> sense.

I hope I could make it clear why I think there's no viral quality, no
infectious quality, and also no "self"-propagating quality. It propagates
to derivative works, that's true, but it does so in a similar way as (just
less restrictive than) most commercial source licenses -- under full
control of the creator.


Think of another aspect. Let's look at one of the basic assumptions of the
capitalist model: the informed buyer. With software, one of the major
quality criteria is the quality of the sources. I guess I don't have to
explain that. When buying a substantial piece of software, source code
quality would be a major criterion, if the sources were accessible. (Of
course not every buyer would examine millions of source code before buying,
but the general quality of a product's sources would be part of the expert
review of that application and therefore a more long-term known fact.)

Without the sources, one essential piece for a functioning software market
is missing. In a way, you could look at the open source movements as
capitalism at work :) -- the system adjusting itself, to create what's
missing for it to work.

Gerhard

2005\12\20@172027 by Neil Cherry

picon face
William Chops Westfield wrote:
> On Dec 19, 2005, at 1:47 PM, Peter wrote:
>
>> Linus Torvalds, Tenenbaum (of Xinu fame) and the BSD people were
>>  started by the closure of access to unix sources afaik.
>>
> "open source" had a long life well before this.  See things like
> DECUS, for instance, or CPMUG.  The careful wording and codifying
> of "social agendas" in to OS licenses probably had a lot to do with
> things like the closure of access to unix sources, though...
>
> BillW

I'd say at minimum that Open Source mentality goes back into
the 1970's. Of course we didn't call it that. I can say this
with first hand knowledge I as I posted my source code to BBS's
back that far. I'm pretty certain that this kind of thinking
went on before that and probably the computer age.

Sorry I've tried to stay out of this, too many 'religious'
folk who only see the world their way. BTW, can you guess
which side of the fence I stand on? Answer is: neither. This
stuff just tools, each tool has it's place. As far as social
agenda's are concerned, well all zealots scare me. Everything
in moderation, even moderation.

--
Linux Home Automation         Neil Cherry       KILLspamncherryKILLspamspamlinuxha.com
http://www.linuxha.com/                         Main site
http://linuxha.blogspot.com/                    My HA Blog
http://home.comcast.net/~ncherry/               Backup site

2005\12\20@183340 by Wouter van Ooijen

face picon face
> > I think the term viral is use here to point out how the GPL
> > propagates itself.  
>
> It's not really propagating itself. Somebody decides to
> create GPL code;
> only then the resulting code is GPL code.

A biological virus does not replicate itself either. It enters its DNA
into a cell, and then it effectively dies (there is nothing left but an
empty sack). Now the cell (of its own decision) starts producing
virusses. I think this perfectly matches the GPL model of propagation.
Of course it's the individula programmer's decision (consious or
unconcious) to use GPLed source and hence to infect his application. If
he does so that is a compliment to the GPL and the quality of GPLed code
in question :)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@183340 by Wouter van Ooijen

face picon face
> Again, I have no problem with either of these two agreements.
Software
> developers should be free to charge whatever "price" they want for
their
> software, and others are free to accept that price or not.  My beef is
when
> people call the first one evil and the second one noble.

I agree that both (GPL and commercial) are asking a price for the use of
their software, but IMHO the GPL price is much less selfish. They have
identified their ideal and are willing to work hard for it, without
direct gain. And even without owing them anything at all you can use
their applications (not their sources!), so they do give something for
nothing! The gain of the commercial types is much more direct. I can't
deny that I am the commercial type most of the time.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@183344 by Wouter van Ooijen

face picon face
> The BSD is not "almost PD", it's a licence, it's not like
> expired copyright, even if it seems more "PDish" to you.

On the scale of PD-ishness BSD is definitely much much more PD than GPL
in the sense that it alloww the user much more.

> We definitely can't turn something in
> the PD into something proprietary; it's not because we can
> use a PD work in our own productions that we are allowed to
> claim that this work is ours and
> apply any licence we want. A PD work can't have a license
> anymore, it's game over.

No. You can change one letter and claim that change as yours. Everyone
who wants to build upon your change is bound by your license decisions.
Contrast this with attempting the same thing with GPLed software.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@183344 by Wouter van Ooijen

face picon face
> Further, there's a relevant interview with RMS from a couple of days
> ago (on Slashdot today, which is where I saw it) illustrating the
> motives behind the guy who's behind the GPL. [snip]
> Much like any radical, he has good ideas taken to a bad extreme.

Whatever else yoy think of Stallman, he has created the (original) GPL
and GCC, which I bothy consider major landmarks in the software history.

> Danny, who, in protest, refuses to call it "GNU/Linux"

You are right, it should be called [em]GNU[/em][small]linux[/small] :)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\20@191952 by Howard Winter

face
flavicon
picon face
Gerhard,

On Mon, 19 Dec 2005 21:45:12 -0200, Gerhard Fiedler wrote:

> Wouter van Ooijen wrote:
>
>...
> > I certainly won't deny that there are too many occasions where patent
> > lawsuits are in effect a battle of attrition (is that the right word?).
>
> Battle of legal budgets, you could also say :)

I think "Attrition" works fine!  :-)

> Exactly my point. And this definitely won't be getting easier with software patents.

Which makes it all the better that the European Parliament voted against them!  Most of the discussions in
here of patents applies to the USA only.

Cheers,



Howard Winter
St.Albans, England


2005\12\20@194929 by Marc Lavallée

flavicon
face
Le 20 Décembre 2005 18:33, Wouter van Ooijen a écrit :
> > The BSD is not "almost PD", it's a licence, it's not like
> > expired copyright, even if it seems more "PDish" to you.
>
> On the scale of PD-ishness BSD is definitely much much more PD than GPL
> in the sense that it alloww the user much more.

Your conception of copyright is flawed. There's no such thing as a license
"more or less" in the Public Domain. What you mean is that the BSD is more
permissive, in a way that pleases you. It's possible to put a work directly
in the Public Domain, if you prefer.

> > We definitely can't turn something in
> > the PD into something proprietary; it's not because we can
> > use a PD work in our own productions that we are allowed to
> > claim that this work is ours and
> > apply any licence we want. A PD work can't have a license
> > anymore, it's game over.
>
> No. You can change one letter and claim that change as yours. Everyone
> who wants to build upon your change is bound by your license decisions.

Right. The change is yours. hat's why changelogs are important. :)

> Contrast this with attempting the same thing with GPLed software.

Well, the GPL is a license!
You know, a license, this sort of a copyright thingy? ;-)

Of course, PD is compatible with the GPL; you can change a letter and use
the GPL to distribute "your new" work. But still,it doesn't invalite the
fact that the original work will forever stay in the Public Domain; it's
just a matter of proving it if required...
--
Marc

2005\12\20@203417 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> It's not really propagating itself. Somebody decides to create GPL code;
>> only then the resulting code is GPL code.
>
> A biological virus does not replicate itself either. It enters its DNA
> into a cell, and then it effectively dies (there is nothing left but an
> empty sack). Now the cell (of its own decision) starts producing
> virusses. I think this perfectly matches the GPL model of propagation.

I don't think so -- you just described it. I've never seen a GPL license
"enter" some code by itself, in the way a virus enters a cell. It's also
not quite common that code reproduces GPL licenses, like the cell
reproduces viruses. (Not even software viruses do that; they may reproduce
themselves but they don't reproduce licenses.)

You see that the presumed analogy between a virus and the GPL is quite
difficult to put in words. Being such a bad analogy, it seems much more a
propaganda trick to create (negative) associations -- and as such it has
been quite successful. (And of course reproduces itself through infected
minds, just like a virus... :)

This GPL licensing process could be seen as similar to passing on one's
values to one's children -- but this analogy doesn't create the desired
negative associations. Of course, this also could be thought of as "viral"
(as part of the values could be that they might want to pass them on to
their children...)

Gerhard

2005\12\20@212751 by Neil Cherry

picon face
Wouter van Ooijen wrote:

> You got the order of events all wrong. There was no PD software as we
> know it know at the moment Stallman created the GPL.

I'm not sure I understand this comment (the 'as we know it part').
There was a PD before Stallman started pushing the GPL (i.e. before
GCC). Compuserve, AOL, Prodigy, BBS's and others had loads of it.
I think you're linking Open Source to Public Domain, which is not
even close. One of the problems with Public Domain is that others
have taken the PD work and change the info and called it their
work. This had happened to me on 2 occasions. One was some
EPROM burner software I wrote, the other for an article about
building an external keyboard for the Atari ST. Neither were
very important to anyone other than me but it changed my view
of PD.

I'm not sure someone has mentioned this yet (been a bit tough
keeping up with the circles on this thread) but
Copyright != License. Unfortunately Copyleft does include
License and Copyright. I had to investigate Open Source Licenses
for the HCS II project. I learned a lot about Copyright and
Licenses (but that doesn't make me an expert or Lawyer ;-).

--
Linux Home Automation         Neil Cherry       RemoveMEncherryTakeThisOuTspamlinuxha.com
http://www.linuxha.com/                         Main site
http://linuxha.blogspot.com/                    My HA Blog
http://home.comcast.net/~ncherry/               Backup site

2005\12\21@015806 by William Chops Westfield

face picon face

On Dec 20, 2005, at 8:31 AM, Gerhard Fiedler wrote:

> I don't see anything more "viral" in this license than
> in most commercial source code licenses.

It's "viral" in that a small amount of GPL code can "infect" a large
amount of proprietary software.  I've worked with quite a bit of
licensed
proprietary code; none of it makes ANY claim on my code that gets
"statically linked" with the licensed code (beyond that it print
a copyright notice for the licensed SW as well.)

I find it somewhere between surprising and sinister that any compiler
would claim to change the license nature of source code just because
that code was linked with standardized library code that supports the
language.  Vendors were widely harassed for wanting to charge royalties
for runtime environments for more complex languages; the GPL is worse
than that.  Apparently the avr gcc libraries have a "special" license
to avoid this.  Good for them.  I wonder which gcc libraries DON'T have
such a modified license...  (and I don't think I should HAVE to wonder!)

BillW

2005\12\21@022328 by Wouter van Ooijen

face picon face
> There was no PD software as we
> know it know at the moment Stallman created the GPL.
>
> I'm not sure I understand this comment (the 'as we know it part').

I think I have to withdraw my statement in the light of the evidence
produced, mylord :)

> One of the problems with Public Domain is that others
> have taken the PD work and change the info and called it their
> work. This had happened to me on 2 occasions.

Indeed. But of course the problem is yours, not theirs: by releasing PD
you had this 'bug' in your license code, which allowed it to operate in
a manner you did not intend.

> I'm not sure someone has mentioned this yet (been a bit tough
> keeping up with the circles on this thread) but
> Copyright != License.

Of course not, why would anyone assume that?

> Unfortunately Copyleft does include License and Copyright.

Copyleft it not a well-defined term (it could mean at the very least
either LGPL or GPL). It is true that the texts used to implement for
instance the GPL include a copyright notice, but AFAIK so do all other
source-level licenses. There must be some right to license out, and that
is (in the sense of source level-licenses) mostly (always?) the
copy-right. With some addition of patent rights, in case of the (L)GPL.

> I had to investigate Open Source Licenses
> for the HCS II project. I learned a lot about Copyright and
> Licenses (but that doesn't make me an expert or Lawyer ;-).

No, in that case you would be on the lawlist. I wonder how the
equivalent of James, Olin and Russel (just to name a few) on that list
behave?

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@022329 by Wouter van Ooijen

face picon face
> It's also
> not quite common that code reproduces GPL licenses, like the cell
> reproduces viruses.

No, the GPL-ed code induces the programmer to produce more GPL-ed code
:)

> You see that the presumed analogy between a virus and the GPL is quite
> difficult to put in words.

not at all

> it seems much more a
> propaganda trick to create (negative) associations -- and as
> such it has been quite successful.

For me it does not hold *negative* associations at all. It just
increases my admiration for the 'programmer' of this virus.

> (And of course reproduces itself
> through infected
> minds, just like a virus... :)

Aha, after all you *are* capable of regognising virus-like properties
when you see them!

> This GPL licensing process could be seen as similar to
> passing on one's
> values to one's children -- but this analogy doesn't create
> the desired
> negative associations.

That's a bit too restricted: the propagation if those values is rather
limited. I think an aggressive form of faith (which request from it
followers that they convince at least 2 other people of the faith in
question) is a better - but still flawed - alternative. (note: no, I do
NOT have any particular faith in mind!)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@022329 by Wouter van Ooijen

face picon face
> Your conception of copyright is flawed. There's no such thing
> as a license "more or less" in the Public Domain.

IMHO there is, but leave that 'battel of words' aside

> What you mean is that the BSD is more permissive,

yes

> in a way that pleases you.

no, at least not more than GPL

> But still,it doesn't invalite the
> fact that the original work will forever stay in the Public
> Domain; it's
> just a matter of proving it if required...

True, but the phrase 'taking something propriatary' in this area does
not mean that the original work somehow leaves the PD, but that you can
change one letter (or even: change nothing), and use it in a product
that is not PD at all. Like Apple did for parts of OS/X. You can do that
with PD and BSD software, but not with GPL software.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@075210 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> It's also not quite common that code reproduces GPL licenses, like the
>> cell reproduces viruses.
>
> No, the GPL-ed code induces the programmer to produce more GPL-ed code
> :)

But only if she thinks that the GPL code is better suited -- including the
license. So if it's better suited, who is to say that this isn't a result
of the GPL license? That the better suited code wouldn't even be around
without GPL? And why should she use something she thinks is not as well
suited (including the license)? Is using the better suited product "viral"?

Or is the problem that so much GPL code is "better suited"? If that's the
case, then maybe one should start to think why that may be -- and whether
this is not actually the result of the GPL.


>> You see that the presumed analogy between a virus and the GPL is quite
>> difficult to put in words.
>
> not at all

Maybe not, but you didn't succeed (nor anybody else, so far).


>> it seems much more a propaganda trick to create (negative) associations
>> -- and as such it has been quite successful.
>
> For me it does not hold *negative* associations at all.

Come on... you can't be so naive. "Virus" is negatively associated. If you
want to know about how terms are generally associated, look at advertising.
You won't see "virus" in a positive context.


>> (And of course reproduces itself through infected minds, just like a
>> virus... :)
>
> Aha, after all you *are* capable of regognising virus-like properties
> when you see them!

There was the ":)". And this situation is different from the GPL situation
-- unless you think the whole /idea/ behind GPL (not the license itself) is
the virus. You get exposed to ideas and they creep into your head, whether
you want that or not, and you have only limited control over what they do
there once you are exposed. That's the viral part. A license, GPL or other,
doesn't just creep into your code. You put it there, if and only if you
want to. That's why this is not quite viral.


>> This GPL licensing process could be seen as similar to passing on one's
>> values to one's children -- but this analogy doesn't create the desired
>> negative associations.
>
> That's a bit too restricted: the propagation if those values is rather
> limited. I think an aggressive form of faith (which request from it
> followers that they convince at least 2 other people of the faith in
> question) is a better - but still flawed - alternative. (note: no, I do
> NOT have any particular faith in mind!)

No GPL code requires anybody to convince anybody else. To stick with your
faith analogy, it's more like saying "You can come and make your lunch in
our kitchen with our supplies, without having to pay anything, but you'll
have to pray with us after lunch. You can also come and pick up as many of
our free ready-made takeout lunches as you like, and you won't even have to
pray with us." Sounds fair to me. And not viral. Don't like their way to
pray? Get your own kitchen and supplies. Or eat their ready-made meals...

Gerhard

2005\12\21@080359 by Gerhard Fiedler

picon face
William ChopsWestfield wrote:

>> I don't see anything more "viral" in this license than in most
>> commercial source code licenses.
>
> It's "viral" in that a small amount of GPL code can "infect" a large
> amount of proprietary software.  I've worked with quite a bit of
> licensed proprietary code; none of it makes ANY claim on my code that
> gets "statically linked" with the licensed code (beyond that it print a
> copyright notice for the licensed SW as well.)

I think you are comparing the proverbial apples and oranges here. GPL is a
source code license, so when comparing it to commercial licenses, you
should compare it to commercial source code licenses, not to binary
licenses. (The use, commercially or otherwise, of the binaries created from
GPL code is AFAIK not restricted in any way. And I don't think that
proprietary programs that call binaries that are created from unaltered GPL
code count as derivative works.)

I assume you have never tried to publish your modifications of the
commercially licensed 3rd party sources, for example. The commercial source
code licenses I've worked with would not allow me to publish my derived
works (derived from that 3rd party licensed source code). That restriction
would continue to be valid for all derived works -- if there were any. But
due to that restriction, there won't be derived works, because the sources
can't be published. Sounds like a severe restriction to me. The same way a
GPL license for an included module can "infect" an application that is
supposed to be closed source, such a commercial license in an included
module can "infect" an application that is supposed to be open source.

So the so-called "viral" quality of the GPL license can only be called
viral because it makes it possible to create derived works. With a
commercial source license, there are no derived works available, not
commercially nor otherwise. Such a commercial source code license is, in
sticking with the analogy, a much more deadly virus: it doesn't only
infect, it makes the infected entity sterile (or dead, depending on how you
apply the analogy) on contact.


> I find it somewhere between surprising and sinister that any compiler
> would claim to change the license nature of source code just because
> that code was linked with standardized library code that supports the
> language.  

WRT the standard compiler libraries, they seem to be usually licensed under
the LGPL or maybe a similar license that even allows static linking -- but
still allows you to publish your modifications to the library, something
that commercial licenses usually don't allow.


> Apparently the avr gcc libraries have a "special" license to
> avoid this.  Good for them.  I wonder which gcc libraries DON'T have
> such a modified license...  (and I don't think I should HAVE to wonder!)

I think you should have to wonder. As long as there is something like
copyright, one has to license code and this implies that you have to wonder
about the license terms -- and should do so before you use the code. In a
way, saying that you shouldn't have to wonder about the license terms is
like saying there shouldn't be a copyright; both are intimately linked: the
existence of copyright implies that you have to license code from the
copyright holder if you want to use it. And the existence of a license
implies that you have to wonder about the license details of the code you
want to use.

This is the dual standard attitude towards GPL that I don't understand.
Whenever I want to use 3rd party source code in a product, the first thing
(after identifying suitable candidates) is to check out their licenses --
to see whether they match my requirements. I don't see why that should be
different with /any/ code. And I don't see commercial source code licenses
being much more permissive or less "viral" than the (L)GPL and similar
copyleft licenses.

Gerhard

2005\12\21@083245 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

> Copyleft it not a well-defined term (it could mean at the very least
> either LGPL or GPL).

I'm not sure whether there are other definitions I don't know about that
muddy the waters, but it seems to be well-defined. Check out
http://www.gnu.org/gnu/thegnuproject.html and scroll down to the section
"Copyleft and the GNU GPL". An excerpt:

'The central idea of copyleft is that we give everyone permission to run
the program, copy the program, modify the program, and distribute modified
versions--but not permission to add restrictions of their own. Thus, the
crucial freedoms that define "free software" are guaranteed to everyone who
has a copy; they become inalienable rights.'

A more general discussion is in the Wikipedia
http://en.wikipedia.org/wiki/Copyleft. Seems to be rather well-defined,
even if you take a larger context than Stallman into account.

Gerhard

2005\12\21@084629 by Danny Sauer

flavicon
face
Wouter wrote regarding 'RE: [OT] OSS' on Tue, Dec 20 at 17:36:
> > Much like any radical, he has good ideas taken to a bad extreme.
>
> Whatever else yoy think of Stallman, he has created the (original) GPL
> and GCC, which I bothy consider major landmarks in the software history.

The software's good.  The blind ideology and total inflexibility's bad.

> > Danny, who, in protest, refuses to call it "GNU/Linux"
>
> You are right, it should be called [em]GNU[/em][small]linux[/small] :)

The systems work pretty much the same if I'm using busybox and uClibc
compiled with Intel's compilers.  The user interface and programs most
users interact with wouldn't be any different on one of the BSDs (I
used to use BSDi for what I do on Linux), on HPUX, etc.

Should it be GNU/FreeBSD just because they use gcc?  Should Sun be
pushing GNU/Solaris, since they ship with the gnu buildchain?  If not,
how much GNU software is required before some zealot wants to tack GNU
on to the *front* of the name of another sucessful project?  Is it the
use of GNU ifconfig and occasionally GNU ps which makes a GNU system
GNU?

I saw the smily, but the concept is still irritating.  Yeah, RMS is a
smart guy, and drove the development of some good stuff.  He's also a
conceited ass if he thinks he can stick his pet project's name in
front of any project's name as soon as that project gets sucessful,
just because it happened to use some of his stuff.  I can't think of
any other group who does that.  I'll call ps GNU/ps if he wants, and I
really prefer the GNU/top variant.  How about that? :)

--Danny

2005\12\21@090317 by Wouter van Ooijen

face picon face
> But only if she thinks that the GPL code is better suited --
> including the
> license. So if it's better suited, who is to say that this
> isn't a result of the GPL license?

did I say so?

> That the better suited code wouldn't even
> be around without GPL?

maybe/probably not (did I say something to the contrary?)

> Is using the better suited product "viral"?

Is reproducing RNA from DNA viral?

> >> You see that the presumed analogy between a virus and the
> GPL is quite
> >> difficult to put in words.
> >
> > not at all
>
> Maybe not, but you didn't succeed (nor anybody else, so far).

In my opinion I did.

>> For me it does not hold *negative* associations at all.
>
> Come on... you can't be so naive. "Virus" is negatively
> associated. If you
> want to know about how terms are generally associated, look
> at advertising.
> You won't see "virus" in a positive context.

I reserve the right to form and have my own opinion, but thank you for
your contribution towards it. Be aware that English is not my native
language, and that 'virus' is a term in many languages. It is quite
normal for me to say that 'an idea spreads like a virus'. There is talk
about using special-purpose viri to cure genetic defects. There are
zillions of virusses that make our life possible by holding the bacteria
population in check.

> -- unless you think the whole /idea/ behind GPL (not the
> license itself) is the virus.

Maybe in another sense, but I don't think that idea is spreading so
fast.

> A license, GPL or other, doesn't just creep into your code. You put it
there, if and
> only if you want to.

Jup. Exactly like an infected cell producing new virusses.

{Quote hidden}

You won't catch me saying that GPL (or any other license) is unfair.

> And not viral. Don't like their way to
> pray? Get your own kitchen and supplies. Or eat their
> ready-made meals...

Hmmmm that sounds somewhat like (a part of) the Hamas. Not incidently
they are very popular in the poorer regions of the middle-east (that is:
their ideas spread).

You hold a very narrow idea about what could be called viral. I hold a
much wider view. Right or wrong (for either view), I don't think these
two opinions will merge as a result of further discussion.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@090317 by Wouter van Ooijen

face picon face
> Such a commercial source code license is, in
> sticking with the analogy, a much more deadly virus: it doesn't only
> infect, it makes the infected entity sterile (or dead,
> depending on how you apply the analogy) on contact.

Viruses that quickly kill their host are the least dangerous ones.

> WRT the standard compiler libraries, they seem to be usually
> licensed under
> the LGPL or maybe a similar license that even allows static
> linking

It allows any form of linking, but when you link statically it requires
you to publice the code (your code!) you link against it. I don't know
any commercial license that imposes any requirement on code you link
with the licensed code.

>> Apparently the avr gcc libraries have a "special" license to
>> avoid this.  Good for them.  I wonder which gcc libraries DON'T have
>> such a modified license...  (and I don't think I should HAVE to
wonder!)
>
> I think you should have to wonder.

Hey, for a change I agree with Gerhard! I can't imagine why you "don't
think I should HAVE to wonder". You use something that is not yours, so
you must stick by the rules the owner imposed on using it. Anything else
would be rediculous.

> This is the dual standard attitude towards GPL that I don't
> understand.
> Whenever I want to use 3rd party source code in a product,
> the first thing
> (after identifying suitable candidates) is to check out their
> licenses --
> to see whether they match my requirements. I don't see why
> that should be
> different with /any/ code.

I agree completely.

> And I don't see commercial source
> code licenses
> being much more permissive or less "viral" than the (L)GPL and similar
> copyleft licenses.

I disagree completely. I don't know of any 'commercial' license that
imposes requirements on the code you link with the licensed code. This
is the unique (and viral) aspect of the GPL and somewhat less ( :) ) the
LGPL.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@093402 by Neil Cherry

picon face
Wouter van Ooijen wrote:
>> There was no PD software as we
>> know it know at the moment Stallman created the GPL.
>>
>> I'm not sure I understand this comment (the 'as we know it part').
>
> I think I have to withdraw my statement in the light of the evidence
> produced, mylord :)

He He. I was just trying to limit the arguement that, today, some folks
view Open Source as PD.

>> One of the problems with Public Domain is that others
>> have taken the PD work and change the info and called it their
>> work. This had happened to me on 2 occasions.
>
> Indeed. But of course the problem is yours, not theirs: by releasing PD
> you had this 'bug' in your license code, which allowed it to operate in
> a manner you did not intend.

I think you release the Copyright to the PD not the license. I'm
not sure how the rules of licensing apply to PD software.

>> I'm not sure someone has mentioned this yet (been a bit tough
>> keeping up with the circles on this thread) but
>> Copyright != License.
>
> Of course not, why would anyone assume that?

Because Copyright and License are easily confused. This whole
discussion has demonstrated that a few times (including my
mistake above :-).

>> Unfortunately Copyleft does include License and Copyright.
>
> Copyleft it not a well-defined term (it could mean at the very least
> either LGPL or GPL). It is true that the texts used to implement for
> instance the GPL include a copyright notice, but AFAIK so do all other
> source-level licenses. There must be some right to license out, and that
> is (in the sense of source level-licenses) mostly (always?) the
> copy-right. With some addition of patent rights, in case of the (L)GPL.

Originally I think Copylefting was seen as a way to 'stick it to
the man'.  I seem to recall that the GNU foundation wanted you to
assign your Copyrights to the foundation and then use their license
(the GPL). I prefer to keep my Copyrights as that gives me a
choice of which way I license my code. Technically this is all
mute (?) because the company I work for owns all my ideas. :-/
At least that's what the agreement said when I signed it. Trust
me that has given me some items to reflect on over the years.
Luckily the company prefers to concern itself with subjects that
are much closer to what I do for a living (communications). I just
hope the new company has similar views.

--
Linux Home Automation         Neil Cherry       spamBeGonencherryspamBeGonespamlinuxha.com
http://www.linuxha.com/                         Main site
http://linuxha.blogspot.com/                    My HA Blog
http://home.comcast.net/~ncherry/               Backup site

2005\12\21@100659 by Wouter van Ooijen

face picon face
> Should it be GNU/FreeBSD just because they use gcc?

That would be one point, but the bigger point is where the source for
all the tools come from. Linux uses the FSF tools, and the Linus kernel.
Both are compiled with the FSF GCC (which is part of the tools too).

> I saw the smily, but the concept is still irritating.  Yeah, RMS is a
> smart guy, and drove the development of some good stuff.  He's also a
> conceited ass if he thinks he can stick his pet project's name in
> front of any project's name as soon as that project gets sucessful,
> just because it happened to use some of his stuff.  I can't think of
> any other group who does that.  I'll call ps GNU/ps if he wants, and I
> really prefer the GNU/top variant.  How about that? :)

You can call anything any name you want (but on this list, only within
James limits!), but likewise RS is free to proclaim that is should be
called what he thinks is right. And on the GNU - Linux argument I think
looking at the weight of the contributions that he is right.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@110118 by William Chops Westfield

face picon face

On Dec 21, 2005, at 4:51 AM, Gerhard Fiedler wrote:

> I think you are comparing the proverbial apples and oranges here. GPL
> is a
> source code license, so when comparing it to commercial licenses, you
> should compare it to commercial source code licenses, not to binary
> licenses. (The use, commercially or otherwise, of the binaries created
> from
> GPL code is AFAIK not restricted in any way.)

Huh?  I'm talking about source licenses.  Distributing a binary
containing
(any amount of) statically-linked GPL code obligates me to also release
SOURCE of my proprietary code.  So for instance, before LGPL (and maybe
thereafter, depending), any proprietary code using standard C libraries
is required to make the source available.  Or rewrite the C libraries
locally, or obtain libraries with a different license...

I appreciate the desire and effectiveness of GPL and related licenses
as applied to preventing public code from being made "private"; I don't
much appreciate their efforts toward making private code public...

BillW

2005\12\21@111019 by Wouter van Ooijen

face picon face
> I think you release the Copyright to the PD not the license. I'm
> not sure how the rules of licensing apply to PD software.

As far as I remember from my Law subjects when you publish some work you
have the copyright. You can't even avoid that. Starting from there: you
have the copyright, so you can grant what you want to whom you want to
grant it. When you grant everything to everyone that has the same effect
as if the copyright was already expired, which is what is referred to as
PD.

But PD means realy free (as in free beer), including the right to use it
in a commercial product and make $$ of it without even saying 'thanks'
to the original author.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@113830 by Wouter van Ooijen

face picon face
> I appreciate the desire and effectiveness of GPL and related licenses
> as applied to preventing public code from being made
> "private"; I don't
> much appreciate their efforts toward making private code public...

There effort is to 'lure' you to make your own code public. Whatever the
viral properties when seen from a distance, it is still *you* who makes
your own code public. No-one forces you to link in any (L)GPL-ed code!

And as always: put your money (or effort) where your mouth is! Stallman
did this, but he is not they only one who can pull this trick. Instead
of the standard (L)GPL-ed GCC library there is newlib, which is not GPL
but more like BSD (actually a collection of licenses for the individual
parts, but none that force you to open up the code you link to it).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2005\12\21@120317 by Neil Cherry

picon face
Wouter van Ooijen wrote:
>> I think you release the Copyright to the PD not the license. I'm
>> not sure how the rules of licensing apply to PD software.
>
> As far as I remember from my Law subjects when you publish some work you
> have the copyright. You can't even avoid that. Starting from there: you
> have the copyright, so you can grant what you want to whom you want to
> grant it. When you grant everything to everyone that has the same effect
> as if the copyright was already expired, which is what is referred to as
> PD.
>
> But PD means realy free (as in free beer), including the right to use it
> in a commercial product and make $$ of it without even saying 'thanks'
> to the original author.

Ah, now you reminded me of why I got upset when I released my 'stuff'
into the PD. I wasn't bothered that they had taken it. What bothered
me was that they changed it to state that they create it. That's
plagiarism and that's what bothered me. I'm not sure how that's
addresses by PD (I guess it is, just not how I'd like it :-) but I no
longer worry about it since I don't release code into the PD anymore.

--
Linux Home Automation         Neil Cherry       TakeThisOuTncherryEraseMEspamspam_OUTlinuxha.com
http://www.linuxha.com/                         Main site
http://linuxha.blogspot.com/                    My HA Blog
http://home.comcast.net/~ncherry/               Backup site

2005\12\21@125419 by Danny Sauer

flavicon
face
Wouter wrote regarding 'RE: [OT] OSS' on Wed, Dec 21 at 09:09:
> You can call anything any name you want (but on this list, only within
> James limits!), but likewise RS is free to proclaim that is should be
> called what he thinks is right. And on the GNU - Linux argument I think
> looking at the weight of the contributions that he is right.

So it should be called Debian, RedHat, SuSE, etc, since the GNU
toolchain and Linux kernel would be worthless without someone to put
them together?  Or it should be called XOrg/GNU/Linux when someone
needs to use a graphical interface - maybe XFree/QT/KDE could run on
top of XOrg/GNU/Linux, since TrollTech would be less of a company
without the initial work of the X consortium, and KDE wouldn't exist
in its current form without QT?  Where does it stop? :)

I dunno, I'm all for giving credit where credit's due (where's Alan
Cox's name in "Linux"?), but that's ultimately something for the
distributions to decide how they want to handle it - most of them just
say Linux.  IMHO, calling Linux "GNU/Linux" just makes it sound like
the whole "system" is a GNU project, and that's misleading.  The
kernel is GPL'd, but it's not a GNU project.  The GNU project's
"operating system" is Horde.  Horde has been in development for years,
and has never been useful for anything more than wasting time /
learning.  Maybe one day, but it doesn't look that way...

So a few years back when the big stink over putting GNU in front of
Linux started (it wasn't a big deal when I started using Linux,
apparently since Linux wasn't real popular) it really seemed more like
RMS trying to get more attention for his pet project; screaming "hey,
look at us!  How come no one ever pays attention to us!"...  That kind
of thing rubs me the wrong way, and I think it really goes against the
Linux community in general, not to mention RMS's supposedly altruistic
goals.  Suddenly, programming for GNU isn't about doing "the right
thing" just because it's the right thing, but doing it because of the
adulation you'll get.  That's just not the Linux community way, and
it's not how RMS presents the GNU foundation.

Now, in the case of Debian GNU/NetBSD, it makes a little more sense to
me, since that's Debian's naming convention (userspace/kernel) and
there needs to be some clarity as to what Debian one's referring to.
When someone says "Linux", though, "everyone" knows that they are
probably referring to the Linux kernel with the GNU userspace.  It's
excessively more than needs to be explicitly said redundantly to spell
out the GNU part.  Of course, that's just my opinion, but since I
stated it in public I guess I should have a reason.  So there you go.
:)

--Danny

2005\12\21@125808 by Byron A Jeff

face picon face
On Mon, Dec 19, 2005 at 09:02:01PM +0800, Xiaofan Chen wrote:
> On 12/19/05, Tad Anhalt <RemoveMEtjaspamTakeThisOuTnetins.net> wrote:
>
> >   Check the license that was distributed with avrgcc.  It's been a
> > while since I had a reason to look into this (and not for avr at any
> > time), but the "standard" libraries probably have a specific exemption
> > allowing you to link them however you want.
>
> I think you mean avr-libc. It is a nice license --> modified BSD license.

I've sat on the sidelines for most of this discussion. Gerhard has
done a fine job arguing against the "viral" connotation. Olin, Gerhard,
and Wouter have all pointed out that there's no gun to any developer's
head to to use any codebase.

I have a problem with "nice license" above. It's a license. It has its
pros and cons just like every other license.

Nice would be from a developer/reseller perspective. One can keep one's
own source closed and use this library. I don't have a problem with that.

However, this type of license is subject to VAR abuse. This occurs when
some VAR takes it upon themselves to modify/improve the library itself
and then resell the updated library without releasing the source. I have
a problem with that, and so I have a problem with licenses that allow
for that to happen.

The LGPL does not allow for this type of abuse. Any modifications to the
library itself are considered GPL and the source must be released. However
simply using the library unmodified via (unfortunately dynamic only) linking
does not impose any license requirements upon the developer generated code.

> I guess it should be the preferred to distribute embedded MCU
> libraries.

I disagree. There's a hole in the Open Source license structure for this
situation. The license needs to balance several elements:

1. The library itself should be open source. Specifically modifications to
the library proper should have redistribution requirements imposed.

2. The developer's code that uses the library should have no license
imposition.

3. (A debatable issue): The end user should be able to utilize
updates/improvements in the library.

So let's see how two common licenses stack up:

BSD

1. No
2. Yes
3. No

Not so nice in my opinion because it skews to far towards the developer.

LGPL

1. Yes.
2. Most of the time. Exceptions governed by...
3. Yes.

Not so nice in my opinion because it skews to far towards the user.

What I'd like to see is a license with...

1. Yes.
2. Yes.
3. No.

Fundamentally this last license is the LGPL with the update requirement
removed.  It has a balanced tradeoff where the developer must published
changes to the library itself. However, in return they have no license
restrictions imposed upon their applications code, regardless of linking
method, and they are relieved of the responsibility of having to provide end
users with firmware updated with the latest library updates. So the library
isn't completely open, but mostly so from the source standpoint, nor
completely closed, because all modifications to the library itself will be
released.

Does such a license exist? This would be a nice license for embedded systems
because everyone would be able to benefit from the library itself without
having to release application code to use it.

BAJ

2005\12\21@125914 by Peter

picon face


> for runtime environments for more complex languages; the GPL is worse
> than that.  Apparently the avr gcc libraries have a "special" license
> to avoid this.  Good for them.  I wonder which gcc libraries DON'T have
> such a modified license...  (and I don't think I should HAVE to wonder!)

That's good because the gcc libraries all have such a derogation, as has
gcc itself. There are a few unexpected exceptions, however. Gnu bison
comes to mind here.

Peter

2005\12\21@130522 by Wouter van Ooijen

face picon face
> What bothered
> me was that they changed it to state that they create it.

Mayeb you intended differently, but by putting your code in PD you
allowed that.

> That's plagiarism and that's what bothered me.

Plagiarism is AFAIK something entirely different: it is claiming that
you wrote something. A common vice in the academic world. But quite
different from claiming that you wn (the rights to) something.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\21@132846 by Byron A Jeff

face picon face
On Mon, Dec 19, 2005 at 10:17:58AM +0100, Wouter van Ooijen wrote:
> > Darn, I was going off of their original plan which was to use GPL
> > licenses or similar.  But now the site isn't responding from here and
> > you have me curious if they were able to stick with it.
>
> The problem with GPL + hardware is that GPL talks about things like
> combining into an application, which is more or less well-defined when
> talking about software. But how would you interpret that when deling
> with hardware? That is a very important question, because it is the
> boundary of the GPL viral/contaminating effect. With software the
> boundary is the linked application. But were would the boundary be with
> hardware? Chip level? PCB level? Apparatus level?

I'll take a stab at it.

First off GPL probably doesn't fit unless you are releasing a complete
solution. In that case then all modiciations to the complete solution
would need to be published. However if you're talking about a component
architecture, then an LGPL level license is more appropriate. However, as
I stated in my other post it has its issues. The LGPL has three basic
tenets:

1. Changes to the LGPL component itself is treated as GPL. As such
modifications to LGPL components must be published.

2. Developer generated components are generally not subject to license
restriction. So the developer can license their component any way they
like.  In addition the combined work can be licensed as the developer sees
fit. There is one exception...

3. Since the LGPL component is considered to be free, the end user
retains the right to update/improve the LGPL component even when that
component is embedded in a complete solution with other non LGPL
components. In addition developers who use the LGPL component must
provide facilities for the end user to update/impove the LGPL component.
This can be up to and including the publishing of the developers components
if they are required for this update/improvement to occur.

So take for an example a LGPL ASIC that's integrated into a larger board.
As an LGPL component, any changes made to the VHDL of the ASIC would be
need to be published. However, none of the other components would necessarily
have to be published. On the other hand, the developer still must provide
facilities (JTAG interface for example), so that the end user can update
the ASIC.

Personally I have a problem with the third tenet. While I understand that it
keeps the LGPL component free from a principle standpoint, there's just too
much baggage associated with it. A streamlined license that keeps the first
two tenets and ditches the third would be a good balance. End users would no
be able to update and developers would be able to keep their code to
themselves while having to publish changes to the LGPL component.

BAJ

2005\12\21@135703 by Marc Lavallée

flavicon
face
Le 21 Décembre 2005 11:10, Wouter van Ooijen a écrit :
> > I think you release the Copyright to the PD not the license. I'm
> > not sure how the rules of licensing apply to PD software.
>
> As far as I remember from my Law subjects when you publish some work you
> have the copyright. You can't even avoid that.

May I suggest this reading?
http://en.wikipedia.org/wiki/Public_domain
(Yes, it's a good enough reference...)

What you mention was not always the case, at least in the USA; before 1989,
works without a copyright notice were automatically put in the Public
Domain.

Some works can't be copyrighted, like baseball statistics. Abstract ideas
can't be copyrighted either (althought many are wrongly patented).

It's also mentionned that the  USA allows "public domain shareware". Odd...

An excerpt:
"With regards to patents on the other hand, publishing the details of an
invention before applying for a patent will generally place an invention in
the public domain and (in theory) prevent its subsequent patenting by
others. There is an exception to this, however: in U.S. (not European) law,
an inventor may file a patent claim up to one year after publishing it (but
not, of course, if someone else published it first)."

Trade secrets are forever protected, except if they are disclosed they enter
the public domain.

Trademarks are renewable unless they become generic terms like aspirin.

Here's the algorithm of Public Domain in the US:
http://www.bromsun.com/practices/copyright-portfolio-development/flowchart.htm

Interesting, isn't it?
--
Marc

2005\12\21@141414 by Peter

picon face


On Wed, 21 Dec 2005, William Chops Westfield wrote:

{Quote hidden}

But the libc is specifically under the LGPL, not the GPL, and what you
need to provide is an object module which the customer can relink is
needed, *not* source. I.e. the application.o file that is built during
your compilation, before linking it with ld. F.ex. apropos floating
point, /usr/include/math.h specifically says it is LGPL.

I just gave an example of an item that is not under the LGPL: GNU bison.

Peter

2005\12\21@143401 by Byron A Jeff

face picon face
On Wed, Dec 21, 2005 at 12:33:38AM +0100, Wouter van Ooijen wrote:
> > The BSD is not "almost PD", it's a licence, it's not like
> > expired copyright, even if it seems more "PDish" to you.
>
> On the scale of PD-ishness BSD is definitely much much more PD than GPL
> in the sense that it alloww the user much more.

I would rephrase it as the restricts the user much less.

{Quote hidden}

Bingo. PD is only appropriate when the author absolutely does not care
what is done with the resulting work.

BAJ

2005\12\21@154505 by Byron A Jeff

face picon face
On Sun, Dec 18, 2005 at 08:11:34PM -0500, Olin Lathrop wrote:
> Gerhard Fiedler wrote:
> >Here's where you seem to fall short. How about open source software? I'm
> >pretty sure many developers publish their code as GPL because they want
> >to see more open source code, and they are willing to contribute to
> >that goal -- under the condition that whoever uses their code
> >contributes also.
>
> But that's exactly the social agenda I was talking about.

I think there's a middle ground. I find I can get a direct benefit
releasing code because then others can help work on it. It's not
profit based. Nor is it socially motivated. It's based simply
on the fact that I'd like to see something done without having to
pay in its entirety in money or sweat.

BAJ

2005\12\21@183505 by Gerhard Fiedler

picon face
William ChopsWestfield wrote:

> Huh?  I'm talking about source licenses.  

That wasn't completely clear to me.

> Distributing a binary containing (any amount of) statically-linked GPL
> code obligates me to also release SOURCE of my proprietary code.  

In the same way as pretty much any statically linked commercial source code
impedes you to release the code of the application. Where's the difference?
Why is one called "viral" and the other not?

Gerhard

2005\12\21@184058 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

> It allows any form of linking, but when you link statically it requires
> you to publice the code (your code!) you link against it. I don't know
> any commercial license that imposes any requirement on code you link
> with the licensed code.

For one, you can't publish your application's source code (that includes a
derivation of the commercially licensed code). For me, that is almost an
exact mirror image of the GPL requirement to release the code.

Gerhard

2005\12\22@080315 by Howard Winter

face
flavicon
picon face
Wouter,

On Wed, 21 Dec 2005 17:10:17 +0100, Wouter van Ooijen wrote:

> > I think you release the Copyright to the PD not the license. I'm
> > not sure how the rules of licensing apply to PD software.
>
> As far as I remember from my Law subjects when you publish some work you
> have the copyright. You can't even avoid that.

Yes, that's my understanding too - you can't "Copyright" something, it has copyright intrinsically as soon as
it's created (it doesn't even have to be published).

Licences aren't necessarily a part of copyright at all.

> Starting from there: you
> have the copyright, so you can grant what you want to whom you want to
> grant it. When you grant everything to everyone that has the same effect
> as if the copyright was already expired, which is what is referred to as
> PD.

Yes, you have the copyright, and you can do with it as you see fit, including giving away all your rights.

> But PD means realy free (as in free beer), including the right to use it
> in a commercial product and make $$ of it without even saying 'thanks'
> to the original author.

Yes, when something is "placed in the Public Domain" it means that anyone can do anything with it, including
using it commercially without reference to the originator.  I can make a copy of a picture of a bison on a
cave wall without permission from Ug of the Cave, because even if it had existed, his copyright in it expired
a long time ago, and PD is just like that.  But you can't say "I place this in the Public Domain as long as
anyone uses it credits me as the creator" - PD is "without emcumbrance" - if you want to impose conditions it
isn't PD.

If you want to do anything less than placing something in the Public Domain, then you need to devise a
licence, which is a contract between you and the user, and to which they have to agree before they do
something covered by your copyright.  But licences are a convenient shortcut to contract negotiations, they
aren't an intrinsic part of copyright - nobody can do *anything* with your work unless you allow it, and you
don't have to do so.  There is no "use it or lose it" aspect to copyright, the way there is with patents.

Cheers,

Howard Winter
St.Albans, England


2005\12\22@165350 by Wouter van Ooijen

face picon face
> > (apply (L)GPL to hardware)

> I'll take a stab at it.

I can follow your reasoning and I agree with you in what you would like,
but I think the result is un-GPL-ish because it is not viral (duck
-ough-) sorry, does not contaminate (and, as you pointed out, it does
not preseve the right to modify/update).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu



2005\12\22@165350 by Wouter van Ooijen

face picon face
> Does such a license exist? This would be a nice license for
> embedded systems
> because everyone would be able to benefit from the library
> itself without
> having to release application code to use it.

BAJ, did you read the license I proposed? Repeated below, comments are
welcome.

Note that with this license when you use a modified version of the
library in your product, you are obliged to publish your modifications.

Wouter van Ooijen

==================================================================

Embedded Free License version 1

This work is free, in the sense that you can redistribute it and/or
modify it under the terms of the GNU General Public License
version 2 as published by the Free Software Foundation.

As a special exception to the GNU General Public License version 2:
if other works instantiate templates or use macros or inline functions
from this work, or you compile this work and link it with other works
to produce a resulting work based on this work, this does not by itself
cause the resulting work to be covered by the GNU General Public
License.
However the source code for this work must still be made available in
accordance with section (3) of the GNU General Public License,
under this Embedded Free License.

This exception does not invalidate any other reasons why a work based on

this file might be covered by the GNU General Public License.

You should have received a copy of the GNU General Public License
version 2 along with this work; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301
USA.

This work is distributed in the hope that it will be useful, but WITHOUT

ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
version 2 for more details.

This Embedded Free License is (c) Wouter van Ooijen 2005
(wouterEraseMEspam.....voti.nl).
Unmodified use is allowed. Mofidied use is allowed, but use a different
name.

2005\12\22@170451 by Wouter van Ooijen

face picon face
> What you mention was not always the case, at least in the
> USA; before 1989,
> works without a copyright notice were automatically put in the Public
> Domain.

Yeah, but after that date you conformed to the rest of the world :)

> Some works can't be copyrighted, like baseball statistics.

At least in my country they can, but what you copyright in that case is
what you 'add' to the world: you select which statistics to publish, and
that selection bears copyright, not the figures themselves.

> Abstract ideas can't be copyrighted either (althought many are wrongly
patented).

Sorry, you don't seem to understand copyright. It is about the text
itself, not what the text means. If I describe something and have
copyright on that text, you can describe the same thing *in your own
words* and have a completly independent copyright on your text.

> Interesting, isn't it?

Yes, I thought so when I choose law classes at the university way back
when :)

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\22@170642 by Wouter van Ooijen

face picon face
> > It allows any form of linking, but when you link statically
> it requires
> > you to publice the code (your code!) you link against it. I
> don't know
> > any commercial license that imposes any requirement on code you link
> > with the licensed code.
>
> For one, you can't publish your application's source code
> (that includes a
> derivation of the commercially licensed code). For me, that
> is almost an
> exact mirror image of the GPL requirement to release the code.

? you can relase what you have written!

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\22@190913 by Nate Duehr

face
flavicon
face
Wouter van Ooijen wrote:

> Embedded Free License version 1

Wouter,

Side-note here:

Any place you reference the GPL, you need to reference the VERSION
number of the version of the GPL you're referring to.

The GPL has changed/morphed over time, and future changes could
"de-reference" your references to particular sections, etc.

No huge deal, just a point I thought of, reading this.

(And it pretty effectively demonstrates why GOTO's are bad in code, too!
 LOL!)

Nate

2005\12\22@195500 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>>> It allows any form of linking, but when you link statically it requires
>>> you to publice the code (your code!) you link against it. I don't know
>>> any commercial license that imposes any requirement on code you link
>>> with the licensed code.
>>
>> For one, you can't publish your application's source code (that
>> includes a derivation of the commercially licensed code). For me, that
>> is almost an exact mirror image of the GPL requirement to release the
>> code.
>
> ? you can relase what you have written!

No. I usually license source code because I want to modify it. Source code
licenses are almost always more expensive than binary licenses of products
that provide similar functionality. That expense can only be justified if
it is necessary -- and it's only necessary if I want to modify the licensed
code for my application. Usually.

In almost all cases of commercial source code licenses, you won't be
allowed to publish those modified, licensed sources.

In addition, even if you use unaltered commercial source code in your
application, you can publish the parts that you have written, but not the
complete application. That's often not really helpful -- at least not if
you want to publish it so that it can be compiled and run. (Which is
usually the purpose of publishing sources.)

Gerhard

2005\12\22@212725 by Byron A Jeff

face picon face
On Thu, Dec 22, 2005 at 10:53:47PM +0100, Wouter van Ooijen wrote:
> > Does such a license exist? This would be a nice license for
> > embedded systems
> > because everyone would be able to benefit from the library
> > itself without
> > having to release application code to use it.
>
> BAJ, did you read the license I proposed? Repeated below, comments are
> welcome.

Nope. Will do now.

>
> Note that with this license when you use a modified version of the
> library in your product, you are obliged to publish your modifications.

to the library I presume.

{Quote hidden}

Hmm. I don't think the GPL as a base fits. It would seem that the LGPL would
be a better starting base as the LGPL doesn't generally attempt to extend
itself to combined works.

Also I would suggest using the term combined instead of "instantiate or link".

License writing is difficult. I'm not sure that the above license clearly
delineates the difference between the EFL code and the code combined to it.

What holes would exist if we took the LGPL and struck section 6, which is the
"recombine" requirement? How would it differ from the proposed EFL above?

Just to reiterate, the basic ideas behind this license are:

1. Code under the proposed EFL is essentially GPL code.

2. Code combined with EFL code via compilation or static linking are
not subject to the license. Also executables of such combinations are
not subject to the license.

3. The end user does not get the right to recombine updated EFL code with
non EFL code.

BAJ

2005\12\22@213700 by Byron A Jeff

face picon face
On Thu, Dec 22, 2005 at 10:53:47PM +0100, Wouter van Ooijen wrote:
> > > (apply (L)GPL to hardware)
>
> > I'll take a stab at it.
>
> I can follow your reasoning and I agree with you in what you would like,
> but I think the result is un-GPL-ish because it is not viral (duck
> -ough-) sorry, does not contaminate (and, as you pointed out, it does
> not preseve the right to modify/update).

The LGPL only "contaminates" only for the recombine requirement to the end
user. Without it there is no contamination aspect at all to code outside of
the LGPL code.

BAJ

2005\12\22@221041 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> However that reduction in cost was gained by utilizing community work
>> without making a direct contribution back to the community.
>
> But that "community work" was PD, which means "everything allowed".

Butting in... If I understand this correctly, you both are correct. People
put "in good faith" stuff in the PD without looking too closely at the
legal side. That's why they later may have felt taken advantage of (their
good faith has been misused), and that's why this may have been legal (good
faith is not enough).

> If you want to interpret PD as "everything allowed - well not everything,
> you must ...", then which term is left for "realy realy everything
> allowed"?

No, "everything allowed" remains "everything allowed". But the ones who
felt taken advantage of probably now don't put anything into PD, they
license it with GPL or something else -- no matter whether a copyleft or a
copyright based license, but a license.

Gerhard

2005\12\22@223210 by Tad Anhalt

picon face
Wouter van Ooijen wrote:
> It allows any form of linking, but when you link statically it requires
> you to publice the code (your code!) you link against it. I don't know
> any commercial license that imposes any requirement on code you link
> with the licensed code.

  _My understanding_ of the situation is that the act of linking itself
is what LGPL allows.  Static/dynamic isn't the issue except that there
must be a way provided to re-link with (a) different version(s) of the
LGPL library(s).

  Dynamic linking allows the end-user to replace the shared object
(so/dll/whatever) directly and therefore meets this requirement by
default.  Static linking does not meet the requirement by default and in
this case, the developer has to provide this functionality.  That may be
as simple as providing the (proprietary) object file separately in
addition to the linked version or even just the object file itself with
the assumption that it will be linked by the party you are distributing
the (still proprietary) compiled code to.

  Libraries that are under GPL don't allow any linkage at all (for
proprietary code) and, frankly, the license doesn't try to hide that
fact.  In fact, the verbiage is surprisingly blunt on that point.

  The extra wrinkle with glibc is that it specifically removes the
requirement for re-linking support.  glibc++ has a different exception
entirely that is AFAICT supposed to provide the same results when the
situation is complicated with lots of inlined code (eg: templates).

  Don't take my word for it though.  I've probably missed something
critical and should be roasted over high heat. YMMV, etc, etc.

>>>Apparently the avr gcc libraries have a "special" license to
>>>avoid this.  Good for them.  I wonder which gcc libraries DON'T have
>>>such a modified license...  (and I don't think I should HAVE to
> wonder!)

  I don't know who wrote this, the attributions were lost but...
Seriously, you don't have to wonder.  Look at the license declarations
in the header files.  Go read the license.  The FSF licenses are the
first hit on google and the (very strongly worded) FAQ is linked from there.

  To more directly answer the indirect question posed, there doesn't
appear to be anything distributed with gcc that would require you to
distribute your source code just through the act of compiling it using
the included tools.  There is actually a FAQ entry that states the same
thing on the FSF page.

  Going one step further, I don't recall ever seeing a GPLed library
that didn't have huge warnings all over the place letting you know not
to link it into proprietary code so it's not clear to me where the
problem comes in ?

  My comments on this thread are solely from the perspective of a
developer attempting to satisfy standard professional due-diligence into
licensing issues.  Please don't make assumptions based on what I've said
here about my position on what you should or shouldn't do with your
code, personal feelings on the FSF, etc.

  In other words, it looks like this thread is dying and I really don't
want to unnecessarily prolong it.  I would be very interested in
reasoned critiques of where I've gone wrong in my interpretation though.

Tad Anhalt

2005\12\22@225203 by Neil Cherry

picon face
Wouter van Ooijen wrote:
>> What bothered
>> me was that they changed it to state that they create it.
>
> Mayeb you intended differently, but by putting your code in PD you
> allowed that.

That's what I ended up learning. After that I came up with a few
'Licenses' which basically said, you can use it for free (if not
used in a commercial product) and give me credit. At least up
until the GPL came out.

>> That's plagiarism and that's what bothered me.
>
> Plagiarism is AFAIK something entirely different: it is claiming that
> you wrote something. A common vice in the academic world. But quite
> different from claiming that you wn (the rights to) something.

Well in my mind I thought I did. I considered it my work. Of course
placing it in the public domain wasn't a good idea. Live and
learn.

--
Linux Home Automation         Neil Cherry       RemoveMEncherryEraseMEspamEraseMElinuxha.com
http://www.linuxha.com/                         Main site
http://linuxha.blogspot.com/                    My HA Blog
http://home.comcast.net/~ncherry/               Backup site

2005\12\22@233922 by Xiaofan Chen

face picon face
On 12/23/05, Tad Anhalt <RemoveMEtjaspam_OUTspamKILLspamnetins.net> wrote:
> Wouter van Ooijen wrote:
> > It allows any form of linking, but when you link statically it requires
> > you to publice the code (your code!) you link against it. I don't know
> > any commercial license that imposes any requirement on code you link
> > with the licensed code.
>
>   _My understanding_ of the situation is that the act of linking itself
> is what LGPL allows.  Static/dynamic isn't the issue except that there
> must be a way provided to re-link with (a) different version(s) of the
> LGPL library(s).

I am more confused than before about LGPL and GPL. ;-(

> >>>Apparently the avr gcc libraries have a "special" license to
> >>>avoid this.  Good for them.  I wonder which gcc libraries DON'T have
> >>>such a modified license...  (and I don't think I should HAVE to
> > wonder!)
>
>   I don't know who wrote this, the attributions were lost but...
> Seriously, you don't have to wonder.  Look at the license declarations
> in the header files.  Go read the license.  The FSF licenses are the
> first hit on google and the (very strongly worded) FAQ is linked from there.

Please read the license here:
http://www.nongnu.org/avr-libc/LICENSE.txt
avr-libc is under a modified BSD license.

They asked all the indivisual people to relicense their code
with this license last year (some of the components were
originally under GPL/LGPL).

I just read one more interesting story of originally GPLed software
changing to a different license and what would this cause.

www.linuxtoday.com/security/2005122202226OSLL
LinuxInsider: Nessus 3.0: The End of the Age of
Open-Source Innocence?

Quoted from the story:
"This announcement is testament to the fact that though
single projects like Nessus may need make dramatic shifts
in order to secure a viable future, open source overall is alive
and well; continuing to gather more and more support."

That is good to know.

Regards,
Xiaofan

2005\12\23@020222 by William Chops Westfield

face picon face

On Dec 22, 2005, at 7:38 PM, Tad Anhalt wrote:

>   _My understanding_ of the situation is that the act of linking itself
> is what LGPL allows.  Static/dynamic isn't the issue except that there
> must be a way provided to re-link with (a) different version(s) of the
> LGPL library(s).
>
Which of course wrecks havoc with any embedded sort of application.
"Here's the .o module of our proprietary code that can be linked
with this list of LGPL libraries.  And there's your box with its
OTP soldered in chip containing the old code.  Have fun!"

BillW

2005\12\23@021545 by Wouter van Ooijen

face picon face
> Yes, that's my understanding too - you can't "Copyright"
> something, it has copyright intrinsically as soon as
> it's created (it doesn't even have to be published).

AFAIK publication is necesarry.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@021545 by Wouter van Ooijen

face picon face
> Well in my mind I thought I did. I considered it my work. Of course
> placing it in the public domain wasn't a good idea. Live and
> learn.

It is realy just like programming (or designing electrical circuits):
you do something wrong and you get burned (or crashed or zapped).

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@065719 by Gerhard Fiedler

picon face
Xiaofan Chen wrote:

> I just read one more interesting story of originally GPLed software
> changing to a different license and what would this cause.
>
> www.linuxtoday.com/security/2005122202226OSLL
> LinuxInsider: Nessus 3.0: The End of the Age of
> Open-Source Innocence?

This seems to show that there is not much of a virus in GPL :)

If they can do that, they developed their application with care towards the
licenses of the code they used in it. What anybody should do...

Gerhard

2005\12\23@072040 by Howard Winter

face
flavicon
picon face
Wouter,

On Fri, 23 Dec 2005 08:15:38 +0100, Wouter van Ooijen wrote:

> > Yes, that's my understanding too - you can't "Copyright"
> > something, it has copyright intrinsically as soon as
> > it's created (it doesn't even have to be published).
>
> AFAIK publication is necesarry.

AFAIK, not in the UK - I believe the critical phrase is "rendered into tangible form" - writing, drawing,
painting, typing (not sure about typing into a computer, but I imagine this is included these days),
audio/video recording, photographing, carving, whatever.  So if you stand on a street corner and say a poem,
you don't have copyright unless you audio-record it or write it down.  

As an example, someone could break into your place and make copies of your paintings (which you did for
yourself and which have never left the house) and sell the copies.  If publication was necessary, they would
be guilty of breaking and entering, but not of copyright violation.  If they passed on the copies to a dealer
who sold them, you could not stop the dealer doing so.  I don't think this is the case.  Remember that in the
case of patents, publication prohibits a patent being granted (outside the USA anyway).  If you wanted
copyright and patent on something (assuming it is possible for the item concerned) it would be a very
difficult juggling act if publication was necessary for one, but banned for the other!

Cheers,


Howard Winter
St.Albans, England


2005\12\23@073119 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Abstract ideas can't be copyrighted either (althought many are wrongly
>> patented).
>
> Sorry, you don't seem to understand copyright. It is about the text
> itself, not what the text means. If I describe something and have
> copyright on that text, you can describe the same thing *in your own
> words* and have a completly independent copyright on your text.

Isn't that the exact same thing Marc wrote? Abstract ideas can't be
copyrighted, only a specific formulation of the ideas...

So why would you say that this shows that he doesn't seem to understand
copyright?!?

Gerhard

2005\12\23@075343 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Yes, that's my understanding too - you can't "Copyright"
>> something, it has copyright intrinsically as soon as
>> it's created (it doesn't even have to be published).
>
> AFAIK publication is necesarry.

Maybe the question is what "publication" means in this context.

Let's say I write up something, put it in an envelope, seal that envelope
and put it in a lock box to which I don't have a key, and make sure that I
have a definite proof (through the lock box owner, by a sufficiently
"definite" procedure) that I put that specific work in the lock box on
2005-12-23.

Can I later claim that I "published" this work for copyright uses? I'm not
sure. I think the idea is that if you want to invoke copyright protection,
you need to prove that somebody copied your work. Without it being
"published", it can't be "copied". But in this sense, "published" seems to
mean only that the work existed, potentially accessible to somebody. In the
example above, my work surely didn't get "published" in the general sense
of the word. But if somebody claims to have a year later created something
that is beyond reasonable doubt an exact copy of my work, I think I may
have copyright protection -- if there was a possibility that he could have
gotten access. I think that's what "publication" might mean in this
context: the (even remote) possibility of access to the work. Without that,
there's probably no possibility of a copyright.

I also think that I could use it as prior art in later patent disputes
(which of course is a different question than copyright).

Gerhard

2005\12\23@082718 by Howard Winter

face
flavicon
picon face
On Fri, 23 Dec 2005 10:30:56 -0200, Gerhard Fiedler wrote:

> Wouter van Ooijen wrote:
>
> >> Abstract ideas can't be copyrighted either (althought many are wrongly
> >> patented).
> >
> > Sorry, you don't seem to understand copyright. It is about the text
> > itself, not what the text means. If I describe something and have
> > copyright on that text, you can describe the same thing *in your own
> > words* and have a completly independent copyright on your text.
>
> Isn't that the exact same thing Marc wrote? Abstract ideas can't be
> copyrighted, only a specific formulation of the ideas...
>
> So why would you say that this shows that he doesn't seem to understand
> copyright?!?

Well actually *no* ideas can be copyright, abstract or otherwise.  It is the picture / words / sounds / ...
themselves that have copyright, not the thing they represent.  A nonsense poem, for example, is copyright, as
is a drawing of a machine.  But if the drawer has only drawn and not made the machine, the machine itself
isn't covered - as long as you don't reproduce the drawing, but you build the machine represented in it, you
have not broken the drawer's copyright.  

Of course, if the drawing was part of a patent application then you have probably broken the patent, but that
is an entirely seperate issue.

Cheers,


Howard Winter
St.Albans, England


2005\12\23@084041 by Wouter van Ooijen

face picon face
> Let's say I write up something, put it in an envelope, seal
> that envelope
> and put it in a lock box to which I don't have a key, and
> make sure that I
> have a definite proof (through the lock box owner, by a sufficiently
> "definite" procedure) that I put that specific work in the lock box on
> 2005-12-23.
>
> Can I later claim that I "published" this work for copyright
> uses?

AFAIK not, at least not in my country. Your proof might be handy in some
other situations (like claiming that the ideas in your writing were
yours at that date, hence not developed lateron in some employee
situation), but no-publication == no-copyright established.

> you need to prove that somebody copied your work.

AFIK you need to prove that someone *could* have copied your work,
because it was publicly available. Whether someone actually did this is
not important, although it might be difficult to prove your publication
unless some independent party recorded it :)

> I also think that I could use it as prior art in later patent disputes
> (which of course is a different question than copyright).

Again, no. Prior art is to be interpreted as prior art in the common
(accessible) domain ("as accessible or know to a crafstman in the
applicable field").

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@085626 by Wouter van Ooijen

face picon face
> Any place you reference the GPL, you need to reference the VERSION
> number of the version of the GPL you're referring to.

done

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@090222 by Wouter van Ooijen

face picon face
>> Isn't that the exact same thing Marc wrote? Abstract ideas can't be
>> copyrighted, only a specific formulation of the ideas...
>
> So why would you say that this shows that he doesn't seem
> to understand copyright?!?

Because he seemed surprised by this. It is the very essence of
copyright, so anyone who finds this surprising has IMHO somehow not
grasped it.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@101826 by Byron A Jeff

face picon face
On Thu, Dec 22, 2005 at 09:38:08PM -0600, Tad Anhalt wrote:
> Wouter van Ooijen wrote:
> > It allows any form of linking, but when you link statically it requires
> > you to publice the code (your code!) you link against it. I don't know
> > any commercial license that imposes any requirement on code you link
> > with the licensed code.
>
>    _My understanding_ of the situation is that the act of linking itself
> is what LGPL allows.  Static/dynamic isn't the issue except that there
> must be a way provided to re-link with (a) different version(s) of the
> LGPL library(s).

Correct. The arugment is that the relink requirement allows the library
to remain "free" and updatable by the end user.

Personally I find it cumbersome to follow for embedded systems work.

>    Dynamic linking allows the end-user to replace the shared object
> (so/dll/whatever) directly and therefore meets this requirement by
> default.

Bingo.

> Static linking does not meet the requirement by default and in
> this case, the developer has to provide this functionality.  That may be
> as simple as providing the (proprietary) object file separately in
> addition to the linked version or even just the object file itself with
> the assumption that it will be linked by the party you are distributing
> the (still proprietary) compiled code to.

All on point.

One missing point is that it's the devlopers responsibility to provide the
end user whatever tools required to make the relink, including propietary
compilers and linkers.

As I said it's too cumbersome.

>    Libraries that are under GPL don't allow any linkage at all (for
> proprietary code) and, frankly, the license doesn't try to hide that
> fact.  In fact, the verbiage is surprisingly blunt on that point.

Which makes GPL libraries rare indeed. Any applications code written to
use such a library would have to be GPL. While that's probably Stallman's
desire, most resonable folks would find it to be overboard.

>    The extra wrinkle with glibc is that it specifically removes the
> requirement for re-linking support.  glibc++ has a different exception
> entirely that is AFAICT supposed to provide the same results when the
> situation is complicated with lots of inlined code (eg: templates).
>
>    Don't take my word for it though.  I've probably missed something
> critical and should be roasted over high heat. YMMV, etc, etc.

I think your points are on point.

[snip]

>    To more directly answer the indirect question posed, there doesn't
> appear to be anything distributed with gcc that would require you to
> distribute your source code just through the act of compiling it using
> the included tools.  There is actually a FAQ entry that states the same
> thing on the FSF page.

Right. Now there was a time when bison/flex output was GPL code. But
even that restriction has been relaxed.

>    Going one step further, I don't recall ever seeing a GPLed library
> that didn't have huge warnings all over the place letting you know not
> to link it into proprietary code so it's not clear to me where the
> problem comes in ?

My question is why would someone make a library GPL if you really wanted
folks to use it? Stallman argues GPL domination here:

http://www.gnu.org/licenses/why-not-lgpl.html

"Using the ordinary GPL for a library gives free software developers an
advantage over proprietary developers: a library that they can use, while
proprietary developers cannot use it."

He posits that the LGPL is for common libraries. But for uncommon ones:

"However, when a library provides a significant unique capability, like GNU
Readline, that's a horse of a different color."

The zealotry definitely bleeds through. I don't buy it. I take the more
moderate stance that everyone should contribute to the community code but
they can do what they want with their own personal code and the combinations
of community and personal code (for Wouter: regardless of the method of
combination).

>    My comments on this thread are solely from the perspective of a
> developer attempting to satisfy standard professional due-diligence into
> licensing issues.  Please don't make assumptions based on what I've said
> here about my position on what you should or shouldn't do with your
> code, personal feelings on the FSF, etc.

Great job. I really appreciate it.

>    In other words, it looks like this thread is dying and I really don't
> want to unnecessarily prolong it.  I would be very interested in
> reasoned critiques of where I've gone wrong in my interpretation though.

Nothing wrong AFAICT.

BAJ

2005\12\23@111223 by Gerhard Fiedler

picon face
Byron A Jeff wrote:

> Correct. The arugment is that the relink requirement allows the library
> to remain "free" and updatable by the end user.
>
> Personally I find it cumbersome to follow for embedded systems work.

Right. I think here we see one of the reasons why people came up with the
distinction between software and firmware. This distinction has become less
"firm" over the years :) but it still exists -- and it shows when trying to
apply the GPL license (created for software) to firmware.

There are many terms in the (L)GPL licenses that seem to be very specific
to software development (and specifically in C or languages that build
similarly, with object files and object file libraries and compile/link
stages). The whole license doesn't seem to make a lot of sense in
situations where this is not given.


> One missing point is that it's the devlopers responsibility to provide the
> end user whatever tools required to make the relink, including propietary
> compilers and linkers.

In the case of software (let's say this includes that it runs on a
sufficiently standardized hardware platform and on a sufficiently
standardized operating system), this is assumed to be provided, somehow.

This got me thinking... Does it make any sense at all to publish, say, a C#
(or, to stay closer to home, a HiTech PICC or Microchip C) library under
LGPL? This seems to require that any application that uses it (even when
only linking in the library) provides the tools required to rebuild it,
which contradicts the licensing terms of the tools:

<http://www.gnu.org/copyleft/lesser.html>
'For an executable, the required form of the "work that uses the Library"
must include any data and utility programs needed for reproducing the
executable from it.'

Sounds as if such a library could not be used by anybody, because nobody
would be able to publish all "utility programs needed for reproducing the
executable"... ?

Gerhard

2005\12\23@111305 by Peter

picon face

On Fri, 23 Dec 2005, Xiaofan Chen wrote:

> I am more confused than before about LGPL and GPL. ;-(

What's so hard. GPL is acutely viral. Source code under GPL remains open
source. LGPL is latently viral. An aggregate work (as in, linked against
it) need not be open source, but it must be provided in linkable form
for relinking (i.e. .obj file) by the user.

Peter

2005\12\23@112716 by Peter

picon face

On Thu, 22 Dec 2005, William Chops Westfield wrote:
> On Dec 22, 2005, at 7:38 PM, Tad Anhalt wrote:
>
>>   _My understanding_ of the situation is that the act of linking itself
>> is what LGPL allows.  Static/dynamic isn't the issue except that there
>> must be a way provided to re-link with (a) different version(s) of the
>> LGPL library(s).
>>
> Which of course wrecks havoc with any embedded sort of application.
> "Here's the .o module of our proprietary code that can be linked
> with this list of LGPL libraries.  And there's your box with its
> OTP soldered in chip containing the old code.  Have fun!"

I do not see a problem with that. The LGPL does not imply non-OTP
devices. Specify flash to the client and provide the .o file and a copy
of the LGPL on the documentation cdrom and just forget about it imho.

Peter

2005\12\23@113203 by William Couture

face picon face
On 12/23/05, Peter <RemoveMEplpTakeThisOuTspamspamactcom.co.il> wrote:

> What's so hard. GPL is acutely viral. Source code under GPL remains open
> source.

The GPL is a voluntary infection.

Most things come with a price.  You pay $$$ for a compiler/library, that's
the cost that the author has decided to set.

For GPL software, the cost is not money.  The cost is that code that
uses GPL software must be relased under the same GPL.

If this cost is unacceptable to you, don't use it.

However, unlike a purchased compiler/library, if you don't like the
cost, you can still look at the code and take ideas.  As long as you
don't copy the code, you can create your own using the ideas of
the GPL software without invoking the GPL for your software.

Bill

--
Psst...  Hey, you... Buddy...  Want a kitten?  straycatblues.petfinder.org

2005\12\23@113816 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Let's say I write up something, put it in an envelope, seal that
>> envelope and put it in a lock box to which I don't have a key, and make
>> sure that I have a definite proof (through the lock box owner, by a
>> sufficiently "definite" procedure) that I put that specific work in the
>> lock box on 2005-12-23.
>>
>> Can I later claim that I "published" this work for copyright uses?
>
> AFAIK not, at least not in my country.

Are you definitely sure? What would constitute a "publication" in this
context in your country? How "public" must it be? For example, I know that
generally copyright retention is an issue in work for hire (not sure about
your country, though). Work for hire rarely gets published, though...

en.wikipedia.org/wiki/Berne_Convention_for_the_Protection_of_Literary_and_Artistic_Works
only speaks of "creation", not of publication. But IANL... :)


> AFIK you need to prove that someone *could* have copied your work,
> because it was publicly available.

Again, the question is what means "publicly available". Anything in my
house is available to a burglar (see Howard's post). Something I published
in a trade magazine that's only available to certain qualified companies
may not be available to that burglar. So, from the burglar's point of view,
my "publication" is less "available" than my lab book with my private
notes.


>> I also think that I could use it as prior art in later patent disputes
>> (which of course is a different question than copyright).
>
> Again, no. Prior art is to be interpreted as prior art in the common
> (accessible) domain ("as accessible or know to a crafstman in the
> applicable field").

This seems to be different in different countries. For example, in the USA
it seems to count as prior art if somebody else (in the USA) knew about it
before the claimed date of invention. In that case, the contents of my
envelope would suffice.

>From <www.tms.org/pubs/journals/JOM/matters/matters-9106.html>:
'[...] a person is not entitled to a patent if the invention was "known or
used by others in this country, or [...]'

But then, <http://www.iusmentis.com/patents/priorart/> seems to agree with
you and require some form of "publication".

Finally, the Wikipedia <http://en.wikipedia.org/wiki/Prior_art> is somewhat
in between those two extremes (at least for the USA)... Doesn't seem to be
an issue that could be resolved without a lawyer :)

Gerhard

2005\12\23@115209 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> So why would you say that this shows that he doesn't seem to understand
>> copyright?!?
>
> Because he seemed surprised by this.

I must have missed this... :)

Gerhard

2005\12\23@133149 by Wouter van Ooijen

face picon face

> >> So why would you say that this shows that he doesn't seem
> to understand
> >> copyright?!?
> >
> > Because he seemed surprised by this.
>
> I must have missed this... :)

or maybe I imagined it.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@133149 by Wouter van Ooijen

face picon face
>>> Can I later claim that I "published" this work for copyright uses?
>>
>> AFAIK not, at least not in my country.
>
> Are you definitely sure? What would constitute a "publication" in this
> context in your country?

I checked an I was wrong. Creation is sufficient.

> This seems to be different in different countries. For
> example, in the USA it seems to count as prior art if somebody
> else (in the USA) knew about it
> before the claimed date of invention. In that case, the contents of my
> envelope would suffice.

(the next is translating a Dutch text to english, so this might not
apply outside my country)

[at patent must be on something new] "New is everything that does not
belong to the stare of the art and at the moment the patent is filed is
not accessible by description, use or otherwise, in any country."

Accessible does not seem to be easily defined, a sideline of this text
explains that showing something in a fair is making it accessible,
unless that fair is a "Universal Exhibition".

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@140634 by Gerhard Fiedler

picon face
Peter wrote:

>>>   _My understanding_ of the situation is that the act of linking itself
>>> is what LGPL allows.  Static/dynamic isn't the issue except that there
>>> must be a way provided to re-link with (a) different version(s) of the
>>> LGPL library(s).
>>>
>> Which of course wrecks havoc with any embedded sort of application.
>> "Here's the .o module of our proprietary code that can be linked
>> with this list of LGPL libraries.  And there's your box with its
>> OTP soldered in chip containing the old code.  Have fun!"
>
> I do not see a problem with that. The LGPL does not imply non-OTP
> devices. Specify flash to the client and provide the .o file and a copy
> of the LGPL on the documentation cdrom and just forget about it imho.

It seems you are right in that the LGPL doesn't require you to provide the
user with a possibility to actually write the code into any chip. (Which
kind of shows that the GPL was not intended to be applied to firmware. The
actual requirement, see below, doesn't really make much sense in such a
situation.)

But it seems to require you to provide the user with all tools to create
your application executable. Which seems to exclude applications written
for any commercial compiler. Which seems to imply that you can't license a
library that's written for a commercial compiler with LGPL. (Well, you of
course can, but it seems that then nobody would be able to use it.)

(This issue doesn't seem to have to do much with embedded or not.)

Gerhard

2005\12\23@150849 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

> "[...] is not accessible by description, use or otherwise, in any
> country."

Let's face it, without a worldwide administration of prior art (which kind
of implies at least the start of a world government), there won't be a fix
to the patent system. Whether that can fix it though is not really clear :)

There's no way currently that any patent office (nor any inventor) can
research prior art "in any country" reasonably (reasonable certainty,
reasonable timeframe, reasonable price, reasonable everything :).

Which then makes the value of a patent that what it currently is, mostly:
it's worth as much as you can spend to defend it.


> Accessible does not seem to be easily defined

Is there ever a law that is in its entirety "easily defined"? :)

Gerhard

2005\12\23@151046 by William Chops Westfield

face picon face

On Dec 23, 2005, at 4:53 AM, Gerhard Fiedler wrote:

> Let's say I write up something ..and put it in a lock box ...
> Can I later claim that I "published" this work for copyright uses?

No.  Don't forget that both patent and copyright law are supposed to be
incentives for people to make their works available "at large."  A sort
of "you show us all what  you've done, and we'll protect your ability
to make a profit from it."

BillW

2005\12\23@154226 by Wouter van Ooijen

face picon face
> There's no way currently that any patent office (nor any
> inventor) can research prior art "in any country" reasonably

The current practice is AFAIK that you can file a patent and get without
any researhc in to "newness", but when you want to sue someone for
infringement you must supply a newness-research-report.

> Which then makes the value of a patent that what it currently
> is, mostly: it's worth as much as you can spend to defend it.

Or the inverse: inversely proportional to the amount your opponent can
spend to attack it.

> Is there ever a law that is in its entirety "easily defined"? :)

No, but there are gradations.

I once went to a real law lesson (not the law lessons for techies, but
one for first year as-yet-to-be lawyers). I was so fascinated by the
subject that I bought the book and swallowed it in a few days. It was
about the history of legislation (the writing down of laws). There are
two conflicting principles that law has to satisfy: fairness and
predictability.  So we start with very simple, well defined, predictable
laws, which are unfortunately after a short time felt unfair. New laws
are added, and more freedom is give to judges, to enhance the fairness.
But this leads to unpredicatability, and in the end people ask again for
simple, predicatable laws and the cycle starts again. In ancient Rome
there have actually been a few cycles of this.

I in opinion the current laws are to far to the unpredictable side. I
would accept quite some amount of unfairness in exchange for more
predicatbility.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\12\23@192445 by Marc Lavallée

flavicon
face
Le 23 Décembre 2005 13:31, Wouter van Ooijen a écrit :
> > >> So why would you say that this shows that he doesn't seem
> >> > to understand
> > >> copyright?!?
> > >
> > > Because he seemed surprised by this.
> >
> > I must have missed this... :)
>
> or maybe I imagined it.

Here's another suggested reading:
"GPL: Viral Infection or Just Your Imagination?" :
http://www.linuxinsider.com/story/33968.html
--
Marc

2005\12\23@214900 by Gerhard Fiedler

picon face
William ChopsWestfield wrote:

>> Let's say I write up something ..and put it in a lock box ...
>> Can I later claim that I "published" this work for copyright uses?
>
> No.  

This doesn't seem to be as dry and cut as you put it. There seems to be a
number of legal comments that use "creation" instead of "publication".

See "http://en.wikipedia.org/wiki/Copyright", in the section about
obtaining copyright in the USA:

"Once an idea has been reduced to material form, for example by securing it
in a fixed medium (such as a drawing, sheet music, photograph, a videotape
or a letter), the copyright holder is entitled to enforce his or her
exclusive rights."

See also the interesting chart
http://www.bromsun.com/practices/copyright-portfolio-development/flowchart.htm
and the distinction between creation and publication it makes.

I'm glad they abandoned publication as criterion; defining "publication" in
a suitable way seems to be pretty complicated.


> Don't forget that both patent and copyright law are supposed to be
> incentives for people to make their works available "at large."  

While I agree for patents, I don't think that holds for copyright.
Copyright seems to be following more an "I created it, so it's mine"
feeling, along the lines of limited property rights.

It's difficult to quantify, but I would guess that most copyrighted
material never becomes available "at large".

Gerhard

2005\12\23@225418 by D. Jay Newman

flavicon
face
> > Don't forget that both patent and copyright law are supposed to be
> > incentives for people to make their works available "at large."  
>
> While I agree for patents, I don't think that holds for copyright.
> Copyright seems to be following more an "I created it, so it's mine"
> feeling, along the lines of limited property rights.
>
> It's difficult to quantify, but I would guess that most copyrighted
> material never becomes available "at large".

Copyright does *eventually* expire. I seem to recall 75 years in some
cases in the US.
--
D. Jay Newman           ! _Linux Robotics: Building Smarter Robots_
EraseMEjayspamspamspamBeGonesprucegrove.com     ! To be released soon to unsuspecting bookstores
http://enerd.ws/robots/ ! everywhere.

2005\12\24@065139 by Gerhard Fiedler

picon face
D. Jay Newman wrote:

>>> Don't forget that both patent and copyright law are supposed to be
>>> incentives for people to make their works available "at large."  
>>
>> While I agree for patents, I don't think that holds for copyright.
>> Copyright seems to be following more an "I created it, so it's mine"
>> feeling, along the lines of limited property rights.
>>
>> It's difficult to quantify, but I would guess that most copyrighted
>> material never becomes available "at large".
>
> Copyright does *eventually* expire. I seem to recall 75 years in some
> cases in the US.

Yes, it does expire, but it's not that simple :)  For works "fixed" after
1978-1-1, it expires 70 years after the author's death, if the author is
the copyright holder. In other cases, 95 years after publication or 120
years after creation, whichever comes earlier.

Again, this already mentioned chart shows the various possible expiration
dates
http://www.bromsun.com/practices/copyright-portfolio-development/flowchart.htm


In case this was a comment re becoming or not available, I'm just as
convinced that most copyrighted material doesn't become available, not
before and not after the copyright expires. It's just that most copyrighted
material is proprietary, and it won't be available "at large". Ever.


Here's a phrase from that chart above about "publication" in the context of
copyright:

""Published" is a technical term that is both broader and narrower than the
lay definition. For instance, in general it requires, among other things,
the copyright owner's authorization. In some situations, distribution of
even a single copy of a work may constitute publication for copyright
purposes."

Gerhard

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...