Re: patent on O_ATOMICLOOKUP [Re: [PATCH] loopable tmpfs (2.4.17)]

Re: patent on O_ATOMICLOOKUP [Re: [PATCH] loopable tmpfs (2.4.17)]

[Adam J. Richter]

	If Red Hat is able to leave the licensing issues for their
Linux patents unresolved and they still manage to be regarded as being
in good standing with most contributors and would-be customers, then,
in the coming months, many other Linux companies will take this is a
green light to file for many patents and remain silent when asked to
explictly grant permission to the public to practice the patents in
free software.  There is little or no business reason to publicly
grant such permission if one can get away with not doing so.

	Although other companies today already have many patents that
they could argue are infringed by certain free software components.
The Linux patents are different as a practical matter, however, in
that the chance of prevailing in that argument will be greater when
if alleged infringer is using the code for which the patent was
originally submitted.

	Eventually, as some companies are bought or go out of
business, it is a statistical certainty that some of these patents
will pass into the control of parties that do not care about the GPL's
penalties for enforcing a software patent (after all, that would allow
litigation only against copiers of the software, and a copyright owner
would have to sue, which is approximately already the level of danger
one has with an unlicensed software patent).

Adam J. Richter     __     ______________   575 Oroville Road
adam@yggdrasil.com     \ /                  Milpitas, California 95035
+1 408 309-6081         | g g d r a s i l   United States of America
                         "Free Software For The Rest Of Us."

business models [was patent stuff]

[Larry McVoy]

Jumping on the chance to cause more discussion...  This probably ought to
be off on some other list, but I don't know what that is.  Alan will be 
sending out the guards to herd me back into the ward, but I've escaped
for the moment and I'll make the best of it :)

On Mon, May 27, 2002 at 02:52:13PM -0700, Adam J. Richter wrote:
> 	Eventually, as some companies are bought or go out of
> business, it is a statistical certainty that some of these patents
> will pass into the control of parties that do not care about the GPL's
> penalties for enforcing a software patent 

If the free software community is ever going to really compete with the
non-free software community, they simply have to come up with a better
business model than giving it away and trying to make money on support.
It's economics 101 - a free market will go to whomever can provide the
needed service most cheaply.  With no barrier to entry, that means as
soon as the price gets high enough, someone will resell the product for
less.  Which results in razor thin profits, if any at all.

In my opinion, it's time for the free software fanatics to ease off and
let some moderates come in and try and define a reasonable compromise.
We all need to realize that we can let businesses figure it out themselves
and we may not like what they decide, or try and define a compromise
that can be lived with.  Whether you like it or not, the patent ploy
works and there isn't a damn thing you can do about it if it is a legit
patent.

If you hold the "It's GPL or bugger off" position, people will figure out
how to work around it and it is virtually certain you won't like what they
do.  If you offer them some sort of reasonable compromise, I'll bet they
take it.  If you don't, you get to live with whatever their nasty evil
business minds dream up.

This whole line of reasoning is why I detest the OSI and Eric Raymond
in particular.  They had a chance to define a "Buiness Source License"
or some other compromise and get the world to consider it as an option.
Instead, they just made a lot of noise and for what?  Ask yourself -
how much more open source is there in the world today versus 5 years ago.
Imagine that pile.  Now take the OSI out of the picture and tell me what
would not be in that pile anyway.  Bloody little.  The OSI squandered a
golden opportunity to really change the world, and I find that depressing
beyond words.

Sorry, I'm wandering.  The discussion I'd like to see is one in which
people explored the values they hold dear and tried to come up with a
business model which preserves those values and allows that business
to compete with the likes of Microsoft.  Yeah, yeah, I know that Linux
kicks butt for print serving and web serving, but the 99% reason it
does is price.  It's not because Linux has new compelling features that
Microsoft doesn't have, it's because it's hard for Microsoft to compete
with stuff that costs zero dollars.  The problem isn't where we are,
the problem is where we are going (or more to the point, not going).
How is Linux and open source ever going to be a leader, producing new
applications, new protocols, new languages, new markets when it doesn't
generate the incredible amounts of revenue needed to build all that?

Ask yourself - how much open source is a reimplementation of what has
already been designed and implemented, and how much is fundamentally new?
That new stuff costs huge dollars, not because of the cost of building it,
but because of the cost of building all the crap that turned out bad
but provided the insight that lead to the new stuff.  It's really not
that hard to reimplement something, open source has proven that beyond
all doubt.  What it hasn't proven is that open source leads to new ideas,
products, and markets.  So far, open source follows, it doesn't lead.
A reasonable business model might change that.  There may be other ways
to change it, but something needs to change or 20 years from now there
will be open source versions of all the current popular apps, but still
playing catch up on the next generation.

My 2 cents.
-- 
---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Re: business models [was patent stuff]

[Karim Yaghmour]

I agree that this should go on some other list, but 'til then, here are my 2c.

No offense Larry, but many of your arguments are the same used by Microsoft
to push their vision of publicly available source.

Larry McVoy wrote:
> It's economics 101 - a free market will go to whomever can provide the
> needed service most cheaply.

I will take the liberty of rephrasing to illustrate my point of view:
"a free market will go to whomever can provide the needed service."

The question is: Can you provide the service? Of course if the service
you are selling is common knowledge, then you've got nothing to sell
your competitor can't. But if you're part of building the technology
then you're certainly in a know-how monopoly position. This is why I
don't think any of the kernel developers will ever be out of a job.

The drawback to this is that you simply can't scale a knowledge-based
company the way you do with a classical intellectual-property-based
company. The way I see it, the software industry will look increasingly
like that of other speciality fields such as law and medicine. Sure, any
doctor can administer a serum, but not every doctor can actually perform
robotic heart-surgery. He who can perform robotic heart-surgery can
offer something other doctors can't. Same will be with the software field.

> In my opinion, it's time for the free software fanatics to ease off and
> let some moderates come in and try and define a reasonable compromise.

Can you suggest a list of names of some "moderates"?

> If you hold the "It's GPL or bugger off" position, people will figure out
> how to work around it and it is virtually certain you won't like what they
> do.  If you offer them some sort of reasonable compromise, I'll bet they
> take it.  If you don't, you get to live with whatever their nasty evil
> business minds dream up.

I predict the inverse. Of course, people will actually try to use patents to
restrict free and open source software. And of course, they will push this
as hard and as far as they can. The community, however, will always find
alternative ways to obtain the same results and, in the end, no client wil
use the patent holder's products or services. Instead, they will use the
community's alternative solutions.

You would like the open source and free software communities to get used
to having their rights being violated. I think the software "manufacturers"
better get used to the fact that they can't outsmart the community,
regardless of the legal/political/financial tools they use.

As I said earlier, the current software business model is an endangered
species.

Best regards,

Karim

===================================================
                 Karim Yaghmour
               karim@opersys.com
      Embedded and Real-Time Linux Expert
===================================================

Re: business models [was patent stuff]

[Austin Gonyou]

Man..it's like Farenheit 451, but more like "Hex 0xffff"

On Mon, 2002-05-27 at 18:15, Karim Yaghmour wrote:
...
> I predict the inverse. Of course, people will actually try to use patents to
> restrict free and open source software. And of course, they will push this
> as hard and as far as they can. The community, however, will always find
> alternative ways to obtain the same results and, in the end, no client wil
> use the patent holder's products or services. Instead, they will use the
> community's alternative solutions.
> 
> You would like the open source and free software communities to get used
> to having their rights being violated. I think the software "manufacturers"
> better get used to the fact that they can't outsmart the community,
> regardless of the legal/political/financial tools they use.


I tend to agree with Karim here for sure. It's human nature to overcome
adversity, and this is adversity in one frame of mind, not from
different points of view. (Just to clarify the MPAA see adversity in
certain aspects of the OSS community I think, but they're the opressors,
not us).

Also, in the paragraph about "GPL or bugger off", if you feel it's over
the top, then it's possible others will too, and then it's simply
survival of the fittest. People must change to a degree, but the message
they carry is, and probably will remain, the same. 

Austin

Re: business models [was patent stuff]

[Peter Wächtler]

Karim Yaghmour wrote:
> I agree that this should go on some other list, but 'til then, here are my 2c.
> 
> No offense Larry, but many of your arguments are the same used by Microsoft
> to push their vision of publicly available source.
> 

Huh?

Just a quick suggestion: keep Linux open (THE operating system) as basis
for future development, enhancement. I think it's called "public facility"
like power plants and infrastructure that is needed for a healthy society
(at least politicians shall work for the publics benefit).

Applications and other value add are produced and provided by third
parties and SOLD. You can do that legally TODAY. With the laws and
licenses (e.g. LGPL) used TODAY.

Microsoft wants to control almost everything (on computers).
If they place a new API on their system, the whole industry has to
follow. Hey, what's up with the laws that should protect me (and others)?
Uh, they have to provide a hook for another email/web client!
Who thinks that this is adequate? Who is guilty of what?
Who cares?

Re: business models [was patent stuff]

[Mark Mielke]

On Tue, May 28, 2002 at 11:29:24AM +0200, Peter Wächtler wrote:
> Karim Yaghmour wrote:
> > I agree that this should go on some other list, but 'til then, here
> > are my 2c.  No offense Larry, but many of your arguments are the
> > same used by Microsoft to push their vision of publicly available
> > source.
> Just a quick suggestion: keep Linux open (THE operating system) as basis
> for future development, enhancement. I think it's called "public facility"
> like power plants and infrastructure that is needed for a healthy society
> (at least politicians shall work for the publics benefit).

Just for a quick interjection... under the model you describe, the
governments of this world should be paying for linux kernel
development. :-)

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: business models [was patent stuff]

[Peter Wächtler]

Mark Mielke wrote:
> On Tue, May 28, 2002 at 11:29:24AM +0200, Peter Wächtler wrote:
> 
>>Karim Yaghmour wrote:
>>
>>>I agree that this should go on some other list, but 'til then, here
>>>are my 2c.  No offense Larry, but many of your arguments are the
>>>same used by Microsoft to push their vision of publicly available
>>>source.
>>>
>>Just a quick suggestion: keep Linux open (THE operating system) as basis
>>for future development, enhancement. I think it's called "public facility"
>>like power plants and infrastructure that is needed for a healthy society
>>(at least politicians shall work for the publics benefit).
>>
> 
> Just for a quick interjection... under the model you describe, the
> governments of this world should be paying for linux kernel
> development. :-)
> 

Do you vote against that? I'm not.

It does not have to be financed by the government - but it can.
And I think that already happens in smaller amounts in Germany.
The BSI (my translation: ministry for security in information technology)
already sponsors the project SPHINX for developing encrypted email
services (based on GNU gpg). This is only one example.

Why not? Computers are getting more and more important in our daily life.

It's obvious to protect common goods. Microsoft can still develop
their non disclosed products for "more security".
I thought moms and dads would still teach their children:
"If a stranger talks to you, offering chocolate and says: trust me -
run away and ask other for help"

Oh, is that too old fashioned?  ;-)

Re: business models [was patent stuff]

[Gilad Ben-Yossef]

On Tue, 2002-05-28 at 01:24, Larry McVoy wrote:

> If the free software community is ever going to really compete with the
> non-free software community, they simply have to come up with a better
> business model than giving it away and trying to make money on support.
> It's economics 101 - a free market will go to whomever can provide the
> needed service most cheaply.  With no barrier to entry, that means as
.........^^^^^^^^
> soon as the price gets high enough, someone will resell the 
> product for
..^^^^^^^
> less.  Which results in razor thin profits, if any at all.

Software is not a "product" any more then a lawyer argument in a case is
a "product" or an architect plan for a building is a "product".  That's
the whole problem in a nut shell. If you continue to view software as a
sellable object then the business model indeed doesn't make sense. Then
again, if you try to look at a lawyers argument in a case as the
"product" of the law firm then that same business plan doesn't make
sense either. After all, in most countries I'm aware of, a lawyer
argument cannot be effectivly copyrighted. Of course, this doesn't stop
lawyers from make a living. Just ask anyone at Microsoft legal
depratment... ;-)

Most "Open Source service companies" that I'm aware off don't understand
the implications. They're not built like a small law firm that may be
large one day; they're built like a company that tries to be a 90 pound
world class gorrila backed by Vulture Capital. Sorry, but it doesn't
work like this: you have to walk before you can run.

So there's a business model and it works for years in other fields.
Maybe the problem is that people this path is it isn't "sexy" - they
will have to lose the dream of being High Tech millionares who built a
megacorporation and start thinking  like a small time law firm, which
might one day be big. Maybe they don't like to work in a consultant kind
of way. I can understand that and it's OK, but there is a business
model.

The way Open Source (or Free) software works is to multiply the "use
value" of software by killing the "sell value". But here is the
interesting part: it increases the "use value" of the open/free software
in a given niche, but kills the "sell value" of ALL software competing
in that niche. This is why it's going to win, cold hard economic facts
and no need for FSF like ideology (not that there's anything wrong with
that... ;-))))

And BTW I'm OK with patents as long as their licensed for free to GPLed
software (maybe add a clause that makes this irreversable?).

> How is Linux and open source ever going to be a leader, producing new
> applications, new protocols, new languages, new markets when it doesn't
> generate the incredible amounts of revenue needed to build all that?

A. Who cares? We're having fun.
B. It already is in many senses.

I even think that A leads to B, but that's just me...

> Ask yourself - how much open source is a reimplementation of what has
> already been designed and implemented, and how much is fundamentally new?

Ask yourself - how much science is a reimplmentation of what has 
already been researched and implmented, and how much is fundamentally
new?

If I remember correctly Linus quoted some dead guy about this once,
something about standing on the shoulders of giants...

OK, I'll now go back to being a quiet little lurker... ;-)

-- 
Gilad Ben-Yossef <gilad@benyossef.com>
http://benyossef.com
"Hail Eris! All Hail Discordia!"

Re: business models [was patent stuff]

[Daniel Phillips]

On Tuesday 28 May 2002 09:53, Gilad Ben-Yossef wrote:
> And BTW I'm OK with patents as long as their licensed for free to GPLed
> software (maybe add a clause that makes this irreversable?).

Indeed, an irreversible license is essential, but that's not all.  Not
only must the GPL be accommodated, but all open source licenses.  And it
is not enough to be free as in "without cost": it must also be free of
additional restrictions.

-- 
Daniel

Re: business models [was patent stuff]

[Alan Cox]

On Tue, 2002-05-28 at 09:57, Daniel Phillips wrote:
> On Tuesday 28 May 2002 09:53, Gilad Ben-Yossef wrote:
> > And BTW I'm OK with patents as long as their licensed for free to GPLed
> > software (maybe add a clause that makes this irreversable?).
> 
> Indeed, an irreversible license is essential, but that's not all.  Not
> only must the GPL be accommodated, but all open source licenses.  And it
> is not enough to be free as in "without cost": it must also be free of
> additional restrictions.

I don't think you can realistically expect all open source licenses like
the BSD one to be accomodated. Otherwise people would ship binary apps
linked with a BSD licensed libpatent.o/c that was useless to anyone. The
GPL restrictions happen to work very nicely in terms of making a patent
available for free software (or one definition thereof), the BSD license
alas doesn't.

Re: business models [was patent stuff]

[Ingo Oeser]

On Mon, May 27, 2002 at 03:24:52PM -0700, Larry McVoy wrote:
> If the free software community is ever going to really compete with the
> non-free software community, they simply have to come up with a better
> business model than giving it away and trying to make money on support.

Why? Users are always in the need of support. They just don't
like to pay for it. That's why they ask the UseNet, Webforums and
the like.

> It's economics 101 - a free market will go to whomever can provide the
> needed service most cheaply.  With no barrier to entry, that means as
> soon as the price gets high enough, someone will resell the product for
> less.  Which results in razor thin profits, if any at all.
 
The barrier of entry is building up a support (and testing) group for the
intended audience. That's not easy, so many companies fail to
offer it at a reasonable price and thus with those razor thin
profits. But it is still being tried.

Selling knowlegde works better, then selling IP and leads to
more innovation, because you ALWAYS have to think and rebuild
your knowledge and cannot just develop (or buy) sth. and be done
with it.

> Whether you like it or not, the patent ploy works and there
> isn't a damn thing you can do about it if it is a legit patent.

This stuff works only for lawyers revenues, not for getting
something done. Attacking other and defending a your own property
is important and effective in the short run, but your army will
be out of food soon, if you have more soldiers then farmers
getting sth. done.

Look at the hold times of patents and look at the things you can
do with it. You are not forced to market it, you will not loose
it in a reasonable timeframe (which would be 2-3 years for all
software) or at reasonable events (>70% market share) and you
might even not be forced to license it.

Patents are intellectual property, where intellectual property
can be copied without limits (only the intellectual capacity of
people ;-)) without loosing its public value.

Even now IP doesn't matter that much, because on the biggest
market (the consumer one) it depends on marketing, how big your
revenues are and not really on the superiority of your IP.

And now lets check, what revenue streams Patents create:
Creation:
- If you have an idea, you must check whether it's patented and
  a reasonably new idea
  -> revenue stream from creator to a lawyer doing the patent research
  -> revenue stream from creator to a expert checking the idea
  whether its new
  -> revenue stream from creator to the patent office for
     registering it

While you own it:
- Defending it
  -> revenue stream from owner to a lawyer for defending you on court
  
- Using it
  -> revenue stream from owner to laboratory to develop it further
  -> revenue stream from owner to factory to produce it
  -> revenue stream from owner to marketing dept. to market it
  -> revenue stream to owner from sales dept. for selling it

- License it 
  -> revenue stream to owner from licensee for using it

Getting rid of it:
- Selling it:
  -> revenue stream from new owner to old owner

- Trading it for another one:
  -> no revenues involved or only involved on one side

While you not own it anymore:
- If you encounter the same problem again
   + you have to license your old patent again 
   -> revenue stream from creator to owner (stupid laws?)

   + you can think hard, to solve it better this time
   -> creation of new patent, revenue streams as described there

   + you can think not so hard, to solve it just different enough
      to not violate your old patent (common practise)
   -> creation of new patent, revenue streams as described there

- If you want to use a product with your patent included
   + You have to pay the license like everyone else
   -> revenue stream from creator to owner (stupid laws?)

So you can easily see, how much the creator who PROVIDED the IP
for the patent will get. 

Patents where intended to protect the creators, but are in fact
protecting the owners and make the creators replacable with
cheaper engineers (doing the testing and further development)
after they wrote down the idea and the patent is filed. 

This is not very motivating for doing real in depth research or
basic research and leads to the things we see now with OS
development: We benchmark a lot, tune a lot but real ideas are
rare, because they are expensive and nobody pays them.

Analyzing the revenue streams without patents or with more limited
patents (xref: s.a.) is left as an exercise for you to better
defend your arguments ;-)

> If you hold the "It's GPL or bugger off" position, people will figure out
> how to work around it and it is virtually certain you won't like what they
> do.  If you offer them some sort of reasonable compromise, I'll bet they
> take it.  If you don't, you get to live with whatever their nasty evil
> business minds dream up.

The FSF is fighting against software patents. Better would be to
fight for reasonable adopting software patents to the life cycle
of software and fighting for more revenues to the creators (who
actually getting the work done).
 
> Ask yourself - how much open source is a reimplementation of what has
> already been designed and implemented, and how much is fundamentally new?

Reimplementation from specification is needed to improve the
quality of the implementation itself. Software reimplementation
is an important part of its life cycle. Also new variants of an
interface and behavior description (which can be patented) are
needed to provide the user with more options and the programmer
with more practise.

Thats the quest of free and open source: Trying to be better.
Trying to be new is the quest of the commercial development,
because new things are the only thing the consumer buys and they
will simply not pay for you correcting your mistakes.

> That new stuff costs huge dollars, not because of the cost of building it,
> but because of the cost of building all the crap that turned out bad
> but provided the insight that lead to the new stuff.

Right. So please make this patented crap (not marketed) and also
the new stuff (marketed) public knowlegde after 1 year and 3 years.

> It's really not
> that hard to reimplement something, open source has proven that beyond
> all doubt.  What it hasn't proven is that open source leads to new ideas,
> products, and markets.  So far, open source follows, it doesn't lead.

It merely collects and implements public knowledge, that is
scattered everywhere in the world. I look at it as a source code
library for my own (or my employers) needs.

> A reasonable business model might change that.  There may be other ways
> to change it, but something needs to change or 20 years from now there
> will be open source versions of all the current popular apps, but still
> playing catch up on the next generation.

Do we want to change that? I don't. 

I have no problems paying for bleeding edge and paying nearly
zero for being 2 years behind. People that need bleeding edge are
enough and willing to pay. But I'm also willing to pay for decent
support ;-)

Regards

Ingo Oeser
-- 
Science is what we can tell a computer. Art is everything else. --- D.E.Knuth

Re: business models [was patent stuff]

[Jonathan Corbet]

> Jumping on the chance to cause more discussion...  This probably ought to
> be off on some other list, but I don't know what that is.  

For what it's worth, something like the Free Software Business List, which
has a (rather minimalistic) web page at http://www.crynwr.com/fsb/,
welcomes this sort of discussion.

jon

Jonathan Corbet
Executive editor, LWN.net
corbet@lwn.net

Re: business models [was patent stuff]

[Eric W. Biederman]

Larry McVoy <lm@bitmover.com> writes:

> Jumping on the chance to cause more discussion...  This probably ought to
> be off on some other list, but I don't know what that is.  Alan will be 
> sending out the guards to herd me back into the ward, but I've escaped
> for the moment and I'll make the best of it :)
[snip]
> If you hold the "It's GPL or bugger off" position, people will figure out
> how to work around it and it is virtually certain you won't like what they
> do.  If you offer them some sort of reasonable compromise, I'll bet they
> take it.  If you don't, you get to live with whatever their nasty evil
> business minds dream up.

To a certain extent other people's business model is none of your
business.  The only people who have a reasonable right to ask that
question is the company itself, and it's investors.  And support is
far from being the only viable GPL compatible business model. 

Most new businesses fail, even when they are using known proven
business models.  It is execution, and being in the right place at the
right time that make a successful company. 

For the GPL or bugger off folks, unless you have a business model
that delivers GPL'd code to them, you don't make any money from them.
Not offering them the GPL violates ``The customer is always right''.
Of course it is perfectly fine to concentrate on those people who you
can figure out how to make money from.

There is only one test for a successful business model, is the company
here making money 20 years from now.  That test is going to take a
while.

Eric

Re: patent on O_ATOMICLOOKUP [Re: [PATCH] loopable tmpfs (2.4.17)]

[Alan Cox]

On Mon, 2002-05-27 at 22:52, Adam J. Richter wrote: 
> 	If Red Hat is able to leave the licensing issues for their
> Linux patents unresolved and they still manage to be regarded as being

Nobody intends to leave in unresolved. It's simply a matter of coming up
on the Friday of a long weekend when most folk are not back until
Tuesday. I think everything will have been appropriately addressed
including the questions of the 'what if IBM/AOL/Sony/.. buy Red Hat' [1]
 kind

Alan
[1] Pick a bogus rumour, any rumour...

Re: business models [was patent stuff]

[Gilad Ben-Yossef]

On Tue, 2002-05-28 at 14:51, Paratimer@aol.com wrote:
> In a message dated 5/28/2002 3:55:00 AM Eastern Daylight Time, 
> gilad@benyossef.com writes:
> 
> 
> > On Tue, 2002-05-28 at 01:24, Larry McVoy wrote:
> > 
> > > If the free software community is ever going to really compete with the
> > > non-free software community, they simply have to come up with a better
> > > business model than giving it away and trying to make money on support.
> > > It's economics 101 - a free market will go to whomever can provide the
> > > needed service most cheaply.  With no barrier to entry, that means as
> > .........^^^^^^^^
> > > soon as the price gets high enough, someone will resell the 
> > > product for
> > ..^^^^^^^
> > > less.  Which results in razor thin profits, if any at all.
> > 
> > Software is not a "product" any more then a lawyer argument in a case is
> > a "product" or an architect plan for a building is a "product".  
> 
> Actually, a lawyer's argument or an architects plan can certainly
> be work products, and if I hire a lawyer to draw up a contract for
> me or an architect to design a building for me, unless our agreements
> say otherwise, I own the contract or the building plans.
> 

Of course you *can*, but do you (or anyone else) actually does this?
AFAIK this is not a very common business model. 

Gilad.
-- 
Gilad Ben-Yossef <gilad@benyossef.com>
Code mangler, senior coffee drinker and VP SIGSEGV
Qlusters ltd.

"A billion flies _can_ be wrong - I'd rather eat lamb chops than shit."
	-- Linus Torvalds on lkml

Re: business models [was patent stuff]

[Adam J. Richter]

Alan Cox responds to Gilad Ben-Yossef about licensing of Linux-based patents:
>I don't think you can realistically expect all open source licenses like
>the BSD one to be accomodated. Otherwise people would ship binary apps
>linked with a BSD licensed libpatent.o/c that was useless to anyone. The
>GPL restrictions happen to work very nicely in terms of making a patent
>available for free software (or one definition thereof), the BSD license
>alas doesn't.

	You could license all programs that consist entirely of
free software.  That way, BSD, LGPL, and MPL software that did
not link in proprietary software would be allowed too, but your
example of a proprietary program that linked in the BSD'ed
libpatent.o/c would not be covered by this permission.

	Licensing all programs that consist entirely of free
software would address issues like GPL'ed software that contains
content that is GPL compatible but not GPL'ed, future versions of
the GPL, what happens if a court ruling opens a loophole in the
GPL, and practicing the patent with free software in other free
operating systems.

Adam J. Richter     __     ______________   575 Oroville Road
adam@yggdrasil.com     \ /                  Milpitas, California 95035
+1 408 309-6081         | g g d r a s i l   United States of America
                         "Free Software For The Rest Of Us."

Re: business models [was patent stuff]

[Daniel Phillips]

On Tuesday 28 May 2002 19:13, Adam J. Richter wrote:
> Alan Cox responds to Gilad Ben-Yossef about licensing of Linux-based patents:
> >I don't think you can realistically expect all open source licenses like
> >the BSD one to be accomodated. Otherwise people would ship binary apps
> >linked with a BSD licensed libpatent.o/c that was useless to anyone. The
> >GPL restrictions happen to work very nicely in terms of making a patent
> >available for free software (or one definition thereof), the BSD license
> >alas doesn't.
> 
> 	You could license all programs that consist entirely of
> free software.  That way, BSD, LGPL, and MPL software that did
> not link in proprietary software would be allowed too,

That's a good start, but it's not enough.  There remains the question of
additional restrictions relating to other programs/hardware/whatever.  The 
Fsmlabs RTLinux patent license is a good example of such additional 
restrictions.  This license restricts the royalty-free grant to software 
that is not realtime, or to software that cannot be freely modified.  The
license does not discriminate on the basis of linking, as the GPL does, but
rather, on the nature of the application.

While it's an open question whether or not this is a suitable way for
Fsmlabs to license their patent, it seems clear that similar restrictions
would be entirely inappropriate for a Linux vendor.

-- 
Daniel

Re: business models [was patent stuff]

[Alan Cox]

On Tue, 2002-05-28 at 18:13, Adam J. Richter wrote:
> 	You could license all programs that consist entirely of
> free software.  That way, BSD, LGPL, and MPL software that did
> not link in proprietary software would be allowed too, but your
> example of a proprietary program that linked in the BSD'ed
> libpatent.o/c would not be covered by this permission.

Define "free software" using only legally defined phrases which have
precedent. In fact put four people in a room and get them to define free
software.

> software would address issues like GPL'ed software that contains
> content that is GPL compatible but not GPL'ed, future versions of

If its linked then it is GPL in the linked form, otherwise you wouldn't
be allowed to link it

Alan

Re: business models [was patent stuff]

[Andre Hedrick]

There is a major problem with your argument.

"FREE" is covered in BLACK's Book.

"RELEASE" is not.  RMS specifically picked "RELEASE" in the GPL to force a
litigation in court.  "RELEASE" is subjective to the reviewing body.

Cheers,

On 28 May 2002, Alan Cox wrote:

> On Tue, 2002-05-28 at 18:13, Adam J. Richter wrote:
> > 	You could license all programs that consist entirely of
> > free software.  That way, BSD, LGPL, and MPL software that did
> > not link in proprietary software would be allowed too, but your
> > example of a proprietary program that linked in the BSD'ed
> > libpatent.o/c would not be covered by this permission.
> 
> Define "free software" using only legally defined phrases which have
> precedent. In fact put four people in a room and get them to define free
> software.
> 
> > software would address issues like GPL'ed software that contains
> > content that is GPL compatible but not GPL'ed, future versions of
> 
> If its linked then it is GPL in the linked form, otherwise you wouldn't
> be allowed to link it
> 
> Alan
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Andre Hedrick
LAD Storage Consulting Group

Re: business models [was patent stuff]

[David Weinehall]

On Tue, May 28, 2002 at 07:22:16PM +0100, Alan Cox wrote:
> On Tue, 2002-05-28 at 18:13, Adam J. Richter wrote:
> > 	You could license all programs that consist entirely of
> > free software.  That way, BSD, LGPL, and MPL software that did
> > not link in proprietary software would be allowed too, but your
> > example of a proprietary program that linked in the BSD'ed
> > libpatent.o/c would not be covered by this permission.
> 
> Define "free software" using only legally defined phrases which have
> precedent. In fact put four people in a room and get them to define free
> software.
> 
> > software would address issues like GPL'ed software that contains
> > content that is GPL compatible but not GPL'ed, future versions of
> 
> If its linked then it is GPL in the linked form, otherwise you wouldn't
> be allowed to link it

An interesting question is whether the licensing terms used for this
patent would pass the DFSG (Debian Free Software Guidelines)?!


/David
  _                                                                 _
 // David Weinehall <tao@acc.umu.se> /> Northern lights wander      \\
//  Maintainer of the v2.0 kernel   //  Dance across the winter sky //
\>  http://www.acc.umu.se/~tao/    </   Full colour fire           </

Re: business models [was patent stuff]

[Greg KH]

On Wed, May 29, 2002 at 11:44:45PM +0200, David Weinehall wrote:
> 
> An interesting question is whether the licensing terms used for this
> patent would pass the DFSG (Debian Free Software Guidelines)?!

According to the folks on debian-legal, there's already other portions
of the kernel that do not pass those guidelines :)

greg k-h

Re: business models [was patent stuff]

[Adam J. Richter]

Alan Cox wrote:
>On Tue, 2002-05-28 at 18:13, Adam J. Richter wrote:
>>       You could license all programs that consist entirely of
>> free software.  That way, BSD, LGPL, and MPL software that did
>> not link in proprietary software would be allowed too, but your
>> example of a proprietary program that linked in the BSD'ed
>> libpatent.o/c would not be covered by this permission.
>
>Define "free software" using only legally defined phrases which have
>precedent. In fact put four people in a room and get them to define free
>software.

	Many if not all legal documents contain more than
"only legally defined phrases which have precedent."  I'm sure Red Hat
has signed many.  You can reasonably find a definition that covers 99%
of what people consider free software and make subsequent grants later.
In the other direction, if you accidentally include some less than
free software, that should not matter much if you are only taking out
these patents for "defensive" purposes.

	Example definitions might be: "public domain or any license
certified by the Open Software Initiative", "a license that has
no more restrictions than version 2 of the GNU General Public License
as published by the Free Software Foundation ("or any subsequent
version"?)."  You could also cut and paste from OSI or Debian bullet
items.

	More importantly, licensing patents only for pure GPL'ed use
is unlikely to become a norm that you can expect broad adoption of
in free software businesses, as many of them tend to be proponents of
slightly different copying permissions.  If we have a bunch of patents
licensed for GPL-only, another bunch for MPL-only, another bunch for
pure-BSD only, then the patent proliferation that I described
yesterday will still probably occur.

	You have a fleeting opportunity to possibly head most of this
off, but you have to look beyond just your favorite license.  Many
developers and even companies' managements identify strongly with their
favorite licenses, and feel personally about their ability to develop
free software under those licenses.  If the GPL developers don't shield
the Apache developers, the X developers, the BSD developers, and the
MPL developers so that their ability to continue with the free software
portion of their activities has been respected, do you really think
they'll shield GPL development from their patents?

Adam J. Richter     __     ______________   575 Oroville Road
adam@yggdrasil.com     \ /                  Milpitas, California 95035
+1 408 309-6081         | g g d r a s i l   United States of America
                         "Free Software For The Rest Of Us."

Re: business models [was patent stuff]

[Alan Cox]

On Wed, 2002-05-29 at 04:21, Adam J. Richter wrote:
> 	Example definitions might be: "public domain or any license
> certified by the Open Software Initiative", "a license that has


The OSI approves things like the BSD license which is a convenient open
door for anyone to use such patents in proprietary code with a tiny
useless BSD licensed library. That doesn't further free software, it 
doesn't provide opportunities for co-operative cross licensing to get
GPL rights to other patents from the proprietary world and so forth.
Being in a position where as part of the normal processes of proprietary
software design we can get other companies to give free software access
to some patents in return for access to use some our stuff proprietary
is a win.

I would worry much more about the million odd patents IBM have, where
IBM have no general statement of this nature than the Red Hat ones.
Perhaps once the Red Hat statement is published IBM can be persuaded to
show willing ?

> 	More importantly, licensing patents only for pure GPL'ed use
> is unlikely to become a norm that you can expect broad adoption of
> in free software businesses, as many of them tend to be proponents of
> slightly different copying permissions.  If we have a bunch of patents
> licensed for GPL-only, another bunch for MPL-only, another bunch for
> pure-BSD only, then the patent proliferation that I described
> yesterday will still probably occur.

I would agree to an extent. Certainly purely GPL is excluding stuff
which has identical 'all of package' rules like db3, Qt free editions,
and much of KDE.

> free software under those licenses.  If the GPL developers don't shield
> the Apache developers, the X developers, the BSD developers, and the
> MPL developers so that their ability to continue with the free software
> portion of their activities has been respected, do you really think
> they'll shield GPL development from their patents?

There is little evidence of that having happened with code, its also
possible to extend lists of licenses, clarify them for specific product
and so forth.

Alan

Re: business models [was patent stuff]

[Jonathan Corbet]

Alan said:

> I would worry much more about the million odd patents IBM have, where
> IBM have no general statement of this nature than the Red Hat ones.
> Perhaps once the Red Hat statement is published IBM can be persuaded to
> show willing ?

For what it's worth, the Red Hat statement is at:

	http://www.redhat.com/legal/patent_policy.html

No patent enforcement against you if you're using a Red Hat-approved open
source license - their list does not include BSD.

jon

Jonathan Corbet
Executive editor, LWN.net
corbet@lwn.net

Re: business models [was patent stuff]

[Jamie Lokier]

Jonathan Corbet wrote:
> Alan said:
> > I would worry much more about the million odd patents IBM have, where
> > IBM have no general statement of this nature than the Red Hat ones.
> > Perhaps once the Red Hat statement is published IBM can be persuaded to
> > show willing ?
> 
> For what it's worth, the Red Hat statement is at:
> 
> 	http://www.redhat.com/legal/patent_policy.html
> 
> No patent enforcement against you if you're using a Red Hat-approved open
> source license - their list does not include BSD.

The Red Hat-approved licences don't include LGPL either.

Red Hat's definition of "Open Source/Free Software" ("means any software
which is licensed under an Approved License.") could do with some
clarification.

I wonder what their stance is if I dual license under BSD and GPL.  Do I
have to be explicit about it?  Or is that too loose - must it be GPL only?

I wonder if that means I have to explicitly list my high performance web
server, which takes advantage of the patented directory lookup of
course, as dual licensed under GPL and LGPL, or if that is taken as
implied.

etc.

-- Jamie

Re: business models [was patent stuff]

[Nicholas Knight]

----- Original Message -----
From: "Jamie Lokier" <lk@tantalophile.demon.co.uk>
To: "Jonathan Corbet" <corbet-lk@lwn.net>
Cc: <linux-kernel@vger.kernel.org>; "Alan Cox" <alan@lxorguk.ukuu.org.uk>
Sent: Wednesday, May 29, 2002 9:15 AM
Subject: Re: business models [was patent stuff]


> Jonathan Corbet wrote:
> > Alan said:
> > > I would worry much more about the million odd patents IBM have, where
> > > IBM have no general statement of this nature than the Red Hat ones.
> > > Perhaps once the Red Hat statement is published IBM can be persuaded
to
> > > show willing ?
> >
> > For what it's worth, the Red Hat statement is at:
> >
> > http://www.redhat.com/legal/patent_policy.html
> >
> > No patent enforcement against you if you're using a Red Hat-approved
open
> > source license - their list does not include BSD.
>
> The Red Hat-approved licences don't include LGPL either.
>
> Red Hat's definition of "Open Source/Free Software" ("means any software
> which is licensed under an Approved License.") could do with some
> clarification.
>

The definition of "defensive patent" could do with some reassesment.

The only way I can see a software patent as truely defensive, is if it's
explicitly
licensed for use by anyone, anywhere, for anything, under any software
license.
The only thing it should be used to defend against is someone else patenting
it
and denying Red Hat the right to use it, it shouldn't be used to prevent
others,
no matter who they are, from using it.

If there's a proper name for what Red Hat now has, it's "Free Software
Foundation Compliant" and "Linux Specific". This isn't a defensive patent,
it's
a patent filed for the express purpose of preventing systems besides Linux
from taking advantage of the techniques covered by the patent.

Off the top of my head, I can think of only one at least somewhat-known OS
licensed under any of their "approved" licenses, and that's AtheOS (GPL).
Relatively few people use AtheOS yet, and I doubt that will change soon.

On the other hand, at the very front of my mind at all times are the names
of
three fairly well-known systems that are in fairly wide-spread use and are
decidedly open source:
FreeBSD (significant widespread use)
NetBSD
OpenBSD

I don't care what anybody's personal opinion of these systems is, they are
Free and Open Souce in every sense of the terms, and a lot of people use
them. There's just one problem, they're frequently in direct competition
with Linux. And since Red Hat is decidedly the front-man of the U.S.
Linux distribution market, they are directly in competition with Red Hat.
Rather convenient that they can't make use of the patent, isn't it?

Of course, if Red Hat went to the logical conclusion of opening the patent
for use, Microsoft would end up able to use the patent, and I doubt anyone
but Microsoft would really want that. Hence this solution:

Explicitly license the patent to anyone and everyone, EXCEPT Microsoft.

Why do I not have a problem with not allowing Microsoft to use it?
Simple, they've been caught, in very recent history, using very illiegal
business practices. They shouldn't have the right to use their *own*
patents.


Of course, this still leaves it open to other competitors, and we can't have
that, now can we?


> I wonder what their stance is if I dual license under BSD and GPL.  Do I

You certainly could, but I doubt there would be any point. The software
has to be under the terms of the GPLv2, IBM PL v1, CPL v0.5, QPLv1,
or any of Red Hat's Open Source licenses in order for the  patented
method to be used.

> have to be explicit about it?  Or is that too loose - must it be GPL only?
>
> I wonder if that means I have to explicitly list my high performance web
> server, which takes advantage of the patented directory lookup of
> course, as dual licensed under GPL and LGPL, or if that is taken as
> implied.

I could see splitting components up, some under GPL, some under LGPL.
But then you have the issue you're already familier with with FSMLabs'
patent, calls to the GPL'd code may have to be made by GPL'd code only.

>
> etc.
>
> -- Jamie

Re: business models [was patent stuff]

[Perry The Cynic]

On Wed, May 29, 2002 at 10:29:40AM -0700, Nicholas Knight wrote:
> The definition of "defensive patent" could do with some reassesment.
> 
> The only way I can see a software patent as truely defensive, is if it's
> explicitly
> licensed for use by anyone, anywhere, for anything, under any software
> license.
> The only thing it should be used to defend against is someone else patenting
> it
> and denying Red Hat the right to use it, it shouldn't be used to prevent
> others,
> no matter who they are, from using it.

There is a much more aggressive use that arguably is still defensive. It's
a formal codification of the way patents are often used by corporations
these days. It goes like this:

"You have a license to use this patent, without fees or restrictions, in
any of your products as long as you do not raise any claims from any of
*your* patents, of any form, against any of our products, except when
such patent includes license terms that are no worse than this license."

In real life, this is hellaciously hard to express in such a way that it
can't be worked around. (One obvious attack is to fork off an "independent"
subsidiary that raises patent claims on your behalf.) Which is why this kind
of strategy is usually implemented by staff lawyers playing "tit for tat"
with the other guy's staff lawyers.

Cheers
  -- perry
---------------------------------------------------------------------------
Perry The Cynic                                             perry@cynic.org
To a blind optimist, an optimistic realist must seem like an Accursed Cynic.
---------------------------------------------------------------------------

Re: business models [was patent stuff]

[Rob Landley]

On Wednesday 29 May 2002 09:15 am, Alan Cox wrote:

> I would worry much more about the million odd patents IBM have, where
> IBM have no general statement of this nature than the Red Hat ones.
> Perhaps once the Red Hat statement is published IBM can be persuaded to
> show willing ?

I'm is off dealing with patents from Sun, Sony, Apple, and Microsoft.  Their 
patent portfolio is like a nuclear arsenal: mutual assured destruction, so 
nobody actually uses their stockpile.

The down side of the GPL language handling patents implies that they more or 
less put a patent into the public domain.  I.E:

> 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. 

However, looking at the text of the license, it seems that it might be 
possible to issue a patent license allowing use for the patent only in GPL 
licensed programs.  (Thus IBM could issue a blanket statement saying that any 
of its patents may be used in GPL code, without necessarily unilaterally 
disarming itself with regards to Sony's patents.)

Opinions from Lessig and Stallman and a few random lawyers would be required 
to really get the issue decently clarified, though.  (It would be best if 
such a license could be a one-paragraph addendum to the GPL itself, perhaps a 
GPL V2.1?)  Using the GPL as a patent pool would be extremely useful to free 
software in the long run...

> > 	More importantly, licensing patents only for pure GPL'ed use
> > is unlikely to become a norm that you can expect broad adoption of
> > in free software businesses, as many of them tend to be proponents of
> > slightly different copying permissions.  If we have a bunch of patents
> > licensed for GPL-only, another bunch for MPL-only, another bunch for
> > pure-BSD only, then the patent proliferation that I described
> > yesterday will still probably occur.
>
> I would agree to an extent. Certainly purely GPL is excluding stuff
> which has identical 'all of package' rules like db3, Qt free editions,
> and much of KDE.

That said, a GPL-only patent license looks like the minimal set of permission 
grants required for a patent to be compatable with the GPL, and such a setup 
does start to turn the GPL into a patent pool anyone can join, which isn't 
entirely a bad thing...

And the existence of a less permissive license doesn't prevent the existence 
of a MORE permissive license.  (Dual GPL/BSD, for example...)

The whole patent issue is largely a matter of big fish using patent pools to 
exclude small fish.  A free software patent pool has been proposed repeatedly 
but never had the critical mass to get off the ground.  (On a related note, I 
entirely understand Red Hat seeking defensive patents due to the mexican 
standoff nature of the modern patent landscape, but hope they plan to use 
them wisely for the good of the community...)

Rob

Re: business models [was patent stuff]

[Adam J. Richter]

Alan Cox writes:
>On Wed, 2002-05-29 at 04:21, Adam J. Richter wrote:
>> 	Example definitions might be: "public domain or any license
>> certified by the Open Software Initiative", "a license that has


>The OSI approves things like the BSD license which is a convenient open
>door for anyone to use such patents in proprietary code with a tiny
>useless BSD licensed library. [...]

	I addressed that previously: you can license just the case
where the entire program is free software.  Patents restrict use, not
just copying.  So, it is hard to argue against the enforceability of
"user does the link" subversion when it comes to patents, which I
infer is your principal objection.


>I would worry much more about the million odd patents IBM have, [...]

	I addressed that previously:

| 	Although other companies today already have many patents that
| | they could argue are infringed by certain free software components.
| The Linux patents are different as a practical matter, however, in
| that the chance of prevailing in that argument will be greater when
| if alleged infringer is using the code for which the patent was
| originally submitted.

[...]
>> If the GPL developers don't shield
>> the Apache developers, the X developers, the BSD developers, and the
>> MPL developers so that their ability to continue with the free software
>> portion of their activities has been respected, do you really think
>> they'll shield GPL development from their patents?

>There is little evidence of that having happened with code,

	Linux distribution vendors make a lot of money from combining
and selling the work of these other free software development
communities.  These other communities currently do a lot of work to
ensure that their software runs under our GPL'ed kernel and LGPL'ed C
library.  The GPL developers are supported in the other direction too:
plenty of GPL'ed software runs under BSD-permission platforms like
tcl, or Apache or on BSD.  Even with the Linux kernel, you have Dave
Hinds relicensing his MPL pcmcia drivers for GPL linking, and sixty
two modules in linux-2.5.18 contain MODULE_LICENSE declarations of
"Dual BSD/GPL", plus two others that are just BSD.

	You would have no Linux distribution if you only shipped
GPL'ed software, or even if the non-GPL free software communities
stopped their continuing substantial efforts to ensure that their
applications run Linux, compile under GCC and binutils, and
interoperate with a host of other GPL'ed components.

	What more should the non-GPL free software communities have
done that now justifies your not licensing your patents for their free
software development?

>its also
>possible to extend lists of licenses, clarify them for specific product
>and so forth.

	If you "extend lists of licenses" _at the outset_, you will
discourage patenting of the non-GPL software you use.  Otherwise, I
think the practice will quickly proliferate if other free software
companies see that their users and contributors tolerate Red Hat
taking out patents on free software implementations and licensing it
GPL-only.  I think subsequent extensions would have less of of a
standards setting effect.


Adam J. Richter     __     ______________   575 Oroville Road
adam@yggdrasil.com     \ /                  Milpitas, California 95035
+1 408 309-6081         | g g d r a s i l   United States of America
                         "Free Software For The Rest Of Us."