Re: Honest does not pay here ...

Re: Honest does not pay here ...

[Adam J. Richter]

Larry McVoy wrote:
>The day that you take over the day to day running of the kernel effort
>is the day that you get to enforce your rules.  Linus provides huge
>value that neither you nor anyone else provides.  So he makes the rules.
>If you don't like it, make a stink and withdraw your patches.  They'll
>get replaced right away.

	If you think no Linux copyright owner other than Linus can
enforce their copyrights, then your understanding of copyright law
is different from mine.

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: Honest does not pay here ...

[Andre Hedrick]

On Sun, 5 Jan 2003, Adam J. Richter wrote:

> Larry McVoy wrote:
> >The day that you take over the day to day running of the kernel effort
> >is the day that you get to enforce your rules.  Linus provides huge
> >value that neither you nor anyone else provides.  So he makes the rules.
> >If you don't like it, make a stink and withdraw your patches.  They'll
> >get replaced right away.
> 
> 	If you think no Linux copyright owner other than Linus can
> enforce their copyrights, then your understanding of copyright law
> is different from mine.

Adam,

This was just dropped in my lap.

---------------------------
Add IBM to this list with their QDIO drivers for OSA-Express on s390(x).

On S/390, IBM did the same as Andre. LCS drivers have been binary-only in
the beginning and have been released under GPL later after 3172 and other
LCS devices became obsoleted by IBM. QDIO is their current standard and
they don't want to open this channel protocol to their competition.

So i think, there is enough precedence even from the "big guys".
---------------------------

Now that you have clearly stated you are revoking usage of your work
except for GPL-ONLY.

Please list the files and functions you claim, under this decision.
If you are the sole copyright holder of a file it is clean and clear.
If you are not then you have a problem to settle with the joint owners.
I will ignore your claim on shared copyrights until you have a settlement,
if all persons holding copyright ownership agree with you then ...

How soon will there be a patch-war for everyone to stake a claim, and
screw the pooch ?

Cheers,

Andre Hedrick
LAD Storage Consulting Group

Re: Honest does not pay here ...

[Trever L. Adams]

On Sun, 2003-01-05 at 15:29, Andre Hedrick wrote:
> Now that you have clearly stated you are revoking usage of your work
> except for GPL-ONLY.
> 
> Please list the files and functions you claim, under this decision.
> If you are the sole copyright holder of a file it is clean and clear.
> If you are not then you have a problem to settle with the joint owners.
> I will ignore your claim on shared copyrights until you have a settlement,
> if all persons holding copyright ownership agree with you then ...
> 
> How soon will there be a patch-war for everyone to stake a claim, and
> screw the pooch ?
> 
> Cheers,
> 
> Andre Hedrick
> LAD Storage Consulting Group

I am ardent supporter of the GPL.  I do have some problems with what
some people are doing (particularly Nvidia, namely because I believe if
I pay for hardware, I pay for the right to use it and to have the specs
on how to use it... i.e. they don't release programming info).  However,
Linus has allowed for binary only modules.

All said and done, while I wish that all modules were GPLed, where Andre
and others do eventually release their code as GPL after development is
paid for... I can't fault them.  More power to them for doing
sustainable development (Because people won't help cover development
costs without this kind of model) and in 18 months or so (From driver
creation) more power to us because they give it away under the GPL.

Now, if it is clear they are making derivitive works, this may be a
problem.  Since I have little idea about if they are, I will leave this
to those that do.

Thank you for what you do do Andre, thank you to all who do things in a
decent way, whether that is GPL from the beginning (Those who can) or
those who delay the GPL until they can afford to do it.

Trever
--
"Life is a comedy for those who think and a tragedy for those who feel."
-- Unknown

Re: Honest does not pay here ...

[Andrew McGregor]

--On Sunday, January 05, 2003 17:28:52 -0500 "Trever L. Adams" 
<tadams-lists@myrealbox.com> wrote:

>
> I am ardent supporter of the GPL.  I do have some problems with what
> some people are doing (particularly Nvidia, namely because I believe if
> I pay for hardware, I pay for the right to use it and to have the specs
> on how to use it... i.e. they don't release programming info).  However,
> Linus has allowed for binary only modules.

>
> Trever

I've had some discussion with an ex-NVidia guy who was there while they 
were doing the driver release.

They wanted to dual GPL/BSD license the kernel part in the first place, 
then they realised they had a problem.  They don't own the copyright on all 
that code themselves, nor do they have the right to redistribute specs for 
all of the hardware without NDA, because it consists in part of purchased 
'IP blocks' (as hardware people call libraries).  So in the end they've 
opened up as far as they were allowed by preexisting constraints.

Remember, the hardware was not constructed with an open source driver in 
mind.  It's fairly easy to build hardware which can have open source 
drivers (you choose your IP block vendors carefully), but NVidia did not do 
that in the first place, and now they are stuck.

So your belief about hardware is just plain false, unfortunately.  You're 
free not to buy their hardware, but I don't think you are being fair to dis 
them when they appear to have gotten the point of open source but been 
stymied by other vendors.  NVidia do try hard to give you the right to use 
their stuff with Linux, but there is only so far they can go.

I expect if Linux makes them enough money, they might buy the rights they 
don't have, and release the driver in full.  But don't expect that to 
happen soon, because if you think proprietary software licenses can be 
expensive, you haven't seen hardware.

Andrew

Re: Honest does not pay here ...

[Trever L. Adams]

On Sun, 2003-01-05 at 19:01, Andrew McGregor wrote:
> I've had some discussion with an ex-NVidia guy who was there while they 
> were doing the driver release.
> 
> They wanted to dual GPL/BSD license the kernel part in the first place, 
> then they realised they had a problem.  They don't own the copyright on all 
> that code themselves, nor do they have the right to redistribute specs for 
> all of the hardware without NDA, because it consists in part of purchased 
> 'IP blocks' (as hardware people call libraries).  So in the end they've 
> opened up as far as they were allowed by preexisting constraints.
> 
> Remember, the hardware was not constructed with an open source driver in 
> mind.  It's fairly easy to build hardware which can have open source 
> drivers (you choose your IP block vendors carefully), but NVidia did not do 
> that in the first place, and now they are stuck.
> 

I was not aware of all of this as being the case.  I am sorry they are
stuck in such a bad position.  It does raise my opinion of them quite a
bit.

> So your belief about hardware is just plain false, unfortunately.  You're 

No, my belief may not reflect what is, but that doesn't make it false. 
I know there were, at least until recently, countries that actually
dictated what I said by law.  Again, how much did reality follow the
laws... your guess would probably be better than mine.

> free not to buy their hardware, but I don't think you are being fair to dis 
> them when they appear to have gotten the point of open source but been 
> stymied by other vendors.  NVidia do try hard to give you the right to use 
> their stuff with Linux, but there is only so far they can go.
> 
> I expect if Linux makes them enough money, they might buy the rights they 
> don't have, and release the driver in full.  But don't expect that to 
> happen soon, because if you think proprietary software licenses can be 
> expensive, you haven't seen hardware.

I expect that IP is expensive to buy.  Anyway, thank you for explaining
the Nvidia situation to me.  I really hope they do figure out some
things soon.  (Even if that is just how to make kernels with their
modules loaded more stable and easier to debug.)

Trever
--
"What makes his world so hard to see clearly is not its strangeness but
its usualness. Familiarity can blind you." -- Robert M. Pirsig

Re: Honest does not pay here ...

[Larry McVoy]

On Sun, Jan 05, 2003 at 12:21:57PM -0800, Adam J. Richter wrote:
> Larry McVoy wrote:
> >The day that you take over the day to day running of the kernel effort
> >is the day that you get to enforce your rules.  Linus provides huge
> >value that neither you nor anyone else provides.  So he makes the rules.
> >If you don't like it, make a stink and withdraw your patches.  They'll
> >get replaced right away.
> 
> 	If you think no Linux copyright owner other than Linus can
> enforce their copyrights, then your understanding of copyright law
> is different from mine.

What I think is that Linus provides a lot more value than you and all
the other yadda-yadda-yadda folks combined.  Because he provides that
value, people follow his lead.  If you want to try and rally a bunch of
people to say that their contributions don't go along with Linus' rules,
by all means, go for it.  See how far you get.

Copyright law, this law, that law, none of it means shit compared to
real work and Linus does way more for the kernel than you ever will.
Everyone recognizes his leadership skills, most of us have disagreed
violently with Linus on more than one occasion, yet we still follow
his lead.  Why?  Because he's the best thing we have going.  He's not
a zealot, he's a reasonable guy, you can get him to change his mind if
he's wrong, he's an extremely effective leader.  You would do well to
emulate him but you are choosing to cause a fuss.   I don't think people
are going to line up behind you, you don't have the same leadership
qualities and people aren't that unhappy with what Linus is doing.
I could be wrong but if I'm not what do you hope to accomplish other
than wasting a lot of time?

The cool thing is that if the community ever changes their mind about all
this, the kernel is GPLed and a new leader can emerge.  Until that time,
how about you go back to coding rather complaining?
-- 
---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Re: Honest does not pay here ...

[Stephen Satchell]

At 07:15 PM 1/5/03 -0500, Trever L. Adams wrote:
>I expect that IP is expensive to buy.  Anyway, thank you for explaining
>the Nvidia situation to me.  I really hope they do figure out some
>things soon.  (Even if that is just how to make kernels with their
>modules loaded more stable and easier to debug.)

Trever,

Just because Nvidia incorporate API-related intellectual property in their 
current products doesn't mean that the next product they do will have the 
same IP restrictions.  Assuming that government and corporations continue 
to examine and adopt Linux over other desktop systems, Nvidia's project 
managers will have some incentive to look at this issue more closely.

As I see it, Andre's problems are a little tougher because there isn't a 
choice of alternatives to the IP that he incorporates in his product, if 
I'm reading his contributions correctly.  (I've not followed the arguments 
closely, so I could very well be in need of disabusement of my incorrect 
notions.)

I'm surprised Richard Stallman didn't remind everyone of the thing that 
started the whole argument for free software:  his inability to drive a 
laser printer because of closed, unpublished specifications.  I won't put 
words in his mouth (RMS is more than capable of speaking for himself) but 
his concern is if the company goes away and there are problems with the 
binary-only module, then people will be forced to junk the hardware or live 
with the problems.

Ok, I'll go back to my hole now.

Satch



-- 
The human mind treats a new idea the way the body treats a strange
protein:  it rejects it.  -- P. Medawar
This posting is for entertainment purposes only; it is not a legal opinion.

Re: Honest does not pay here ...

[Ian Molton]

On 05 Jan 2003 19:15:24 -0500
"Trever L. Adams" <tadams-lists@myrealbox.com> wrote:

> 
> I was not aware of all of this as being the case.  I am sorry they are
> stuck in such a bad position.  It does raise my opinion of them quite
> a bit.

If you happen to believe that...

Re: Honest does not pay here ...

[jw schultz]

On Sun, Jan 05, 2003 at 12:29:16PM -0800, Andre Hedrick wrote:
> Please list the files and functions you claim, under this decision.
> If you are the sole copyright holder of a file it is clean and clear.
> If you are not then you have a problem to settle with the joint owners.
> I will ignore your claim on shared copyrights until you have a settlement,
> if all persons holding copyright ownership agree with you then ...
> 
> How soon will there be a patch-war for everyone to stake a claim, and
> screw the pooch ?

Andre, you just made the point i've been thinking of for
some time.  Let the people who object to having an interface
for binary-only modules stake out their code.  If you have
no code you have no legal standing.  Identify what code you
won't allow the binary modules to use.

I'm a lurker here.  I just don't feel very itchy.  But if
someone actually declares needed code off-limits for the
binary modules i and others could very quickly develop
serious itches to scratch.  The patch-war will be people
replacing code belonging to the extremists.

-- 
________________________________________________________________
	J.W. Schultz            Pegasystems Technologies
	email address:		jw@pegasys.ws

		Remember Cernan and Schmitt

Re: Honest does not pay here ...

[Andrew McGregor]

--On Sunday, January 05, 2003 19:15:24 -0500 "Trever L. Adams" 
<tadams-lists@myrealbox.com> wrote:

> On Sun, 2003-01-05 at 19:01, Andrew McGregor wrote:
>> I've had some discussion with an ex-NVidia guy who was there while they
>> were doing the driver release.

> I expect that IP is expensive to buy.  Anyway, thank you for explaining
> the Nvidia situation to me.  I really hope they do figure out some
> things soon.  (Even if that is just how to make kernels with their
> modules loaded more stable and easier to debug.)

Well, there are people working on it.  I seem to be one of the few people 
for whom a 2.5.x + nvidia kernel really scratches an itch.  (for me, it's 
my laptop + kernel IPSEC)

There are patches at www.minion.de that make things stable up to 2.5.53. 
.54 broke it again, but that will be fixed shortly.

Andrew

Patches for nVidia drivers + 2.5.54 (Was Re: Honest does not pay here ...)

[Brian Davids]

Andrew McGregor wrote:

> Well, there are people working on it.  I seem to be one of the few 
> people for whom a 2.5.x + nvidia kernel really scratches an itch.  (for 
> me, it's my laptop + kernel IPSEC)
> 
> There are patches at www.minion.de that make things stable up to 2.5.53. 
> .54 broke it again, but that will be fixed shortly

It was fixed with an update on www.minion.de on January 3.  I've been 
running 2.5.54 w/ the nVidia drivers since then... ;)


Brian Davids

Re: Honest does not pay here ...

[Trever L. Adams]

On Sun, 2003-01-05 at 20:43, Stephen Satchell wrote:
> As I see it, Andre's problems are a little tougher because there isn't a 
> choice of alternatives to the IP that he incorporates in his product, if 
> I'm reading his contributions correctly.  (I've not followed the arguments 
> closely, so I could very well be in need of disabusement of my incorrect 
> notions.)
> 
> I'm surprised Richard Stallman didn't remind everyone of the thing that 
> started the whole argument for free software:  his inability to drive a 
> laser printer because of closed, unpublished specifications.  I won't put 
> words in his mouth (RMS is more than capable of speaking for himself) but 
> his concern is if the company goes away and there are problems with the 
> binary-only module, then people will be forced to junk the hardware or live 
> with the problems.
> 
> Ok, I'll go back to my hole now.
> 
> Satch

I guess a question then might be this:  Andrea, I understand your stance
of needing to make a decent living and fund development.  I think Satch
has a point about the company going away (or the Bus problem... as in
something happens to you).  Is there any way you can feasibly (legal and
monetary concerns included) do a kind of code escrow so if such happens,
your code becomes GPL/BSD?  

BTW, I may be somewhat out of understanding here.  From what I have been
reading it seems the following is true:

1) Andre is making drivers for hardware or protocols
2) He is making them closed until he recoups his costs (I saw 18 mos
somewhere as the time needed...)
3) He then will open them up

If I am wrong, sorry, but this should say where I am seeing all this
from.

Also, I see the following...

1) The problem lies with him including kernel headers (I didn't think
magic numbers and such were really coverable by copyright... so unless
we are talking macros... where is the problem).
2) Interfaces are reverse engineer-able under US law for
interoperability purposes (DMCA may have muddied this)
3) The Interface calls (sys-calls etc.) are LGPL...

So where is the real problem here?

Trever
--
One O.S. to rule them all, One O.S. to find them. One O.S. to bring them
all and in the darkness bind them.

Re: Honest does not pay here ...

[Andre Hedrick]

On 6 Jan 2003, Trever L. Adams wrote:

> I guess a question then might be this:  Andrea, I understand your stance
> of needing to make a decent living and fund development.  I think Satch
> has a point about the company going away (or the Bus problem... as in
> something happens to you).  Is there any way you can feasibly (legal and

Oh, like when I totalled my Porsche 928 and should have died hitting nose
first into a retaining wall @ 5Krpm in 3rd after somebody ran me off the
interstate ?

> monetary concerns included) do a kind of code escrow so if such happens,
> your code becomes GPL/BSD?  

This is all dependent on the issue below.

> BTW, I may be somewhat out of understanding here.  From what I have been
> reading it seems the following is true:
> 
> 1) Andre is making drivers for hardware or protocols
> 2) He is making them closed until he recoups his costs (I saw 18 mos
> somewhere as the time needed...)
> 3) He then will open them up
> 
> If I am wrong, sorry, but this should say where I am seeing all this
> from.
> 
> Also, I see the following...
> 
> 1) The problem lies with him including kernel headers (I didn't think
> magic numbers and such were really coverable by copyright... so unless
> we are talking macros... where is the problem).

If this is a problem, there will be no project to open source.

> 2) Interfaces are reverse engineer-able under US law for
> interoperability purposes (DMCA may have muddied this)
> 3) The Interface calls (sys-calls etc.) are LGPL...

(target)
It calls net, mm, slab, timer, spinlock, semaphore, scsi (structs and
about 5 or 6 functions), misc/char device, module the basics.
ZERO .c files period for the target protocol transport.  ata+sata can
substitute for scsi if selected.

(initiator)
All of the above, and the critical point is "scsi_module.c", but I can
write my own version of the functions below.

This is why I initially put forward this half of the protocol as an issue.
Now since it is only combining and not changing or derived from that file,
GPL has no say or position of adding that copyright to mine.
Regardless, this is a concern as to be totally above board.

Now there are several vendors who attach this file by one means or
another, so this is very gray.

Additionally there is no license in the file, only Copyright.

This is a concern but the logic is simple,

Actually I just realized I have copyright ownership of similar logic in
another file.  WOOHOO !

Therefore, I now no longer have an issue with "scsi_module.c", so I am
totally .c free regardless.  I had taken the precaution to place the the
include line in one of my .h files to insure that object was created first
and my .o's for the protocol are summed first then the two are LD'd at the
very end.  Now I do not have this difficult makefile and order of
operations to worry about now!

Thanks :-)

> So where is the real problem here?

Paranoia of accidently doing something wrong.
Concerned to the Nth degree that I follow the rules of binary module usage
toed exactly to the line, not short not over but exact.  Zero margin for
error is allowed.  Since I know better, I have to be that more careful.

Cheers,

Andre Hedrick
LAD Storage Consulting Group

If there are no problems and I do not have to switch platforms, I have my
sights set on the next generation of storage about three years out.  iSCSI
is expected to fully mature in 2006/2008, its value is reduced.  My
chances of recovering my costs are slipping right now, as customers are
waiting now.