Re: [OpenBIOS] Intel holding back information

[Timothy J. Massey]

On Sun, 8 Aug 1999 12:59:29 +0200 (MET DST), Niklas Ekstr�m wrote:
> I have been thinking about doing that, but wouldn't that make me unable to
> write the bios after that (legally that is) ?

You got it.  If you **EVER** in your entire LIFE disassemble a BIOS, you are
then unable to program a similar BIOS without being open to a legal
challenge.

Even if your BIOS bears no resemblance to the one you disassembled, because
you *did* disassemble it, the makers of the BIOS you disassembled can say to
you, "But you saw our code and did it yourself.  That's illegal."  And
they'd be right.

Tim Massey

-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[James Oakley]

"Timothy J. Massey" wrote:
> Even if your BIOS bears no resemblance to the one you disassembled, because
> you *did* disassemble it, the makers of the BIOS you disassembled can say to
> you, "But you saw our code and did it yourself.  That's illegal."  And
> they'd be right.

This is very true. A couple of years back, some guys who worked for
Matrox quit and started their own video chipset company. Matrox
successfully blocked their entry into the market because they saw
proprietary Matrox engineering information.

I completely agree with Matrox, too. Seeing it is enough. It is pretty
much impossible to implement a clean room solution if you saw
information from another company.

Therefore, OpenBIOS developers, NEVER DISASSEMBLE A PROPRIETARY BIOS.
That would compromise the whole effort.

James Oakley
-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[David J. Coffin]

On Sun, Aug 08, 1999 at 05:35:25PM -0230, James Oakley wrote:
> "Timothy J. Massey" wrote:
> > Even if your BIOS bears no resemblance to the one you disassembled, because
> > you *did* disassemble it, the makers of the BIOS you disassembled can say to
> > you, "But you saw our code and did it yourself.  That's illegal."  And
> > they'd be right.

Reverse engineering is very legal in Sweden, although it
may soon be illegal in the USA.  There wouldn't *be* any BIOS
industry today if Phoenix hadn't reverse-engineered IBM's code.

So Phoenix, which now owns Award, is going to prosecute
some college student in Sweden for doing exactly what they did.
Not likely.
> This is very true. A couple of years back, some guys who worked for
> Matrox quit and started their own video chipset company. Matrox
> successfully blocked their entry into the market because they saw
> proprietary Matrox engineering information.

Not the same thing.  Those engineers had to sign stacks of
non-disclosure and non-competition agreements before they could
see Matrox's designs.  They clearly violated those contracts.

Niklaus would be taking a bigger risk if he signed Intel's
NDA.  Right now, he has no legal obligation to protect anyone's
secrets.

If a disgruntled chemist at Coca-Cola gave you their secret
formula, he would be in deep trouble.  But no one could stop you
from selling cola drink (you couldn't *call* it "Coca-Cola", of
course), or posting the formula on the web.

Dave Coffin  8/8/99
-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[James Oakley]

>  Niklaus would be taking a bigger risk if he signed Intel's
>
> NDA.  Right now, he has no legal obligation to protect anyone's
> secrets.
>  If a disgruntled chemist at Coca-Cola gave you their secret
>
> formula, he would be in deep trouble.  But no one could stop you
> from selling cola drink (you couldn't *call* it "Coca-Cola", of
> course), or posting the formula on the web.

The company will probably try to sue anyway if OpenBIOS cuts into
Award's profits. I don't think any of us has the financial resources to
counter something like that. If they don't, ok we're lucky, but I'd like
to avoid that kind of stuff. There's been a lot of bad patent stuff
going around lately so I checked something out:

http://www.patents.ibm.com/details?pn=US05732268__

That does *not* look good.

James Oakley
-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[Ed Brinker]

But there is a way and that is the way Compaq did it (I believe this is
called the clean room approach).  The OpenBios group would have to be split
into two distinct groups. The first group would reverse engineer the Award,
Phoenix, AMI etc. BIOS' to prepare a list of all of the functions performed
by the BIOS.  The second group would then prepare requirements, design and
code the BIOS from the "set of BIOS functions." If the two groups did not
communicate any information other than the BIOS functions (no details of
algorithms, code details, or implementation) then the new BIOS would be
considered "pristine" and could not sucessfully be legally challenged.

How might we implement such an approach.  It would probably require two
separate mail lists.  One for the group to reverse engineer the other BIOS
and a second for the group creating the new BIOS.  The could be no
communication between the two groups about the BIOS's other than the
"identified functions."

The first group could make no comments about the work in progress of the
second group as that could be intrepreted as providing "guidance."  The
second group    
could not ask if the function was really necessary as that would be
intrepreted as asking for "guidance."

If we are to take this route, then perhaps it is best we split soon to
avoid the process of legal "contamination."

There would be legal restrictions on the membership of each group.  We
would probably need some legal advice.  But I suspect that the
implementation group could have no knowledge of BIOS's other than what is
in the open literature.  That would be restricted to knowing about details
such the hard drive table which is described numerous magazines and books.

After the "genie was out of the bottle," and the BIOS had a "GNU" license
and hundreds of variations might become available.

So I pose the question to our leaders.  Should we do it before we become
legally "contaminated?"

At 10:13 PM 8/8/99 -0230, you wrote:
> >  Niklaus would be taking a bigger risk if he signed Intel's
> >
> > NDA.  Right now, he has no legal obligation to protect anyone's
> > secrets.
> >  If a disgruntled chemist at Coca-Cola gave you their secret
> >
> > formula, he would be in deep trouble.  But no one could stop you
> > from selling cola drink (you couldn't *call* it "Coca-Cola", of
> > course), or posting the formula on the web.
>
> The company will probably try to sue anyway if OpenBIOS cuts into
> Award's profits. I don't think any of us has the financial resources to
> counter something like that. If they don't, ok we're lucky, but I'd like
> to avoid that kind of stuff. There's been a lot of bad patent stuff
> going around lately so I checked something out:
>
> http://www.patents.ibm.com/details?pn=US05732268__
>
> That does *not* look good.
>
> James Oakley
>
> To unsubscribe: send mail to [e-mail]
> with 'unsubscribe openbios' in the body of the message

-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[William Meyer]

> How might we implement such an approach.  It would probably require two
> separate mail lists.  One for the group to reverse engineer the other BIOS
> and a second for the group creating the new BIOS.  The could be no
> communication between the two groups about the BIOS's other than the
> "identified functions."

It would require a rigid protocol. E-mail is not likely to be an acceptable
record of communications. Rigorous note keeping using traditionally accepted
methods would be essential. Those participating in coding would have to be
able to swear out affidavits attesting to their complete prior ignorance of
BIOS coding.

There are paper notebooks which serve well for documenting developments
which are potentially of legal importance. One supplier is the Laboratory
Notebook Company. The notebooks are stitched, the pages are gridded and
numbered, and each page provides space for the author and a witness to sign
and date. I've used these on projects for over 20 years, and they are
generally adopted as a standard in companies where patentability is a
consideration. They should be equally useful in documenting freedom from
infringing practices.
> If we are to take this route, then perhaps it is best we split soon to
> avoid the process of legal "contamination."

True enough, but the split can't be based merely on people's preferences.
Coders have to be free from any prior knowledge of BIOS source code. Many
won't like this, but there is no other way I am aware of to keep the group
legally untainted.

Legal advice would be essential. A lawyer (with considerable knowledge of
trademark, trade secret, and patent law) should draft a protocol, and should
also specify the considerations for membership in the coding group.

One question which occurs to me is whether there is still the potential for
problems from 3rd party contributions subsequent to publication of an open
BIOS. After all, the intellectual property of IBM, Phoenix/Award, and AMI
will persist.

Now the question is whether a geek lawyer can be found who would be
interested in contributing his or her services to a non-revenue open BIOS.

William Meyer

-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[Matthias Wächter]

On Sun, 8 Aug 1999, James Oakley wrote:
> http://www.patents.ibm.com/details?pn=US05732268__
>
> That does *not* look good.

Haven't seen the link, but I think we should consider the following: Many
companies pay their taxes in countries with low taxes, even if the work
was done in high-tax countries. Similarily, many companies try to develop
codes etc. in countries where copyright, trademark and patent law are on
their side.

Why don't we try to get the best of all the laws of the world? Disassemble
code in countries where it is OK, use algorithms in countries where one
cannot patent algorithms etc.

Sehr Wus,
- Matthias

Intel holding back information

[Stephan Mueller]

Am Mon, 09 Aug 1999, schriebst du:
> On Sun, 8 Aug 1999, James Oakley wrote:
> > http://www.patents.ibm.com/details?pn=US05732268__
> >
> > That does *not* look good.

This patent only convers communication over network, analog modems and isdn.
not a simple serial link, as i can read at ibmŽs page.

Ciao,
    Stephan
-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

The Thread: Re: Intel holding back information

[Matthew Sullivan]

Hi All,

May I make a suggestion, simply saying 'OpenBIOS is doomed' is not a good
attitude, the purpose of this list, and the project is to build an opensource
BIOS.  I think that there are a _number_ of open source projects with
proprietary code pieces used in them.  I wouldn't be at all suprised if some are
not the result of copyright theft.  However have these projects been closed?  I
don't recall any.....

Obviously, this doesn't mean we should steal others' code, but we should not let
it stop us writing ours.  We are part of an Open Project, made up of a hugely
diverse group of people, who would be held culpable in the event of
inadvertantly duplicating someone elses code?

<Start $0.02 worth>
When it comes to large companies and Open Projects, the current BIOS
manufacturers are unlikely to chase a  hundreds email addresses to attempt to
prove that a copyright was stolen - baring in mind that if they succeeded, they
wouldn't actually be able to get any pennies out of people, as most are skint
students, or idiots like me who spend all their money on pizza, coffee and
contacts.  No the other BIOS companies are more likely to implement some of our
code within their BIOS's.  Mainly because with all successful Open Source
projects, (Think about Apache, whilst this is the other end of the proverbial
stick..) the result is better, more efficient (and cheaper) code.
</Start $0.02 worth>

Matthew

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Accept that some days you are a pigeon, and some days you are the statue.

(was) Intel holding back information

[Ron Tsur]

I feel bad about apparently, opening the issue of patents and proprietary
information in my response to Niklas on this mailing list.  I don't see that
Niklas and this group of interested individuals should be discouraged from
pursuing the goal of producing free and open-source software.  The work should
proceed.  However, we should be careful about using any known-to-be-proprietary
information and signing non-disclosure agreements.

We should also realize that our work is done in the open.  The code can be
published in the public domain for a review period, when we can ask anyone,
including BIOS companies to provide their comments or shut up!  Technically,
this is already done now.  This may circumvent legal claims because they will
have to take active role in advance.  The burden of proof will also fall on
them, in a more significant way and make any legal move much more complicated
and expensive.  So far, I am not aware of any successful legal claim that was
taken against an open-source developer, who originated their code on their own,
without copying other people's code.

Many times you may NOT want to reinvent the wheel.  Sometimes you HAVE TO. 
Maybe we should look at developing BIOS specifications of our own, avoiding
legacy design problems that plague most existing software and reaching beyond
just compatibility with current commercial BIOS.  There is no reason to drive
the project through code hacking alone.  We may want to start by developing the
specifications first.

Note: We are now entering a new era where 64-bit machines, probably with more
than one CPU will be present on people's desktops and miniature computing
devices fit in people's shirt pockets.  Wireless networking is also emerging. 
Maybe this is an opportunity to create something new by transforming the old.

Ron.

(was) Intel holding back information: Maybe we should move on...

[James Oakley]

Ron Tsur wrote:
> I feel bad about apparently, opening the issue of patents and proprietary
> information in my response to Niklas on this mailing list.  I don't see that
> Niklas and this group of interested individuals should be discouraged from
> pursuing the goal of producing free and open-source software.  The work should
> proceed.  However, we should be careful about using any known-to-be-proprietary
> information and signing non-disclosure agreements.

[deletium]

Well said. The sudden turn to legal issues is, quite frankly, scaring
me. I, personally, can not afford to hire a lawyer with the necessary
qualifications. Maybe if we all chipped in we could... If we continue on
the path we have been going we will *need* a lawyer. Asking around is
simply not enough.

Therefore, I agree with Ron. We should build a specification from
scratch with reference to Linux. Then we get the Linux kernel to support
us. Then we can add features such as APM, improved user setup,
comprehensive hardware enumeration, and, of course, loading
DOS/Win/OS/2.

So... Let's start, shall we? Maybe we should contact specific kernel
maintainers about negotiating a spec. Ask them what bugs them about
current BIOS structure and collectively come up with solutions.
> From the maintainers list:

Primary Mailing List - [e-mail]

APM driver:	Stephen Rothwell - [e-mail]
    	Mailing List - [e-mail]

IDE driver:	Andre Hedrick - [e-mail]

IDE/ATAPI CDROM: Jens Axboe - [e-mail]

PCI Subsystem:	Martin Mares - [e-mail]

PNP:		Tom Lees - [e-mail]
    	List - [e-mail]

SCSI:		List - [e-mail]

Plus you-know-who.

I've probably left out people (specifically networking people).

Also we should contact the LILO maintainers.

I'd rather start doing this than muck around in waters of unknown depth.
I have *some* traditional BIOS information, including snippets of IBM
technical docs in a book on x86 architecture... I think we should wait
on DOS though. Maybe if we do this then we will know a lot more about it
all when we do tackle that situation.

Agree/disagree?

James Oakley
-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

(was) Intel holding back information

[Niklas Ekström]

On Mon, 9 Aug 1999, Ron Tsur wrote:
> I feel bad about apparently, opening the issue of patents and proprietary
> information in my response to Niklas on this mailing list.  I don't see that
> Niklas and this group of interested individuals should be discouraged from
> pursuing the goal of producing free and open-source software.

You shouldn't feel bad about that Ron, I think it's great that the issue
was brought up. Atleast I haven't been discouraged at all, rather gained
some important knowledge. Such as that a) I should _never_ sign any NDA
and that b) it might be alright to disassemble other BIOSs to gain 
knowledge on how to perform a certain task, as long as you never take any
code from the disassembled BIOS.

(However I don't have the intention to disassemble any BIOS; it just seems
safer to never have touched any commercial BIOS at all.)
> So far, I am not aware of any successful legal claim that was taken
> against an open-source developer, who originated their code on their
> own, without copying other people's code.

Neither have I, and I surely wouldn't want to the first one either! =)
> Many times you may NOT want to reinvent the wheel.  Sometimes you HAVE TO. 
> Maybe we should look at developing BIOS specifications of our own, avoiding
> legacy design problems that plague most existing software and reaching beyond
> just compatibility with current commercial BIOS.

I can assure you that GNUBIOS will be anything _but_ a IBM-PC compatible,
legacy BIOS. GNUBIOS will be built in three stages. In the first stage, 
it will be a 32-bit protectedmode BIOS able to boot Linux and other OSs
through the Multiboot standard (no real fancy features though). In the
second stage a BIOS Runtime Interface will be added and the possible to
boot from more devices etc. In the last stage, a full Open Firmware
implementation will be made (Open Boot standard, IEEE 1275). Most of the
BIOS will be rewritten to Forth in this stage (everything but the chipset
setup so that the FCode interpreter can run).

For those that doesn't immidiately see all the cool benefits of Open
Firmware; wait and see! It will rock immensely.

/ Niklas

-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message

Intel holding back information

[Tim Small]

Stephan Mueller wrote:
> Am Mon, 09 Aug 1999, schriebst du:
> > On Sun, 8 Aug 1999, James Oakley wrote:
> > > http://www.patents.ibm.com/details?pn=US05732268__
> > >
> > > That does *not* look good.
>
> This patent only convers communication over network, analog modems and isdn.
> not a simple serial link, as i can read at ibm´s page.

On the same thread, many companies routinely patent anything that any of their
employees can think of - the performance of patent clerks in the US is assessed
on the number of patents approved (not the number processed).

There are vast numbers of patents registered (especially software patents) which
embody ideas that are so obvious they are practically humorous - "infringements"
of such patents would never stand up to legal challenge, and indeed are never
likely to be challenged.  Patents of this type amount to little more than scare
tactics by company legal departments.

There has been much talk of disassembly on this mailing list recently, but is
there *really* need for disassembly for the completion of a free BIOS?  Examples
of usage of practically all BIOS functions should be present in code for which
source code is available, or for which textually descriptions are openly
available - shouldn't it be possible to use a black box approach whilst creating
a new BIOS?

I don't know the exact legal implications of using such an approach but I would
imagine they are fewer, and less serious than disassembly.

Tim.

p.s. Dude, perhaps if you learn assembler, i've heard that's great.. and visit
developer.intel.com ;-)

-
To unsubscribe: send mail to [e-mail]
with 'unsubscribe openbios' in the body of the message