Intellectual property violation?
Professor Ing-Jer Huang
Dear colleagues,
Do you know whether it will violate any intellectual property if we
take the (possibly behavioral) specification from the data sheet
of a commercial company's product (such
as Intel's 8051, MIPS R3000, etc.) and synthesize it?
Will the IP matter differ according to how we use the synthesized
result:
1. education
2. research
3. commercial product
Best regards,
--
Ing-Jer
----------------------------------------------------------------
Associate Professor Ing-Jer Huang
Institute of Computer and Information Engineering
National Sun Yat-Sen University
Kaohsiung, Taiwan 804
Republic of China
PHONE: 886-7-525-2000 ext. 4315
FAX: 886-7-525-4301
EMAIL: ijh...@cie.nsysu.edu.tw
----------------------------------------------------------------
Daniel L. Bates
In article <33B1E9...@cie.nsysu.edu.tw>,
ijh...@cie.nsysu.edu.tw says...
An interesting question. This happens with improved super sets. The
Zilog Z80 could execute Intel 8080 code. However the assembly syntax
was much much different. When NEC came out with the V20, V30 these
were pin compatible with the Intel 8088 and 8086 and could execute thier
code (albeit faster due to hardware effective address calculation and dual
rail microcode engine). These again were supersets of the Intel 8086
instruction set.
Personally I do not see how anyone can copyright an instruction set with
words like "add", or "move" or even register names like "AX"
--
*==================================================*
Dan Bates, Flower Mound, TX, USA
email: danbates "at" flash "dot" net
Pursuant to US Code, Title 47, Chapter 5, Subchapter
II, '227,any and all nonsolicited commercial E-mail
sent to this address is subject to a download and
archival fee in the amount of $500 US. E-mailing
denotes acceptance of these terms.
*==================================================*
timo...@cyberramp.net
Professor Ing-Jer Huang <ijh...@cie.nsysu.edu.tw> wrote:
>Dear colleagues,
>Do you know whether it will violate any intellectual property if we
>take the (possibly behavioral) specification from the data sheet
>of a commercial company's product (such
>as Intel's 8051, MIPS R3000, etc.) and synthesize it?
>Will the IP matter differ according to how we use the synthesized
>result:
> 1. education
> 2. research
> 3. commercial product
>Best regards,
>--
>Ing-Jer
I assume that some self inportant attorny would give you a hard time,
but I would consider it to be legitimate reverse engineering as long
as there was no "inside" information used; I.E. - Hiring an engimeer
that worked on the development of the chip.
Don North
> >Do you know whether it will violate any intellectual property if we
> >take the (possibly behavioral) specification from the data sheet
> >of a commercial company's product (such
> >as Intel's 8051, MIPS R3000, etc.) and synthesize it?
The simple (and complex) answer is 'it depends'. See below.
In article <33b4391c....@news.netcomuk.co.uk>, z...@dserve.com (Peter) wrote:
> Practically, it depends on whether the company finds out. LOTS of
> people have done ASICs containing synthesised CPUs, and if you don't
> tell then it is most unlikely anyone will find out.
This is really a bad approach, especially if you are a commerical
enterprise using the design in a product. If you are doing nothing
but 'research', it is probably not an issue. Even if someone patents
a concept, I can build a machine that uses it (as a lab experiment,
for example). However, I do NOT have the right to reproduce multiple
copies to be sold on the open market, whether for gain or not.
> I am not a lawyer, but I bet you that a lawyer (if you can find one
> who knows what a silicon chip is) would be happy to argue it for you.
> The data sheet info is public knowledge, and in the past a lot of
> reverse engineering of such data (if not actual CPUs) has been proved
> in the courts to be legal.
Using publically available knowledge (ie, a data sheet description) is
not sufficient to guarantee that you do not infringe a patented concept.
If you use 'trade-secret' information, you are on the hairy edge and you
better retain a good lawyer, or document very well how you legally
obtained the necessary information, whether by clean-room reverse
engineering or simultaneous re-invention.
For example, you could design a 486 processor using Intel data sheet
information, but to make a truly plug-compatible 486 you would need
to implement patented Intel concepts. Probably not a problem for a
research project (maybe), but definitely a problem if you are a for-
profit commercial concern. Talk to AMD, CYRIX, etc lawyers... :-)
> Some CPUs have been specially closely guarded by the original vendor.
> E.g. if you synthesised a Z80 in VHDL and offered it to all (free or
> not), Zilog would go nuts. AFAIK similarly for Intel for the 8051. In
> those cases you would presumably have to pay a license.
>
> But I heard that when NEC designed a Z80-lookalike (they called it the
> UPD780) in the early 1980s, they paid nothing to Zilog.
Licenses are for the right to use patented information, or for access
to trade-secret info that you don't want to spend the resources to
(re-)discover, but need to complete your design. NEC obviously required
neither for the Z80 clone, so owed nothing to Zilog. And if Zilog was
not far-sighted enough to obtain patents on key implementation issues,
they are basically SOL with respect to any license or royalty fees.
> There are firms who will sell you a HDL model of various CPUs,
> typically for some $1000s, and I suppose they pay a license to some of
> the original vendors.
Maybe or maybe not. Mostly you are paying for their time and energy to
implement and validate the HDL design. This is not an inexpensive
proposition in most cases that are not trivial to design. Whether they
pay (or more likey YOU) pay a license or royalty fee depends on the
above patent and/or trade secret issues.
---------------------------------------------------------------------------
Donald N. North KD6JTT don....@technologist.com
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
{{{{{{{{ Facts are facts, but any opinions expressed are my own }}}}}}}}}
---------------------------------------------------------------------------
Gareth Baron
In article <33B1E9...@cie.nsysu.edu.tw>, Professor Ing-Jer Huang
<ijh...@cie.nsysu.edu.tw> writes
>Dear colleagues,
>
>Do you know whether it will violate any intellectual property if we
>take the (possibly behavioral) specification from the data sheet
>of a commercial company's product (such
>as Intel's 8051, MIPS R3000, etc.) and synthesize it?
>
>result:
> 1. education
> 2. research
> 3. commercial product
>
3. It might be worth seeing how the Digital Equipment vs. Intel Law Suit
turns out. That might give us all an idea of what is legal and what is
not.
If there are any educational needs it is probably worth asking the Legal
departments in the various OEMs. They may grant you a licence to use
their microcode etc if you wish to produce a microprocessor similar to
theirs. You are best talking to the OEMs though.
Regards,
Gareth Baron
Brian J. Conant
prof,
typically anyone who would complain about this would be tempted to sue
for damages. if you aren't selling a product or offering something
which competes with someone's product then there is nothing to sue for.
i'm not a lawyer of course.
bjc
Professor Ing-Jer Huang wrote:
>
> Dear colleagues,
>
> Do you know whether it will violate any intellectual property if we
> take the (possibly behavioral) specification from the data sheet
> of a commercial company's product (such
> as Intel's 8051, MIPS R3000, etc.) and synthesize it?
>
> Will the IP matter differ according to how we use the synthesized
> result:
> 1. education
> 2. research
> 3. commercial product
>
> Best regards,
> --
> Ing-Jer
--
=====================
Check out my Harley
=====================
name
Professor Ing-Jer Huang (ijh...@cie.nsysu.edu.tw) wrote:
> Dear colleagues,
> Do you know whether it will violate any intellectual property if we
> take the (possibly behavioral) specification from the data sheet
> of a commercial company's product (such
> as Intel's 8051, MIPS R3000, etc.) and synthesize it?
> Will the IP matter differ according to how we use the synthesized
> result:
> 1. education
> 2. research
> 3. commercial product
In the U.S., I don't think there is any problem if the product is used for
education or research. Copyright laws say you can copy articles or parts of
books for educational purposes. Research is usually internal to a company, so
it's sort of educational to the employees, and no profit is being made on the
item.
As for commercial product, I suggest seeking a lawyer and asking him/her some
questions about that application.
GREG
Thomas A. Coonan
I emailed MICROCHIP directly as to whether my synthesized PIC core
would upset them in anyway. I got no answer. If your not a threat,
why would they worry about it, I suppose..
tom coonan
http://www.mindspring.com/~tcoonan
J.T. Conklin
Professor Ing-Jer Huang <ijh...@cie.nsysu.edu.tw> writes:
> Do you know whether it will violate any intellectual property if we
> take the (possibly behavioral) specification from the data sheet of
> a commercial company's product (such as Intel's 8051, MIPS R3000,
> etc.) and synthesize it?
I vaguely remember hearing that either IBM or Moto considered PowerPC
simulators which were written with information from the architecture
manuals to be derrived works. Andrew Cagney <cag...@cygnus.com> the
author of the ppc simulator used in GDB probably knows more.
--jtc
Mike McCarty
In article <don.north-260...@mac.slowdeath.com>,
Don North <don....@technologist.com> wrote:
)
)> >Do you know whether it will violate any intellectual property if we
)> >take the (possibly behavioral) specification from the data sheet
)> >of a commercial company's product (such
)> >as Intel's 8051, MIPS R3000, etc.) and synthesize it?
)
)The simple (and complex) answer is 'it depends'. See below.
)
)In article <33b4391c....@news.netcomuk.co.uk>, z...@dserve.com (Peter) wrote:
)
)> Practically, it depends on whether the company finds out. LOTS of
)> people have done ASICs containing synthesised CPUs, and if you don't
)> tell then it is most unlikely anyone will find out.
)
)This is really a bad approach, especially if you are a commerical
)enterprise using the design in a product. If you are doing nothing
)but 'research', it is probably not an issue. Even if someone patents
)a concept, I can build a machine that uses it (as a lab experiment,
)for example). However, I do NOT have the right to reproduce multiple
)copies to be sold on the open market, whether for gain or not.
)
)> I am not a lawyer, but I bet you that a lawyer (if you can find one
)> who knows what a silicon chip is) would be happy to argue it for you.
)> The data sheet info is public knowledge, and in the past a lot of
)> reverse engineering of such data (if not actual CPUs) has been proved
)> in the courts to be legal.
)
)Using publically available knowledge (ie, a data sheet description) is
)not sufficient to guarantee that you do not infringe a patented concept.
)If you use 'trade-secret' information, you are on the hairy edge and you
)better retain a good lawyer, or document very well how you legally
)obtained the necessary information, whether by clean-room reverse
)engineering or simultaneous re-invention.
[snip]
In an amusing court case several years ago, Motorola sued a Japanese
company (Hitachi, I think) for cloning off one of its processors
(68020, possibly the 68030). Hitachi had licensed rights to manufacture
them, and had obtained the masks. But they claimed that they had also
hired "virgin" engineers (ones who had never seen the masks) and gave
them requirement specs on the proc. They produced a working clone, but
were "clean". Then they set off looking for patent violations on
Motorola's part, and found that THEY owned the rights to VFETs, which
are used in the internal RAM used for cache in those processors.
Motorola countered that the engineers who wrote the requirements -were-
familiar with the masks, and that Hitachi was therefore illegitimately
benefiting from having the trade secrets. The court ruled in both
side's favors, indicating that Hitachi (?) must discontinue
manufacturing the clone, and Motorola had to quit manufacturing 68020s
(or 30s, I forget). Both sides immediately licensed the other to do
what it was already doing.
It was a little scary at the time, because we were manufacturing boards
using the 68020 (o 68030).
In another case, NEC was sued in exactly the same way over the V20/V30
chips by Intel. NEC had the masks for the 8088/8086. In particular,
Intel claimed that they had stolen the multiply (if I recall correctly)
microcode. NEC claimed that they were architecturally constrained to
generate the same microcode, because there was really only one
efficient way to write the code necessary to exactly reproduce the
behavior of the 8088/8086. The judge decided that he did not have
enough training to decide the case, quit the bench for a semester and
went back to graduate school and took a course in computer
architecture, and learned how to write a multiply instruction in
microcode. After he got back to the bench, he listened to the
arguments, and decided that NEC actually -was- architecturally
constrained, and ruled in NECs favor. And so one can still purchase a
V20/V30 chip.
So one can indeed do reverse engineering. But one cannot use trade
secret material EVEN INCIDENTALLY to the development of the reverse
engineered product.
Mike
--
----
char *p="char *p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
I don't speak for DSC. <- They make me say that.
Ben Twijnstra
On Thu, 26 Jun 1997 09:42:28 +0100, Gareth Baron
<gar...@trsys.demon.co.uk> wrote:
>In article <33B1E9...@cie.nsysu.edu.tw>, Professor Ing-Jer Huang
><ijh...@cie.nsysu.edu.tw> writes
>>Dear colleagues,
>>
>>Do you know whether it will violate any intellectual property if we
>>take the (possibly behavioral) specification from the data sheet
>>of a commercial company's product (such
>>as Intel's 8051, MIPS R3000, etc.) and synthesize it?
>>
>>Will the IP matter differ according to how we use the synthesized
>>result:
>> 1. education
>> 2. research
>> 3. commercial product
>>
For either case, if you use a provable 'clean room' approach there
shouldn't be any problem. The specs are published, and therefore
public knowledge.
We had a trainee and a graduate student write a synthesizable model
for the 8085 microcontroller, just based on the specs. We have no
commercial or intellectual contacts with Intel regarding the 8085. So
far, everyone in the educational corner has applauded the students'
efforts. Wehaven't had any comments from Intel yet (haven't asked them
either, but anyhow). Given this really "clean room" approach, this
should mean that for educational purposes everything's OK. However, if
you're able to design a slab of silicon that will do the same thing
but will underprice Intel you have to be _very_careful indeed as to
where you got your information from.
The IBM vs. Phoenix lawsuits will be enlightening if you're after the
principles of "clean room" (re)design.
Grtz,
Ben
-- Remove the nospam to reply by email.
Bernd Paysan
J.T. Conklin wrote:
>
> Professor Ing-Jer Huang <ijh...@cie.nsysu.edu.tw> writes:
> > Do you know whether it will violate any intellectual property if we
> > take the (possibly behavioral) specification from the data sheet of
> > a commercial company's product (such as Intel's 8051, MIPS R3000,
> > etc.) and synthesize it?
>
> simulators which were written with information from the architecture
> manuals to be derrived works. Andrew Cagney <cag...@cygnus.com> the
> author of the ppc simulator used in GDB probably knows more.
Sounds strange. All the second source manuals of processor manuals (or
other manuals) are much closer to be derived work than a simulator
written from a user manual. The copyright law has a point where taking
other (copyrighted) pieces of work and creating something different
doesn't violate the original copyright. Thus when Andy Warhol took a
Coca Cola can and signed it, the created "ready made" was the
intellectual property of Andy Warhol (at least the one who bought it for
lots of $ instead of taking an "unsigned" Coca Cola can out of the next
dust bin honored it, and Coca Cola Inc. would have been a fool to sue
him ;-).
Certainly it depends on what's written in the manual. If AIM wrote a
manual that turns itself into a processor simulator if you tangle it and
run a C compiler on the output (as it's possible with the TeXbook), you
should not read the manual to write a clean-room simulator. But I doubt
that.
--
Bernd Paysan
"Late answers are wrong answers!"
http://www.informatik.tu-muenchen.de/~paysan/
s...@bob.eecs.berkeley.edu
Historians will recall a very early case of this
sort being IBM v. Amdahl -- Amdahl having reverse-engineered
an IBM mainframe (the 360 I think), and the court ruling
this was legitimate. They could not of course clone
the operating system without violating software copyrights;
this is the root of the current industry dichotomy
wherein computers and CPU's are commonly cloned, but
software makers have a lock on their products.
For many years most companies felt it was fair game to clone
a chip; indeed, this was the standard approach to second-sourcing,
and led to healthy competition on things like 74-series parts,
DRAMS, opamps, voltage regulators, etc. It's only when the stakes
got really high with high-priced, high-volume CPU's that
the lawyers took another look at it and figured out new ways
to take legal action against clones.
Steve
Allison
z...@dserve.com (Peter) wrote:
>But I heard that when NEC designed a Z80-lookalike (they called it the
>UPD780) in the early 1980s, they paid nothing to Zilog.
True, the design was clean slate to emulate the z80 perfectly.
Allison
Alli...@world.std.com
Real address is: Allisonp @ world DOT std DOT com
++++BULK Email severely not wanted+++
Olav Woelfelschneider
Peter <z...@dserve.com> wrote in comp.arch.embedded:
P> E.g. if you synthesised a Z80 in VHDL and offered it to all (free or
P> not), Zilog would go nuts. AFAIK similarly for Intel for the 8051. In
P> those cases you would presumably have to pay a license.
Some time ago, Siemens manufactured the 8031/8051 etc. using a costly
license from intel. Today they sell the C501 and friends, which are,
according to a Siemens salesdroid, completely redesigned cpus to avoud
license payments.
Wonder if this is right, if intel went nuts and how it turned out. At least,
the C501 etc. are still available.
--
Olav "Mac" Wölfelschneider wo...@rbg.informatik.th-darmstadt.de
PGP fingerprint = 065F66B32AAD7D2DB719673C95A79DAF
But let your communication be Yea, yea; nay, nay: -- Matthew 5:37
for whatsoever is more than these cometh of evil. on computers
Dick Wilmot
s...@bob.eecs.berkeley.edu writes:
>Historians will recall a very early case of this
>sort being IBM v. Amdahl -- Amdahl having reverse-engineered
>an IBM mainframe (the 360 I think), and the court ruling
>this was legitimate. They could not of course clone
>the operating system without violating software copyrights;
>this is the root of the current industry dichotomy
>wherein computers and CPU's are commonly cloned, but
>software makers have a lock on their products.
I atarted at AMdahl in '79 and understood that the design was to the IBM 370
Principles of Operation (the POO) which was certainly the bible by my time.
We had many ex-IBMers and they were even more stringent POO lawyers than the
rest of us. The POO as architecture spec seems to have been (probably still is)
the explicit agreement between the hardware folk and the software people
(OS, compilers, database, telecomms).
IBM had bundled MVS operating system software into sales to the US Gummint which
put it into the public domain until they later (early '80s?) unbundled it and
made it a separate license. Amdahl shipped copies of the early MVS (and other
operating systems such as VS1, SVS, OS MVT, VM, etc.) to its customers and IBM
licensed later MVS/SP, MVS/EXA, MVS/ESA, VM/SP, ... to Amdahl, NAS, Magnuson
(?), Hitachi, (and now Commercial Data Systems?) customers directly with
some occasional help from Amdahl folks.
Amdahl & IBM wre cross-licensed when I was there and I think that had been
ongoing for some while so you wouldn't expect patent suits.
The openness of the S/360, S/370, S/390 seems to have been good for its
commercial viability and stamina. I think I'll offer to teach my grandson
S/390 JCL and so called High Level Assembler so that he can look forward to
a good retirement after years of fixing the remaining Y2K bugs - no one else
is being trained for this duty ;)
--
Dick Wilmot
Diablo Data Design
Non-spammers (e.g. humans) should dw the dw in my address to reply to my usenet posts
Dan Babcock
On 28 Jun 1997 18:35:09 GMT, s...@bob.eecs.berkeley.edu wrote:
>this was legitimate. They could not of course clone
>the operating system without violating software copyrights;
>this is the root of the current industry dichotomy
>wherein computers and CPU's are commonly cloned, but
>software makers have a lock on their products.
Well, no. Most of the "look and feel" lawsuits were resolved in favor
of the defendents. Unlike the hardware sphere, patents were (until
recently) not an issue.
Since CPUs compete primarily on performance, not features, it is only
natural for them to be clones (the Pentium Pro is a "clone" of the
386, etc). Software competes primarily on features, so clones are the
exception. This has little to do with the legal situation.
>For many years most companies felt it was fair game to clone
>a chip; indeed, this was the standard approach to second-sourcing,
>and led to healthy competition on things like 74-series parts,
>DRAMS, opamps, voltage regulators, etc. It's only when the stakes
>got really high with high-priced, high-volume CPU's that
>the lawyers took another look at it and figured out new ways
>to take legal action against clones.
The laywers have always been involved, it's just a matter of degree.
If I remember correctly, Motorola sued MOS over the 6500, forcing a
change in the bus design (it was to be pin-compatible, but not
software-compatible, with the 6800).
Dan
Toon Moene
all...@see.sig (Allison) wrote:
>z...@dserve.com (Peter) wrote:
>>But I heard that when NEC designed a Z80-lookalike (they called it the
>>UPD780) in the early 1980s, they paid nothing to Zilog.
>True, the design was clean slate to emulate the z80 perfectly.
Interesting ... Did they also preserve the 100+ undocumented instructions ?
--
Toon Moene (mailto:to...@moene.indiv.nluug.nl)
Saturnushof 14, 3738 XG Maartensdijk, The Netherlands
Phone: +31 346 214290; Fax: +31 346 214286
g77 Support: mailto:for...@gnu.ai.mit.edu; NWP: http://www.knmi.nl/hirlam
timo...@cyberramp.net
>Some CPUs have been specially closely guarded by the original vendor.
>E.g. if you synthesised a Z80 in VHDL and offered it to all (free or
>not), Zilog would go nuts. AFAIK similarly for Intel for the 8051. In
Allison
Toon Moene <to...@moene.indiv.nluug.nl> wrote:
>all...@see.sig (Allison) wrote:
>>z...@dserve.com (Peter) wrote:
>>>But I heard that when NEC designed a Z80-lookalike (they called it the
>>>UPD780) in the early 1980s, they paid nothing to Zilog.
>>True, the design was clean slate to emulate the z80 perfectly.
>Interesting ... Did they also preserve the 100+ undocumented instructions ?
I said perfectly! Yes all of them even the high order address buss
twitch the z80 had.
The NEC 78xx series was z80 like internally but different in that it
would not run 8080/z80 code, different opcodes but similar fuctions.
Allison
z...@dserve.com (Peter) wrote:
>>Interesting ... Did they also preserve the 100+ undocumented instructions ?
>No idea, but there were differences. Early UPD780s would power-up with
>interrupts ENABLED.
I supported the nec part at one time. and the answer is, no, it didn't
behave any different from the zilog/mostek parts.
AS to the 100+ there were a lot but, I never actually counted them.
Hans Tiggeler
In article <33b4391c....@news.netcomuk.co.uk>, z...@dserve.com says...
Can they stop it? To learn VHDL I wrote a simplified model of a PIC micro
controller. Like me there must be thousands of engineers and student doing
exactly the same. I won't be long before these designs are freely available on
the web. To add to this, synthesizers are getting more and more powerful for
less money. If they want to go nuts they have to accept to go nuts from this
day onwards :-)
Hans.
Stephen Maudsley
Ben Twijnstra wrote:
> On Thu, 26 Jun 1997 09:42:28 +0100, Gareth Baron
> <gar...@trsys.demon.co.uk> wrote:
>
> >In article <33B1E9...@cie.nsysu.edu.tw>, Professor Ing-Jer Huang
> ><ijh...@cie.nsysu.edu.tw> writes
> >>Dear colleagues,
> >>
> >>Do you know whether it will violate any intellectual property if we
> >>take the (possibly behavioral) specification from the data sheet
> >>of a commercial company's product (such
> >>as Intel's 8051, MIPS R3000, etc.) and synthesize it?
>
> For either case, if you use a provable 'clean room' approach there
> shouldn't be any problem. The specs are published, and therefore
> public knowledge.
>
The fact that something is in the public domain does not mean that you
are free to copy it at will - you need to be careful that you do not
inadvertently breach copyright on the specification by republishing
parts of the origninal.
Also I recollect that patents have been granted on some instruction sets
(although I can't remember the reason why).
--
Stephen Maudsley
esgem limited
Tel: +44-1453-845015
Stephen....@btinternet.com
Eric Ryherd
Peter wrote:
>
> >Interesting ... Did they also preserve the 100+ undocumented instructions ?
VAutomation sells a synthesizable Z80. We didn't preserve the
undocumented
opcodes as this would have been nearly impossible without trade secret
information...
--
Eric Ryherd er...@vautomation.com
VAutomation Inc. Synthesizable VHDL and Verilog Cores
20 Trafalgar Sq. #443 http://www.vautomation.com
Nashua NH 03063 (603) 882-2282 FAX:882-1587
John Eaton
Dan Babcock (da...@nospam.tiac.net) wrote:
: The laywers have always been involved, it's just a matter of degree.
: If I remember correctly, Motorola sued MOS over the 6500, forcing a
: change in the bus design (it was to be pin-compatible, but not
: software-compatible, with the 6800).
That was the MCS6501 that was a drop in for the 6800 but with a different
set of opcodes for 1/4 the price.
Patents are nice but they are limited to a 17 year life. Anything from
the 70's is now fair game.
John Eaton
Don Yuniskis
In article <33ba8f6...@news.netcomuk.co.uk>, Peter <z...@dserve.com> wrote:
>
>>Patents are nice but they are limited to a 17 year life. Anything from
>>the 70's is now fair game.
Hmmm... I thought patent protection was extended recently by some
of these gazillions of international treaties... (?)
>So why is there not a public domain VHDL Z80? Or any other half-useful
>CPU??
I think the 6502 core is widely used in ASICs... but does that really
count as a CPU? :>
>Presumably *copyright* is the bit stopping that. That lasts for 50
>years (?) after the death of the designer, and the designer of e.g. a
>Z80 would have assigned the copyright to his then employer, Zilog.
Again showing my ignorance but wouldn't copyright only extend to a
particular *implementation* of the design? I.e. you can copyright
the masks, a PCB layout, etc. You can copyright the instruction set
but I only think that covers the actual choice of mnemonics (i.e. you
could come up with a functionally identical instruction set but rename
the "LD" instructions to be "LOAD" and the "SET" instructions to be
"TURN_ON", etc. without violating??).
I think this is how the "knock-off" vendors have been able to get
around the issue -- i.e. AMD can clone a 486 as long as they do it
using cleanroom techniques and don't copy microcode, mask, etc.
There are firms here (US) who exist *solely* to mimic products made by
other "mainstream" (i.e. Big Blue) firms and I'm sure if they could
protect their design just by stamping a copyright on it, they would!
Anyone with a better grasp of the actual legalese care to chime in?
--don
Don Yuniskis
In article <5p5bmd$4tp$1...@news.utrecht.NL.net>,
Toon Moene <to...@moene.indiv.nluug.nl> wrote:
>all...@see.sig (Allison) wrote:
>
>>z...@dserve.com (Peter) wrote:
>
>>>But I heard that when NEC designed a Z80-lookalike (they called it the
>>>UPD780) in the early 1980s, they paid nothing to Zilog.
>
>>True, the design was clean slate to emulate the z80 perfectly.
>
I don't believe there were "100+" of these. ;-) Most simply broke IX or
IY into the hi/lo bytes analagous to H and L in HL. (the ED, etc.
opcode prefix actually just toggled whether the NORMAL z80 instruction that
followed operated on HL or IX or IY, etc.
Some of the 8085 clones did *not* emulate the undocumented instructions
therein.
--don
Cardiac Pacemakers, Inc.
It's not because it can't be done for legal reasons, as the message from
VAutomation indicates. It's just that it's a heck of a lotta work. Not
just to design, but to verify operation.
--
Scott Reedstrom
scott.r...@nospam.guidant.com
(remove 'nospam' to reply)
My opions are mine, not my employer's
Peter <z...@dserve.com> wrote in article
<33ba8f6...@news.netcomuk.co.uk>...
>
> >Patents are nice but they are limited to a 17 year life. Anything from
> >the 70's is now fair game.
>
> So why is there not a public domain VHDL Z80? Or any other half-useful
> CPU??
>
> Presumably *copyright* is the bit stopping that. That lasts for 50
> years (?) after the death of the designer, and the designer of e.g. a
> Z80 would have assigned the copyright to his then employer, Zilog.
>
>
> Peter.
>
> Return address is invalid to help stop junk mail.
> E-mail replies to z...@digiserve.com.
>
Tim Shoppa
In article <5pdq2i$8...@news.vcd.hp.com>, John Eaton <jo...@vcd.hp.com> wrote:
>Dan Babcock (da...@nospam.tiac.net) wrote:
>
>: The laywers have always been involved, it's just a matter of degree.
>: If I remember correctly, Motorola sued MOS over the 6500, forcing a
>: change in the bus design (it was to be pin-compatible, but not
>: software-compatible, with the 6800).
>
>That was the MCS6501 that was a drop in for the 6800 but with a different
>set of opcodes for 1/4 the price.
>
>the 70's is now fair game.
It should also be remembered that patents only prevent your competitors
from selling your patented products. They do not prevent your
competitors (or academic or commercial researchers) from building
the patented items for research and testing purposes.
Tim.
Glenn C. Everhart
There may not be a VHDL Z80, but there certainly are multiple Z80
emulators out there. Given the great speed increases of the machines
they're running on, these are able to give very creditable performance
compared with the original, despite the inefficiencies of software
emulation.
The question could be asked, what does one want a free Z80 for?
There are various possible answers, some of which these emulators address.
Tom Thornhill
Peter <z...@dserve.com> wrote in article <33bbfce9...@news.netcomuk.co.uk>...
>
> >The question could be asked, what does one want a free Z80 for?
>
>
>
> Peter.
>
> Return address is invalid to help stop junk mail.
> E-mail replies to z...@digiserve.com.
>
Just a thought -
Most of the intellectual property violation problems
seem to come from microcode, so does this mean that Risc
chips are easier ( legally ) to clone ?
Are Risc vendors keener on licensing because it's harder
for them to prevent reverse engineering using patents?
I can see that the implementation detail of something like
the Alpha can be patented and that this would make it harder
to develop competitive clones, but I'm not sure if you could
prevent someone else implementing the architecture as easily
as if it were Cisc.
Tom.
--
Please email me a copy of anything you post to a news group
.... unless it's unadulterated shite !
Mailto: ttho...@best.ms.philips.com-nospam ( remove the -nospam before replying )
Talkto: Tom Thornhill, Philips Medical Systems Nederland BV
Tel +31 4027 64080 ( Work ), +31 499 371230 ( Hotel - ask for room 11 )
************************ +44 3709 46045 GSM ****************************
**** IMPORTANT: please use the above numbers rather than the GSM *******
**** one if possible as I pay for incoming calls from outside the ****
**** Netherlands. *****************************************************
Mike McCarty
In article <5p7pmp$t...@info-server.surrey.ac.uk>,
Hans Tiggeler <ees...@ee.surrey.ac.uk> wrote:
)In article <33b4391c....@news.netcomuk.co.uk>, z...@dserve.com says...
)
)>Some CPUs have been specially closely guarded by the original vendor.
)>E.g. if you synthesised a Z80 in VHDL and offered it to all (free or
)>not), Zilog would go nuts. AFAIK similarly for Intel for the 8051. In
)>those cases you would presumably have to pay a license.
)
)Can they stop it? To learn VHDL I wrote a simplified model of a PIC micro
)controller. Like me there must be thousands of engineers and student doing
)exactly the same. I won't be long before these designs are freely available on
)the web. To add to this, synthesizers are getting more and more powerful for
)less money. If they want to go nuts they have to accept to go nuts from this
)day onwards :-)
)
)Hans.
I know of cases where courts have ruled that reverse engineering exact
functionality without use of proprietary information is -not- a crime
-nor- is it a tort. Have a look at Motorola vs. Hitachi.