|
|
| The Patent Conversation Continues -- Sun's Jonathan Schwartz |
 |
|
Wednesday, October 20 2004 @ 02:26 AM EDT
|
The most remarkable thing that has come out of the Kodak v. Sun patent lawsuit is that everyone now is talking about software patents. That is mighty fine, especially because everyone seems to be agreeing that the US Patent Office has been granting patents that should never have been granted and that something needs to be done to fix a broken system. I call that progress. In the knock-me-over-with-a-feather department, Monday that chorus was joined by Sun's Jonathan Schwartz, who writes in his blog about the Kodak settlement and says that "[s]ome systemic changes need to be made" to the current patent system. I'm guessing that in his perfect world, those changes would be made retroactive to October 1st, $92 million ago.
Evidently, he reads Groklaw, because he mentions it, and he says he hopes I take no offense at a point he makes. I don't. Welcome to Groklaw, Jon. I'm very glad you expressed your awareness of some of the problems you see in the current patent system, but if I may I'd like to reply, with a question for you. First, what Jon wrote, in part: "Are software patents without controversy?
"No, not at all. Some systemic changes need to be made. But the assertion that patent coverage should be eliminated from the software industry makes no sense.
"There's this great irony: spend time with Sun's microprocessor team, and guess what - they're all writing software. Spend time with our data center switching team - they're all writing software, too. Viewed simplistically, computing hardware is software burned into and onto physical things. And over time, more and more routine software elements end up in hardware, for acceleration or optimization. SSL accelerators, JVM on a chip, you name it. So, where do you draw the line on patents? Firmware? FPGA's? Silicon? Systems?
"Yes, we should reform the system to stop granting spurious patents. And yes, we should disincent spurious litigation. That's what we're advocating in Europe, before the legislature there settles on the worst rather than the best of US practice. Eliminating patent coverage to benefit copyright-only companies is imbalanced - but safeguards are needed if the system is to protect innovation rather than allow predators to tax standards. Safeguards, I agree, that aren't in place today." The elephant in the room in all software patent discussions is GNU/Linux. What about Linux? What about GPL'd software? How can the US patent system be made to fit a world that wants that software? And it does want it. Proponents of software patents need to answer that question. In Europe, they must answer it this exact minute. Because that is the real question. If you have software patents, what happens to GNU/Linux software? I don't see how you can have both. Sooner or later, the Patent Cold War guarantees only one software company in the world, a proprietary one, or a small club of proprietary software giants, will survive. Even Cold Wars are eventually won by somebody, you know, and the patent system is set up so that only the biggest and richest can play, and whoever has the most patents wins in the last battle. Software is math, with only so many ways to do certain fundamental tasks. Once somebody has a patent on every fundamental task, then no one new can write software. Why would the world want that? It's natural that those with stockpiles of patents would love a system where they get value from those patents. We expect you to like that, but the rest of the world has an interest in this question too. What about our interest in having software choice?
It's all very well to say that IP is what makes the world go 'round ("I believe in intellectual property. In my view, it's the foundation of world economies, and certainly the foundation upon which Sun Microsystems was built. Copyright, trademark, patent - I believe in them all"), and that paying off Kodak was your way of backing up your customers, but what about Linux developers? No, really. You use GPL software in your offerings, so you must think it is good software. How do we protect GPL software in a patent-mad world? Sun may have sufficient funds to throw at patent infringement lawsuits, but what about coders who don't have $92 million in their pocket? If the patent system is set up so only those with millions or billions of ducats to pay off the patent highwaymen can continue on the software road, how does something like Linux come to be? Do we want a system whereby only the rich can write software? If we'd had such a system from the beginning, instead of only recently, Linux could never have been written. If large companies enjoy playing Patent Cold War with their stockpiles of patents, whether they use them for offensive purposes or only in self-defense, then I think no one in the FOSS world minds. Fire away at each other all you like, although it looks ridiculously wasteful to the rest of us. But knock yourselves out, if this is your idea of business competition. But find a way, please, if you insist on software patents, to leave FOSS out of it. Let's be honest. Patents are a monopoly grant with a purpose, and the purpose isn't so you can make out like bandits. The whole idea is to benefit the rest of us too while allowing you to reap a fair reward for your labor for a limited period. Where, though, is the benefit to the public in the current system? I see the benefit to large companies, but where is the benefit to small and medium companies? The best they can hope for, under the current system, is to be bought by a bigger fish, and that isn't everyone's dream, you know. What about all the programmers in the world who simply love to write Linux? They do it because they enjoy the creative challenge, and money isn't the object of it all. That's the real reason SCO can't seem to find any infringing code in Linux, by the way. There is no fun in copying code, to those whose motivation is the creative challenge of writing beautiful code. Where is the benefit to the public, if GPL'd code is not allowed to exist, merely because its authors don't have millions or billions of dollars to play Patent Cold War? I can think of a company here or there who might very well like it if Linux had never been born, but should the public be denied the benefit of the incredibly creative work of thousands and thousands of programmers around the world who have produced software a lot of folks love to use just to benefit one company, or even two or three? If the answer to that is no, and it is, then some accommodation needs to be made so that patents don't kill off software that isn't written with money as the goal, so that creativity is allowed to flower wherever it happens to crop up. Do you want a world where software innovation is progressively stifled and GNU/Linux software is snuffed out by the big players? If you do, say so openly. If you don't, and you say you don't like the current system, what do you propose to safe harbor GPL'd code? Of course, Linux will continue somewhere. There will still be places where US-style patent laws don't exist, and FOSS innovation will continue at its dizzying pace, and eventually US software, crippled by patent restrictions, will become Brand X, because it will be second-best. Creativity tends to win, you know. Fighting against human creativity is like trying to eradicate dandelions. No matter what you try, they just grow back, through cracks in cement if they have to. So, Jon, how do you propose to have patents on software but protection for GPL'd software? Or is that not a common goal?
|
|
|
|
| Authored by: moosie on Wednesday, October 20 2004 @ 03:13 AM EDT |
Here is your bridge.
- Moosie.
[ Reply to This | # ]
|
| |
| Authored by: robvarga on Wednesday, October 20 2004 @ 03:20 AM EDT |
With all respect? Why is GPL'd software reserved this special place in this
argument? What about other free licences? What about proprietary code of small
companies?
Any system which would have special cases for free software would be equally
wrong, because it draws a line by providing exemption from the rules, but that
strict line cannot in fact be drawn.
A proprietary company might want to make money on its own software which it has
written on its own. It deserves the same protection.
Therefore no system which takes the licences into account is a good system, so
something different should be devised.
The best approach would be to drop patentability on software, because
independent parallel innovation should not be penalized at all, and patent
system does penalize parallel innovation.
[ Reply to This | # ]
|
| |
| Authored by: robvarga on Wednesday, October 20 2004 @ 03:21 AM EDT |
| "With all respect?" wanted to be "With all respect,"... [ Reply to This | # ]
|
| |
| Authored by: attila_the_pun on Wednesday, October 20 2004 @ 03:31 AM EDT |
Starting with
s/duckets/ducats/[ Reply to This | # ]
|
| |
| Authored by: inode_buddha on Wednesday, October 20 2004 @ 03:38 AM EDT |
I've long believed in patents on physical implementations. If I can drop it on
my foot, it is patentable. I feel that patents in silicon are fair and just.
Patents on fundamental building blocks used in larger works should not be
patentable, just IMHO. That would be like patenting the alphabet in order to get
royalties from Shakespeare's works. The works as a whole may be quite innovative
and protectable, but there are only so many building blocks to logic. There are
only so many ways to say "AND, OR, NAND, XOR..." etc.
---
"When we speak of free software, we are referring to freedom, not price." --
Richard M. Stallman[ Reply to This | # ]
|
| |
| Authored by: muswell100 on Wednesday, October 20 2004 @ 03:41 AM EDT |
"Some systemic changes need to be made. But the assertion that patent
coverage should be eliminated from the software industry makes no sense."
I'm sorry, but Mr Schwartz is wrong about this one. You can't have only 'half a
software patent system'. Patents on software/maths/ideas will always carry the
same sting in the tail, regardless of how much you try to water it down or tidy
the whole mess up. The software patent debacle is ultimately going to prove an
own-goal for the big companies that even only half-support it, as Mr Schwartz
seems to.
Copyrights exist in Europe to protect the work of programmers, artists and
writers across the board. The whole reason software patents came into the mix at
all seems to have been due to the status quo realising that their own buggy,
expensive copyrighted code couldn't compete with the superior combined efforts
of FOSS developers. The obvious (to them) answer is to stop protecting the code
and concentrate on the concepts instead. Put a price tag on ideas, and it
doesnt' matter if your source code is visible for all to see or not.
And, as the software industry giants are starting to realise, the picture is not
going to prove all that rosy.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 03:43 AM EDT |
The GPL is a form of patent. It says: here's my work, you can see it, but you
can only use it under certain conditions.
The patent says: here's my work, you can see it, but you can only use it under
cetain conditions.
The conditions can be money, or other as in the case of the GPL. (But GPL can
also be set aside for money sometimes).
The problems with patents need to be dealt with. But, they have an important
place. Those who've done a lot of R&D understand very well.
The above also shows the obvious answer to defeat MS and a patent attack against
open source:-
Whenever a new tehnology comes out. Write a series of small, relevant pieces of
code for it. And GPL them.
Do this early and often.
If MS comes after a small developer for patent infringement, then you can
counter-attack because there may well be GPL infringement in their source code.
That entitles you to discovery.
If you can do discovery on MS source code, you will get lawyers queued round the
block to work pro bono for you. Who knows what else they'll find?
And MS, knowing this, will back off fast.
Realise (a) GPL is form of patent - you may need to think about this to get your
head around it, and (b) it can be used as a defence against any large software
company trying to unfairly attack you.
Submarine patents are a separate genuine problem, but fortunately affect larger
companies, so they will fight that well.
[ Reply to This | # ]
|
- The GPL *is* a patent - Authored by: PJ on Wednesday, October 20 2004 @ 04:01 AM EDT
- The GPL is NOT a form of patent! - Authored by: Anonymous on Wednesday, October 20 2004 @ 04:34 AM EDT
- The GPL *is* ***NOT*** a patent - Authored by: robvarga on Wednesday, October 20 2004 @ 05:05 AM EDT
- The GPL *is* a patent - Authored by: robvarga on Wednesday, October 20 2004 @ 05:10 AM EDT
- Bzzzt. Sorry. Play again. - Authored by: Anonymous on Wednesday, October 20 2004 @ 05:53 AM EDT
- The GPL *is* a patent - Authored by: eggplant37 on Wednesday, October 20 2004 @ 06:51 AM EDT
- The GPL *is* a patent - Authored by: Anonymous on Wednesday, October 20 2004 @ 10:16 AM EDT
- The GPL *is* a patent - Authored by: Anonymous on Wednesday, October 20 2004 @ 10:55 AM EDT
- The problems I see with the parent - Authored by: penfold on Wednesday, October 20 2004 @ 04:24 PM EDT
- The GPL *is* a patent - Authored by: blacklight on Wednesday, October 20 2004 @ 05:35 PM EDT
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 04:59 AM EDT |
If you got an idea, you can call it a dream and you decide to go and pantent
this idea. That is all you have done.
With software, it is the effort and the industry that leads to the real dream
being possible. And even then by version 4 you might have the bugs cleaned up
enough so that it works right.
That effort can be protected... BY COPYRIGHT.
AND you can make money with this protection.
When you mix GPL software and software with other licenses you have the same
thing in the market as a Chef in the kitchen who goes out and buys the best
fruit, the best meat, the best fish, the best of everything as an part of the
meal that is being prepared. The better the chef, the chances are the better
the prep work prior to the meal (and the better the quality of the art). The
Chef may even keep a garden too... and grow what is wanted for when it is
needed.
We as the ones who go to this Chef's place of business then have a choice of
eating there, or going somewhere else to buy what we thing will be a good meal
that is worth the money we pay for it. I have eaten a sandwich for a few
dollars and I have eaten a meal where going in we knew we were going to pay at
least $150 per person for the quality - and it was worth every cent. Thank
goodness that patents don't exist for those in the food business who create the
foods that we all enjoy the CHOICE of eating and the CHOICE of paying for!
That is all the user of software ever does... is to go out into the market to
get software that is written better than the other software to make a task
easier. If there is a patent protected system where bad software CAN BE and IS
protected (in the US)... then we just have no choice but to choke down bad
software for as long as the monopoly is granted to the patent owner (not always
the AUTHOR) of the software! The users of GPL's softwareware are using it NOT
BECAUSE of the PRICE ... but, because of the fact that there is a LOVE
relationship in the way the meal is prepared. The end result shows thru in the
quality (a level of quality that you only maybe come close to when you have paid
for years of spoon fed and costly upgrades from some proprietary vendors).
Part of our eating experience is the SERVICE we get with the meal. The better
the meal and the service the chances are the more we ACCEPT that we will pay for
this combination. Software is like this too. Users need service. And we pay
for this service EVEN if the food is FREE! Why? Because it is better than we
can do at home... (both the food and the service).
Linux allows both GPL and proprietary (and other) licensed software to run on
top of it. It is a common OS that allows Freedoms of CHOICE. We who eat it up
LOVE that choice. Going back to the days of being force fed by the likes of a
Microsoft (who used us over and over again as beta testers) is simply
unacceptable.
What goes around, comes around. Proprietary software companies, due to their
greedy actions of their pasts created a situation where users started looking
for other choices. Obviously users are not happy with Microsoft and others past
practices (along with their EULA that forced us to choke down their mistakes).
Laws to do away with patents and to limit the language of almost patent creating
EULA's must happen if we are to live and enjoy software (all of us) ever again.
Hey proprietary software companies... are you afraid of creating good software,
software so good that folks will not mind paying a high price to enjoy it?
In the end, bad software (and Microsoft can show us a few examples of that) that
is protected, is the reason why this choice is being demanded. Folks are sick
of being forced to eat what is an ugly situation AND being forced to pay for it
because it is such a monopoly that there is not other choice but to PAY (to
much) for something, that is cases, is so bad! Folks just have decided they
can do it better somewhere else and maybe do it themselves, and so they go to
the market, get some pieces of a GPL's software project... and they innovate and
create to their hearts content. They and their friends enjoy! And some of
their friends even PAY for this enjoyment. How can this be bad?
Competition is what is needed in the authorship of software. Removing software
patents from the battlefield will allow for an aroma to exist that is the best
smell, leading to the best meal, that the world can imagine.
Copyrights will protect the authors of such patent free software. Like books,
the millions of lines of code in sofware tell a story, and different authors can
have different ways to tell that story better and faster.
Patent protection for bad chefs and bad software writers should never be
endorsed by any legal body. Let choice be free again. BAN all software IDEA
patents... NOW.
Let copyright prevail, let quality prevail under it's protection.
ps - let's also set reasonable lenghts of time on copyright protection as well.
Whatever reasonable means is and should be the discussion.
[ Reply to This | # ]
|
| |
| Authored by: bclemmen on Wednesday, October 20 2004 @ 05:18 AM EDT |
I certainly see the problem with the current system, but Jonathan Scwartz is
correct to point out that there is no clear way to distinguish hardware and
software. Thus, "no software patents" isn't a solution unless what
you are really saying is "no patents".[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 05:20 AM EDT |
I think Jonathan Schwartz raises a fair point. After all hardware manufacturers
often have huge sums invested in new bringing new products to market. Any
solution to the patent problem has to account for this as well as FOSS.
If removing software patents ultimately means hardware companies can't protect
their investment, as Jonathan seems to be worried it will, it's going to be very
hard for Sun, Intel, AMD, et all to justify the R&D costs for new projects,
and this will be detrimental to the hardware industry at large.
Of course Jonathan is standing on the side of patents. In his position, he has
to, but that doesn't mean we shouldn't listen to his viewpoint. Reading his
blog, he doesn't seem anti-FOSS, he just makes it clear that he is always going
to act in the best interest of Sun.
A solution to this problem is going to need a lot of debate and input from all
sides. I only hope that a compromise can be found that suits us all.
Ross Smith[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 05:34 AM EDT |
Mr Schwartz raises a good point, though I come to a different conclusion. As he
says, there is often a choice as to whether function should be hard coded in
silicon, implemented as microcode or executed as programs. The algorythms are
much the same. The engineering to translate the algorythms into silicon or
microcode is routine. [The most common tradeoff is flexibility versus
performance. Functions in silicon are fastest, but any bugs are very expensive
to fix. Microcode covers a middle ground. Programs can relatively easily be
patched but are slower.] Should be patentability of algorythms be determined by
the medium in which they are implemented? Mr Schwartz would say "no".
So would I. At this point we part company.
Patenting of algorythms should be banned -- period.
The patents of the past related to innovative ways of engineering things.
Whether the patents really spurred innovation (or as I suspect inventions would
have been made anyway) is arguable. However, it is possible. Most important, a
patent on a new kind of widget cannot prevent others from developing whole
classes of machines. Many software patents have the potential to totally
paralyse the industry.
It should be noted that, with the speed at which software development is
progressing, software patents are ridiculous just because of the time they
remain active. Many generations of software products will be developed, used
and obsoleted during the lifetime of the patent. Noone who thinks about it can
really believe that a software patent will, by revealing its secrets, spur
others to develop software further. On the other hand, copyright provides
appropriate protection. Anyone sitting down to write (say) a new programming
language, without copying the code of others, has a time-consuming task ahead of
them. Patents can prevent a decent new language being created. Copyright
ensures that the new language will only be available in a useable form after a
few years of development.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 05:36 AM EDT |
http://www.nosoftwarepatents.com/en/m/intro/index.html
[ Reply to This | # ]
|
| |
| Authored by: mosborne on Wednesday, October 20 2004 @ 05:43 AM EDT |
Software is simply a unique arrangement created from a finite set of
instructions. Likewise, music is arrangement from a finite (realistically) set
of notes.
MOV 0,R0 is analogous to A#
I fail to see how this these differ. If you can patent software then you should
be able able to patent basic music arrangements as well.
A similar argument could be made for writings. There are finite number of words
as well.
[ Reply to This | # ]
|
- How is software different than, say, music? - Authored by: Anonymous on Wednesday, October 20 2004 @ 07:48 AM EDT
- How is software different than, say, music? - Authored by: EzInKy on Wednesday, October 20 2004 @ 07:56 AM EDT
- How is software different than, say, music? - Authored by: tiger99 on Wednesday, October 20 2004 @ 01:41 PM EDT
- YES! YES! YES! - Authored by: josmith42 on Wednesday, October 20 2004 @ 09:59 PM EDT
- Its a good analogy, but not carried out correctly - Authored by: Anonymous on Wednesday, October 20 2004 @ 11:32 PM EDT
| |
| Authored by: thorpie on Wednesday, October 20 2004 @ 06:00 AM EDT |
From research I have discovered that undertaking particular bodily functions
whilst in the lotus position relieves constipation. This method of relieving
constipation has never before been documented, therefore I can claim a patent on
it. This I have done. Using this method saves individuals the cost of
laxatives, estimated at $3.00 per person for a pack every 2 years
Now
I own this “intellectual property” I have the right to collect royalties. I
have a right to intrude into your life to check whether you are using it and to
claim these royalties. I have a number of options.
First option, I
could put a camera in each and every loo in the world to identify people using
my patented method, and charge accordingly.
Second option, toilet
manufacturers are encouraging people to use my IP. I therefore want a
surcharge imposed on all loo manufacturers based on the estimate that on average
everyone is constipated at least once a year. With a cost of only 20c per use
of my IP, an average of three people using a loo and an expected loo life of 20
years this gives a surcharge of $12.00 per loo.
Third option, I can
charge only commercial usage, every factory, office block, airport, public loo
in the land. An annual fee of $12.00 will apply for these loos because they
are used so frequently.
Of course commercial organizations can avoid the
lotus surcharge by installing the “anti-lotus” attachment to loos. This
attachment stops people sticking their knees out sideways on the loo. It is,
surprisingly, manufactured and marketed by a company I happen own.
Have
a pleasant day, we are looking at undocumented methods used by humans for all
bodily functions. We will one day own an actual part of you.
It isn’t
only software patents that suck, it the whole IP system! NOIP.
---
The
memories of a man in his old age are the deeds of a man in his prime - Floyd,
Pink [ Reply to This | # ]
|
| |
| Authored by: jig on Wednesday, October 20 2004 @ 06:38 AM EDT |
i've probably chosen to write about this a little before
i've formed the
thought completely, but here goes.
i think that the elder, smarter
legislators (the ones that
have lasted long enough to realize investment in
research is generally more important in the long run
than most of the
op-ed stuff they usually win elections
with) have tried very hard to step
back from changing
patent law drastically because of what is happening
with govt. funding in science.
-jig
there is a push from within
for the govt. to incentivize
corporate investment into academic research as
much as
possible becasue the govt. wants to spend its money
elsewhere.
part of the imagined package is for the
companies to rely on an income
stream from patents
gained through the research. if something happens, like
the threat of a large restructuring, that seems to
disrupt that revenue
stream, then they fear the
corporate sponsorship will leave.
that,
coupled with a lack of funds and a general
distraction of domestic issues at
the moment is why
probably nothing will get done in legislation
soon.
more on the same thought: i think some think that allowing
US
researched IP to escape the nantional bounderies
without enriching citizens
directly is bad. they know
that the govt. can't own patents or won't know
what to
do with them if it could, and so one way to 'fix' the
problem
without creating a bunch of bean coutning laws
is to make the research
beholden to american companies,
through patents.
colleges take it
all a step further by trying to keep it
within the schools ip boundaries,
because they know
that the govt is also trying to back away from funding
them as well.
the deal is, software patents are the most bang/buck.
you
buy a genius a couple $2000 computers, and you can be
part of a
patent that controls all interaction on the internet.
so, to fight
software patents, you need to fight the fear,
and the greed, the sense of
leaking competitiveness,
the current disarray of congress, and the national
defecit.
it isn't an impossible fight, but you have to know what you
are fighting enough so that you can argue good solutions.
all
polititions believe in taxes, no matter what they say.
revenue is still
revenue that's taxable even if it
comes from software patents. whatever
'greater good'
argument you want to use has to be very, very
strong.
finally, as an aside, i'd like to ask that you read this
other post
here
so i can get some feedback.
thanks.
[ Reply to This | # ]
|
| |
| Authored by: If_Only on Wednesday, October 20 2004 @ 07:07 AM EDT |
Can I take the opportunity to say wonderful site and an impressively high
level
of discussion
I began without any position on this, but I seem
to have convinced myself
that software patents are not good. Perhaps someone
can correct me.
Patents inhibit competition, that is their purpose
after all, they grant the
right for a limited period to stop others from
making, using or selling the
invention without the permission of the
inventor. They are supposed to provide
a balance between rewarding a
substantial contribution to society { public/state/market
to taste } and
protecting that society from extortion. Of course, failure to
receive a
patent for an idea would not preclude other forms of protection and
reward.
Copyright has been used successfully to protect and reward composers,
authors
and other creators of art for a long time. The software and computer
industries have grown into multi-billion dollar (/euro/yen/pound ....)
enterprises
and major factors in our world without software patents. So the
onus must be
on those who wish to change the rules to first demonstrate the
need for change
At “component” level the individual ideas in
software are either
compulsory
{ mathematical function, definition
by language, externally determined values
(signal to machine, flag to other
program, empirical quantity, etc), etc }
or arbitrary ( eg name
of variable, order of definitions, choice of looping
method, etc )
The creativity in software is in the orchestration of these ideas by
composition,
expression and design (It is no a coincidence nor even an
analogy that good
software is said to possess elegance and creativity).
This is similar to music, economics, politics and for the same reasons has
until
now been excluded from patenting. { Imagine a world where patents on
political
ideas could be enforced or a musical genre was owned }
But
the biggest problem I think is that software development is cumulative
and
iterative. If I make a “Better Mousetrap” that dominates the
market for a while it does not in anyway stop you developing your own entirely
independent “BeTTer mouseTrap”. In software however where a
creation
becomes dominant new programs need to adhere to the protocols etc in
order to
function in the environment. It is important to realise that this
can apply
not just to competing products but to any subsequent invention that
requires
interaction with the original, alternative or other subsequent
programs or just
their files.
Any invention not just software that
becomes fundamental, universal or inherently
monopolistic or where it
conflicts with broader rights it should essentially
become universally
available without restriction for eg by becoming public domain. {
imagine for
eg voting rights dependent on sender ID locked in to a particular
service and
restricted to certain products and the potential for coercion and
corruption}
How ( if ) you compensate the inventor is a separate issue
from the principle
of a free society.
There is also a practical
objection to software patents which is important
and basic. Because software
ideas are free of the physical world, they are not
restricted by the usual
constraints. This means their expressions or implementations
have far greater
degree of freedom than any physical invention could have. (
Any physical
device must obey the rules of nature, an inventor must consider
stress,
material strength etc. ) A patent that practically described a software
creation would have to attempt to cover alternative implementations without
becoming vague enough to encroach and threaten the broader field of endeavour.
Once that happened patents could and would be used to intimidate and
inhibit
competitors. This is when new, independent or smaller operations are
driven
from the market. Of course an individual can still come up with an
idea but
will nearly always be under duress to sell out quickly before
threatened with
litigation.
f [ Reply to This | # ]
|
| |
| Authored by: mickeym on Wednesday, October 20 2004 @ 07:29 AM EDT |
Program Code Is Text
Program code is not different from a novel, or any other kind of literary work,
because it is text. A writer today enters the characters from a keyboard into a
file and then assigns a name to the file. The End.
There is no functionality in the text itself. The text itself doesn't do
anything, and the author didn't actually do what was described in the text,
either, simply by typing the words. At some later time, and then only when
combined with human conduct, there may be measurable activity. Until then, the
only action that has occured is the preparation of text on its way to becoming
fixed into a tangible medium.
If you allow a functional quality to be given to a section of text, the next
stop on the logic ride is where a verb becomes conduct, even construed as human
conduct on the part of the author. Software patents try to create real actions
from action words.
Text Is Not Technology
Program code may appear complicated, and it takes training to understand it
because it has rules, but this is not a distinguishing feature as compared to
poetry or recipes. Program code is referred to as "soft", as opposed
to "hard", for a reason.
We use words like "virtual" so that one may distinguish real things
from unreal things. Virtual reality is not real reality, similar to the way a
calculator program is not a calculator and a cake recipe is not a cake. Text is
not a "device", nor is it a "tool" that can be used to do
something. Software isn't really real, it's just a collection of characters that
describe something.
Conduct Vs Speech
Conduct is more easily regulated than speech. I believe that some of the binds
in which we find ourselves extend from the idea that program code is conduct,
which can then be regulated. This, to me, is where the rub against both free
speech and free press enters the picture, because the ability to regulate text
based on its "functional qualities" allows for freedoms, previously
taken for granted, to be revoked with the backing of the force of law. Now, even
text-based discussion about software routines has been redefined as
"trafficking" in "devices."
A Suggestion For Reform
Once upon a time, there was a requirement for a patent to be "reduced to
practice", where a working model was demonstrated. I used to picture the
patent office waiting room as one with several hopeful inventors, each with an
odd contraption in his or her lap, waiting for a turn to demonstrate a new
widget to the examiner.
This once necessary hurdle is now removed, and, with the definitions of devices,
conduct, technology, speech and literary works nicely scrambled together, the
courts are having a difficult time resolving the outcome. Attempts have been
made to steer patents back into the physical world, but still more virtual
patents are granted. Many of the software patent reform suggestions that I have
read try to leave room for some special cases, but I don't think there are any.
I suggest that we go back to requiring a demonstration of a working model.
Software, in and of itself, doesn't do anything. It's just like a story or a
recipe. The actions described in the text don't happen, and might not ever
happen, until human conduct puts them into play, whether by computer or mixing
bowl. Verbs are not conduct.
mickeym[ Reply to This | # ]
|
| |
| Authored by: Acrow Nimh on Wednesday, October 20 2004 @ 08:30 AM EDT |
Main posts in this thread may only be made by senior managers or attorneys for
"The SCO Group". Main posts must use the name and position of the
poster at "The SCO Group". Main posters must post in their official
capacity at "The SCO Group".
Sub-posts will also be allowed from non-"The SCO Group" employees or
attorneys. Sub-posts from persons not connected with "The SCO Group"
must be very polite, address other posters and the main poster with the
honorific "Mr." or "Mrs." or "Ms.", as
appropriate, use correct surnames, not call names or suggest or imply unethical
or illegal conduct by "The SCO Group" or its employees or attorneys.
This thread requires an extremely high standard of conduct and even slightly
marginal posts will be deleted.
P.J. says you must be on your very best behavior.
---
Supporting Open Sauce since 1947 ;¬)[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 08:31 AM EDT |
PJ, I really loved your metaphor about "playing Patent Cold War". I'm
envisioning a new board game on lines of Monopoly (TM, Milton Bradley?).
Instead of purchasing real estate, you purchase a particular software patent.
Instead of purchasing houses and hotels to "develop" your property,
you "develop" software packages and suites. Anyone landing on a
square with a patent that you own has to pay a royalty. Cards can allow things
like "You've been found in violation of the RICO act, go directly to
Jail." Playing pieces, rather than car, dog, tophat, etc., can be logos of
various companies (fictitious ones, of course; it wouldn't do to infringe on
anyone's trademark).
Actually, I think that this could be quite an effective teaching tool to show
the disastrous effects of software patents. Consider, for example, what it
would be like trying to suddenly enter a Patent Cold War game (or Monopoly, for
that matter) as a new player, with your little bit of stake money, well after
most of the "property" has already been bought up and developed.
You'd last maybe two times around the board before you went bankrupt and lost.
That, unfortunately, is the position that many individual developers and small
companies find themselves in now when trying to develop new software and compete
in the software industry against predatory software giants and submarine patent
holders.[ Reply to This | # ]
|
| |
| Authored by: NZheretic on Wednesday, October 20 2004 @ 08:37 AM EDT |
| as per title [ Reply to This | # ]
|
| |
| Authored by: Pat Pending on Wednesday, October 20 2004 @ 08:38 AM EDT |
Question: Once the IP Gold Rush behemoths start invoking their insane arsenals
given them by the USPTO in some Marvel Comics inspired plan for U.S. companies
to own Earth, how long before no one can develop software for their platforms at
all?
Patent gridlock will eventually stifle development of commercial software as
well as FOSS.
Will a micropayment chip need to be installed in our brains, in case we think of
something?
It looks like mutually assured destruction for the big boys, meaning they will
need to shake down the peasantry to pay for their Gold Rush skirmishes.
Remember kids: duck and cover!
---
Thanks again,[ Reply to This | # ]
|
- The GPL weapon - Authored by: Anonymous on Wednesday, October 20 2004 @ 10:39 AM EDT
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 08:50 AM EDT |
| There's no section yet, so I will start one now. [ Reply to This | # ]
|
| |
| Authored by: psherma1 on Wednesday, October 20 2004 @ 09:27 AM EDT |
| :( [ Reply to This | # ]
|
| |
| Authored by: dcs on Wednesday, October 20 2004 @ 10:08 AM EDT |
It's not simply about GNU/Linux. My question is simple:
Why does the world need software patents?
I mean, why does it *need* them?
Because it's certainly not to give any incentive to software development. See,
for instance, the world, where the benefit of writing your own software is
offering the feature to the public months before the competition, a mighty
benefit indeed.
---
Daniel C. Sobral
[ Reply to This | # ]
|
| |
| Authored by: skip on Wednesday, October 20 2004 @ 10:31 AM EDT |
If all avenues to a particuler technology are blocked by patents or whatever,
this could kill off commercial companies, but not FOSS.
FOSS is a creature of many parts, highly resistant to bits being cut off. If
something gets taken from it, it just produces an alternative.
Some area's of current technology might be blocked due to extreme predation by
patent holding freeloaders, but they couldn't kill off our own OS. Even if some
terrible blow hit Linux,then the unnaffected bits could be taken and used in a
new non-infringing version. It could take a few years in the worst case, but
this wouldn't be an issue, not for FOSS.
Just say, to pick a wildly hilarious and unlikely example, that SCOX got Linux,
or at least managed to prove that they were entitled to have part of it. What
good would this do them? Linux is written by thousands of people, all of whom
hold copyright over their works. Thus each could deny SCOX the right to use it.
This would be a nightmare scenario for McBride, and I imagine was the real
motivation for his trying to get the GPL invalidated. No GPL, no issue, as it
were.
In the fictional event that SCOX took a couple of headers and a file with SMP in
the comments or whatever away from linux (they can't legally take the whole
thing now and close source or binary only distribute it, as the GPL still
stands), the wider FOSS community would re-code the missing bits, and Linux
would go on over the corpse of SCOX. After all, Linux was indeed produced by
many people, many more then SCOX have available, so they'd never manage to keep
up with us.
Of course SCOX would have to say what bits they wanted in order to acheive the
above (Revealing precious code/proof along the way), yep, that old chestnut, and
seriously not their style :-)
I don't think patents are too big a problem. Where's the percentage in spending
vast amounts of money enforcing patents against the wider FOSS community. You
couldn't kill it all, and no sooner would you shut one project up, then another
would spring up to replace it. That's assuming you manage to survive the review
process a'la groklaw or not have you patents invalidated.
A much safer bet is to chase the commercial closed source companies. They have a
lot more to lose, and are far more likely to just pay up, rather then risk their
business.
---
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"Mumbo, perhaps. Jumbo, perhaps not!"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~[ Reply to This | # ]
|
| |
| Authored by: beserker on Wednesday, October 20 2004 @ 10:33 AM EDT |
| Many years ago (back when I was a lot less grey) I read a series of articles on
the patent system written by Don Lancaster. For those of you who have never run
across him, he was a well known '70 era hardware guru and at one time or another
I have had most of his electronic project books. I recently ran across his web
site and he has quite a lot of his old stuff archived there.
His views on
patents are from the point of view of the individual inventor or small startup.
Also note that these articles are from the '90-'96 time frame and software
patents weren't a big topic at the time. Even so, I've found his insights on
patents in general to be useful. A line from one of his articles that has always
stuck with me goes something like (paraphrased) "Patents do not
prevent anybody from copying or stealing an idea. The only
right it gives you is the right to sue an infringer in a civil action."
This link is to a set of patent articles Case Against Patents
His
main site also has lots of stuff that any true EE geek will love:
http://www.tinaja.com
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 10:34 AM EDT |
I remember learning in 9th grade physics that all machines decompose into
six fundamental machines. Chemical engineering is essentially how to coax
an
electron to go from one atom to another. Software isn't math. Software
consists of instructions to a processor to get something from memory (read),
add, multiply, or test, and put something back into memory (write).
Mathematics can provide an understanding into approaches to a problem, but
one is applying human understanding in order to guide the computing device
as to what to
read, test, and write. Think of a program to find the derivative of a second
degree polynomial (ax^2 + bx + c). For those of us who took and remember
freshman calculus, we
know that the value is 2ax + b. The program we write will not care about the
analytical
underpinnings of what happens as the limit of something approaches zero.
(Though that does give a hint as to how to solve the questions for equations
that do not have analytical solutions.) The program will read the tokens of the
equation, check that the input matches a recognized pattern, pass the tokens
into registers, do the adding and multiplying, and write out the result.
Yes, I whole-heartedly agree that something has gone wrong with the US
patent system and it is very much pointed in the wrong direction. But the
trouble with software patents isn't that they are or aren't mathematics. It may
be that all software, in my opinion, is read, test, calculate, and write. (But a
new apple peeler is just adaptation and recombination of the basic machines
of physics, so there's something not quite right about that reduction as well
when understanding the issue).
With high regards and all due respect,
D. Orias, Los Angeles, California[ Reply to This | # ]
|
| |
| Authored by: bsm2003 on Wednesday, October 20 2004 @ 10:41 AM EDT |
| To make the Patent system available for ALL base the fees on a %
percent scale of the net income from the products that are being sold. 10%
should be fair. that way everyone has the oportunity to participate in the
system. [ Reply to This | # ]
|
| |
| Authored by: blacklight on Wednesday, October 20 2004 @ 10:42 AM EDT |
"I believe in intellectual property. In my view, it's the foundation of
world economies, and certainly the foundation upon which Sun Microsystems was
built. Copyright, trademark, patent - I believe in them all") Jon Schwartz
I thought that Sun Micosystems became a Fortune 500 on the back of BSD UNIX - Do
I stand corrected? And what good to Sun are Sun's patents if enough of Sun's
customers are rushing out the door and migrating to Linux to make a systemic,
significant negative impact on Sun's quarterly bottom line? Aside from that, the
foundation of the world's economies is free and fair two-way trade, not one way
buying based on artificial monopolies of dubious legitimacy.
[ Reply to This | # ]
|
| |
| Authored by: qu1j0t3 on Wednesday, October 20 2004 @ 10:53 AM EDT |
One little quibble. I would say a Cold War isn't "won"; it's
eventually "lost".[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 10:58 AM EDT |
Schwartz writes:
"There's this great irony: spend time with Sun's microprocessor team, and
guess what - they're all writing software. Spend time with our data center
switching team - they're all writing software, too. Viewed simplistically,
computing hardware is software burned into and onto physical things. And over
time, more and more routine software elements end up in hardware, for
acceleration or optimization. SSL accelerators, JVM on a chip, you name it. So,
where do you draw the line on patents? Firmware? FPGA's? Silicon?
Systems?"
Schwartz has a point here. But I think he is overstating it a bit. Firmware is
a lot like software. FPGA design is somewhat similar to software, but
microprocessor design is quite different, and probably worthy of some patents.
I have done a bit of hardware design (ASICs & FPGAs) and it is substantially
similar to software development. If you were designing a microprocessor chip
for a cellphone for instance, you would start by writing some code (referred to
as "RTL code") in a hardware description language like Verilog or
VHDL, then you feed it to a compiler (referred to as a "synthesis
tool") which produces a "netlist" -- a list of parts (logic
gates, registers, pads, etc) that implement the code along with a description of
how the parts are interconnected. You can also feed the code to a simulator,
which compiles it to executable code and runs it on your workstation, allowing
you to debug it the same way you would software.
But the similarities between software and hardware end there. When you write
software, once it compiles and runs properly, you are done. When designing a
microprocessor, once it synthesizes and simulates properly, you have done about
20% of the work. After you have completed these steps, you have to worry about
the physical implementation of the hardware. You have to check that the code
you wrote will result in a chip that runs at the required clock speed. You
can't tell this just by looking at the code, you have to do some pretty
time-consuming analysis. This means that you can't just place all the gates
wherever you want on the chip -- if the interconnection between two particular
gates is too long, the signal may not arrive fast enough. Worse, if you expect
two signals to arrive at roughly the same time, and the interconnections are
different length (they pretty much always are) you will have to re-write your
code so that it doesn't care whether the signals arrive at different times. You
also need to consider how much power your chip will use. Is it going to drain
the battery too quickly? Will it overheat? You may need to change your code to
take this into account, or make changes to the layout, or choose a different
manufacturing process. And what about crosstalk? If two wires inside the chip
are close together, you get an effect known as "cross-coupling
capacitance", which can disturb the signals enough to make the design
un-workable. You may need to adjust your layout accordingly. And keep in mind
that manufacturing chips is not a 100% reliable process. A certain percentage
of the chips you make will have manufacturing defects in them, or they may not
run as fast as you hoped. How are you going to determine which ones are
defective or slow? You can't tell by looking at them, and just because it works
at room temperature, doesn't mean it will work in the middle of a hot summer in
Arizona or during a cold winter in Ottawa -- temperature can have a huge effect
on electronics. So you need to design your chip with a built-in self-test that
you can run in a couple seconds (possibly at a couple different temperatures) to
give a "yay" or "nay" before the foundry ships it to you. A
good test should check about 95% of the interconnections inside the chip.
Also, hardware is not portable the way software is. You have to be very aware
of the manufacturing process you are using. If you are going to get your chip
manufactured by TSMC you will have to go about it differently than if you were
working with IBM or LSI. You will have to use different tools, different
libraries, follow different design rules, etc. Once you have completed your
design for TSMC say, you can't just switch to IBM without re-doing at least half
the design work. This is very different than software, where code written in C
can be compiled on many different platforms with only modest effort, provided
that the developers have designed with portability in mind. There really is no
such thing as portability in the hardware world.
Now lets talk about cost. C compilers are free if you know where to look for
them. Even MS offers a stripped-down version of their C compiler for free I
believe. But as of this writing, there are very few open-source hardware
development tools in existence -- and none that are widely used commercially. A
full suite of tools can easily cost a million dollars. The chip foundry may
charge you another million dollars to develop the masks for you chip. Add to
that a couple years worth of salaries for your design team. The payoff is that
once you have finished designing the chip and have working prototypes, you can
manufacture more of them for only a few bucks apiece. But this is still more
expensive than software, which can be mass produced for essentially neglible
cost.
So I think that hardware is still quite different from software. Software is
almost pure thought -- like mathematical equations. But hardware is a physical
thing, and it requires a significant investment to build a piece of hardware, so
I think it may be reasonable to grant patents for hardware, provided that they
are for hardware-related concepts (e.g. a way to reduce cross-coupling
capacitance) as opposed to algorithmic things (e.g. bubble sort in hardware) and
are non-obvious. I think Schwartz is being a bit deceptive by saying that
software and hardware are the same. There are similarities to be sure, but
there are also many, many differences.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 11:01 AM EDT |
When the rules are impossible to follow, they are not followed. When those that
enforce the rules are unable to enforce them, they have no meaning. I don't
know if I am looking foreward the the future if the issues are not resolved now.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 11:06 AM EDT |
I greatly enjoy Groklaw because
of its fine legal analysis and attention to
legal detail in
defending FOSS.
I am keenly interested in seeing
this discussion of legal
analysis and attention to detail applied to the
issue
of software patents as well. I have a number of
questions with respect
to FOSS and patents, including
the possibility of using patents to defend
FOSS.
These questions are also of specific personal interest
to me, because I
fall on two sides of the issue on two
different specific cases. In one case, I
have mostly
given up on some promising research directions in support
of
open-source computing infrastructure because
of patent concerns (i.e., an array
of what I consider
to be undeserved patents in an important area of work).
On
the other hand, I am also contemplating a patent
application for technical work
that may improve
performance dramatically in a particular application
area. I
have in mind the possibility of using such
a patent as a tool to defend FOSS.
Something along
the lines of only allowing FOSS implementations
of the patented
concept unless a specific license
is negotiated.
To me, patent abuse is
a serious problem
just as e-mail abuse is. However, I think
it is reasonable
to be opposed to e-mail abuse
(spam) without being opposed to e-mail.
I think it
also reasonable to be opposed to
patent abuse without being opposed to
software
patents, per se. That is, patents of ideas
that are trivial, obvious
and/or known in the
prior art are clearly abusive. But this does
not mean that
the concepts such as 'novelty'
'lack of prior art' and 'nonobviousness'
cannot
be applied to software. Indeed, I don't see
anything in these concepts
that is fundamentally different
for software than for any other form of
technical
creation. Perhaps it is true that software
may require less
perspiration than invention
in other areas, but isn't it also true that
patents
are only (supposed to be) granted on
the basis of inspiration not
perspiration?
I'm reluctant to go much further. This is PJ's
space and
she has opened with a comment
that patent cold wars guarantee domination
of the
software world by one or a few companies
and the doom of FOSS. I think there
is some
reason for fear and uncertainty. But how
is it even conceivable that
continued patent
cold wars (more and more new patents) could
affect the existing
base of FOSS? Isn't existing
FOSS prior art and hence self-defending
against
future patents? As for current patents,
what is the real state of
serious patent
threats to FOSS? Of the 283 that have
been identified at issue
for Linux, are there
any actual cases being brought? And which
patent-holder
is going to volunteer to be
the next SCO and bring such a case?
Is there really
any reason to believe
that Richard Stallman's bizarre calculation
of 100,000+
patents affecting the GPL'd code
base, or is it just fear-mongering?
In
other venues, I might be OK with FUD
on software patents, because it may be
useful for
raising awareness. But this community is unique
in its broad
interaction between legal and software
professionals and I'd sure like to see
deeper
discussions here.
[ Reply to This | # ]
|
| |
| Authored by: rsteinmetz70112 on Wednesday, October 20 2004 @ 11:13 AM EDT |
<rambleon>
In the software patent debate a couple of themes keep coming up, which I'm not
sure can be supported.
The first is that patents should not apply to software. When the existing patent
system was conceived there was nothing like the programmable computer, although
it had been conceived long ago as a theoretical possibility. Even those concepts
did not envision software as we now understand it. Had Babbage been able to
construct his Computational Engine, it surely would have been patentable, as
were the mechanical computers developed through the first half of the twentieth
century. The first use of electronic computers was to emulate, in a newer faster
technology, these machines. Today we have computers which can emulate virtually
any mechanical system, with the addition of appropriate instructions.
It is hard for me to conceptualize how a mechanical computer can be patented and
the combination of an electronic computer and a software program which performs
the same function cannot. It follows for me that if the electronic computer is
a general purpose device, then the software portion must be a patentable
improvement to the basic device.
It is important to note that, as I understand it, patents cover only a specific
device and that other devices performing similar functions can still be
designed. One need only look at the number of different mouse traps which have
been patented. The concept of trapping a mouse is not protected, only the
specific device for performing the action. There may be some trade offs and
there may be some additional cost or effort, but most patents can probably be
designed around.
The other theme that runs through these discussions is that somehow FOSS or more
specifically GPL should be somehow a privileged class of software. This seems
equally fraught with complications.
It is generally conceded that when working with computers there are usually two
ways to implement any given concept, in hardware, which is faster and more
expensive and in software, which is slower and generally cheaper. If GPL
software were given privileged status, then is would seem trivial for anyone to
bust any computer patent by implementing the patent in software and releasing
it under GPL. A clear end run.
It also seems to me that some innovations are so important that they ought to be
protectable, both for the benefit of the inventor and for the benefit of the
public. As an example let me cite Mosaic, the first widely used graphical web
browser. Developed at NCSA, a publicly funded research center, it was the basis
for Netscape and for Internet Explorer. Had it been patented, NCSA could have
guided the development of the WWW in its early days by licensing the technology,
this might have averted the incompatibilities still plaguing used of the Web.
One of the great achievements of intellectual freedom has been the development
of the GPL which takes advantage of the existing copyright law and allows the
individual author to use their legal rights to decide how the work will be used.
The real genius is that is doesn't require any new law. This is partly because
any work covered by copyright is automatically covered at the time of creation,
requiring nothing of the author.
Unfortunately the patent system does not appear to offer a method of doing the
same thing, due to the registration requirement and the cost of the requirements
for registration. The only way to solve this problem is to reform the patent
system.
The patent system is broken in part because of the high cost of obtaining a
patent and the requirement for registration. This is complicated by the first to
file rule, which presumes that the first one to the patent office is the
inventor. This is obviously unfair and is, if I recall correctly, a concept
imported from Europe, where law is based to some extent on seeking permission.
US patent law used to be based on the first to invent, which is much more in
keeping with American ideas of individual merit, achievement and
responsibility.
A system more similar to the copyright system might help solve this, where
invention is the important event, not registration. In fact I can see no
impediment to a system similar to the current copyright one where registration
is optional, and necessary only to press a claim. The registration could be
challenged at trial by prior art, and obviousness similar to the filtration test
employed in copyright cases. Other types of challenges such as prior invention
(not publication) or abandonment could be inserted to prevent some of the
current abuses.
Perhaps there could be a preliminary examination step, by the patent office to
validate a patent, in the context of the adversarial system. This would be
similar to the idea of a Special Master sometime employed in class action suits
to gather facts on behalf of the court. It could be performed by current patent
examiners and replace the current pre-application patent examination and search.
Those steps only being performed if an actual claim is made.
<rambleon>
---
Rsteinmetz
"I could be wrong now, but I don't think so."[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 11:22 AM EDT |
I read the blog and I see the man for what he is, the head of a wannabe company
that is desperate to be a player. He sees patents only as weapons that he might
be able to use against the overwhelming power of his company's competitors. In
that context, there is no way he would agree that software patents should be
abolished. How then could he level the playing field? To me, he sounds just
like Iran and North Korea telling the United States "See, I have nuclear
weapons too".
His criticism of IBM and Novell is just a transparent lust for their power and
his wish that he could harness FOSS as they have to improve the profitability of
his company. He also knows that if IBM and Novell decided that software patents
should be abolished, he'd be powerless to prevent it unless he sides with
Microsoft against them.
As to his claim that they settled with Kodak "to protect their
customers" I'd say baloney, that was necessary to protect the value of
Sun's only real asset, Java, and to keep Sun from being ravaged by Wall Street.
It will be interesting to see which (if either) of Sun's competitors, IBM or
Microsoft, think there is enough value in Java to buy Sun.[ Reply to This | # ]
|
- Beg to differ - Authored by: Anonymous on Wednesday, October 20 2004 @ 01:24 PM EDT
- IBM and Sun. - Authored by: Brian S. on Wednesday, October 20 2004 @ 01:52 PM EDT
| |
| Authored by: OmniGeek on Wednesday, October 20 2004 @ 11:38 AM EDT |
OK, that's a very strong statement, so it needs a strong argument to back it up.
Here goes an argument based entirely on economic self-interest, which even
bean-counters ought to be able to understand. (There are also powerful ethical
and social arguments for FOSS and against software patents, but others have
addressed them well).
Software patents are fundamentally incompatible with Free and Open Source
software (FOSS), which depends on the free usability of innovative software
concepts. Nation states will, by their laws, inevitably favor either the one or
the other, by permitting or prohibiting patents on inventions purely in
software.
Some nation states will choose FOSS over software patents (think China, India,
Brazil, quite possibly the EU), frequently because of the "Free" part:
they won't be beholden to proprietary, unreliable, insecure American software
products with Heaven only knows what government spyware built into them (if you
think that's tinfoil-hat country, read up on Crypto AG and the NSA; also mnote
India's domestic encryption industry, specifically developed to ensure
independence from suspect US products).
Those countries will enjoy the technical and economic benefits of FOSS, and will
in turn advance its rapid development. There are enough skilled folks (in India
and China alone!) and a large enough software market to ENSURE that Linux and
its kin will be the state-of-the-art operating systems for much of the world.
This development will produce both a technological edge in the software
marketplace AND an educational edge as their schools make educational use of
FOSS as a learning tool; these trends will reinforce one another.
Other nation states (US, Australia, probably the UK) will follow the proprietary
software-is-patentable model, and their software markets will likely be
Microsoft-dominated, 'cause Linux will be illegal there (here). THOSE states
will suffer because THEIR software developers and industries cannot use the
newest, best FOSS technologies, and THEIR schools cannot use FOSS as a learning
tool. (They will also be unable to steal FOSS code for proprietary use, because
its copyright will still be valid even though its use violates software
patents.)
The net result will be that software-patenting, FOSS-hostile nations (most
notably the US) will lose their technological edge to FOSS-friendly nations over
time, and will suffer economically as a result.
In the case of the US, all we really have to offer the world market (aside from
guns and fear) is our technological edge. If we lose that edge, we'll be unable
to sustain the remnants of our high-wage standard of living, and we'll suffer
the kind of decline Britain suffered after World War Two, only worse. We're
already losing highly-skilled jobs to outsourcing; if we put ourselves at a
disadvantage in technology, we will have nothing with which to sustain our
expensive (and wasteful) national lifestyle, and we'll become a miserably poor
backwater. (I think this can fairly be described as a hazard to our national
security...)
From a purely self-interested viewpoint, the US cannot afford to let the
interests of a few software-patent-holding companies -- however deep their
pockets may be -- endanger its national well-being by holding the intellectual
bazaar hostage to a patent system that no longer "advances knowledge in the
useful arts" or serves the public good.
Counterarguments, please?
---
My strength is as the strength of ten men, for I am wired to the eyeballs on
espresso.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 11:52 AM EDT |
If patents would be just about the creation, but they are not as most all
patents today have claims that reach into areas of thought and knowledge locking
out almost everyone. This *win the race to the patent office* does not respect
all IP at all; for many its lost IP, taken from under them. Patents are the
tank, and their claims are the guns. I see a system as patents will do nothing
more then harm.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 11:56 AM EDT |
I just gotta get this one off my chest.
There HAS to be a middle road. Demanding that all ideas, no matter how hard
they are researched and how much time and resources are given to the
implementation of them, be let go into the public domain is not right. Yes, I
said it is NOT RIGHT. Software patents themselves are a bad idea, but a middle
ground that allows a company to patent a killer file system, database
implantation, OS methodology, and things on the low level of software execution
that take huge amounts of time and resources to develop need to have protection
from poaching for at least a time so that companies WANT to make that
investment.
Recently there was an article on the value of the Linux Kernel. Imagine if that
was all done by a proprietary company, the time and money involved. Some
aspects of Linux, yea, there are only a certain number ways of doing it, so it
not patentable, but others that have implemented for the first time (key point
here, first time) that they should be. If someone finds another way to do it,
great, the patent cannot apply due to it being a differing method, but as that
company put the effort and time and money into the first implantation of that
aspect of software, they should have a patent that allow others to license off
of it and to protect their way of implementation.
The Patent Cold War does need to go away, but eliminating the base idea of a
reward for effort, resources and such being driven into new and better ways of
doing things is just going to make any company that does those kind of things
eventually reject the idea out of hand.
We need a middle road, not a knee jerk reaction to kill ‘em all.
Bubba The Barbarian
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 12:04 PM EDT |
If I design a new mouse trap, I have to gave a blue print on how to build it to
the patent office when I apply. Thus the paper work required is:
1) A description of what my device is.
2) A list of claims as to why my device is different.
3) A detailed drawing showing how my device works.
My question is: Why do software patents only require sections 1 & 2?
To be fair, I believe software patents should be required to include a copy of
the code implementing the concept. If the idea behind patents is to make
inventions public once the patent ends, then I should be able to pull up a dead
patent and include it's concepts in my program.
With a device patent, I have a drawing showing me how to implement the device so
I can include it in my device (example: a patent for a coil spring whould
include enough information for me to create my coil over shock along with
intergrated mounting hardware {McPherson's Disposable Suspension System -
commonly called McPherson Struts}.
Why is it that the patent for .GIF files don't include the program for
generating .GIF files? Then the patent dies, I can't just look at the patent and
create a program that generates .GIF file; I have to write the entire code for
making .GIF files from scratch!
My poposal for fixing the software patent system is to require actual working
code to be published along with the patent application. Also note that once the
patent expires, the code would become public domain and would not be covered
under copyright.
Also this would make software patent lawsuits easier, I could just show my code
for doing process and explain why it's different then your way. Your program
generates a sorted list of name and you use a patented sorting method; my
program also generates a sorted list of name. Can you sue me for patent
violations? If you do, all I would have to do is show how my names are sorted
and compare it to your method; if my way is too close to your's you win
otherwise I'm ok.[ Reply to This | # ]
|
| |
| Authored by: Groklaw Lurker on Wednesday, October 20 2004 @ 12:07 PM EDT |
"...There's this great irony: spend time with Sun's microprocessor team,
and guess what - they're all writing software. Spend time with our data center
switching team - they're all writing software, too. Viewed simplistically,
computing hardware is software burned into and onto physical things. And over
time, more and more routine software elements end up in hardware, for
acceleration or optimization. SSL accelerators, JVM on a chip, you name it. So,
where do you draw the line on patents? Firmware? FPGA's? Silicon?
Systems?..."
There is no irony here. The software your microprocessor team and your data
center switching team writes is automatically 100% copyrighted as they write it,
even if they never file for a formal copyright. Every line of that code is fully
and completely protected by the full weight of copyright law.
Adding the protection of patent law to a work that already enjoys the full
protection of copyright law perverts the intention of the framers of the
Constitution by providing dual protection to software, thus hindering progress
instead of promoting it.
This dual protection of both patent and copyright is unique to computer
software. No other form of creative expression is burdened with the protection
of both copyright and patent law. The Congress of the United States did not
intend that both forms of protection be accorded to any single art, whether that
art be software or some not yet conceived form of expression. Nor is such
protection needed.
Lastly, the framers of the Constitution did make one thing perfectly clear. In
Article I, Section 8, Clause 8, the Constitution reads as follows:
"...To promote the Progress of Science and useful Arts, by securing for
limited Times to Authors and Inventors the exclusive Right to their respective
Writings and Discoveries;..."
which very clearly describes the purpose of Article I, Section 8, Clause 8, that
is 'To promote the Progress of Science and useful Arts'. When the application of
patent law to a 'Science' or a 'useful Art' no longer promotes progress, the
application of such law is undeserved. This is obviously the case with the
application of patent law to computer software.
There is really very little dispute, even among the largest software
corporations, that there is inequity and injustice in our present system of
patent law as it is applied to computer software. We have quotes from Bill Gates
saying this, from executives of Sun Microsystems and IBM among many others.
We are faced with a boolean choice here. Either abolish software patents or
perpetuate software patents. If we perpetuate software patents, we have the
boolean choice of reform or adopting a hands off approach and leave it as it is.
I ask you, will any reform ever be able to fully address the inherent
unfairness, inequity and injustice of any application of patent law to the
authors' software and their written code?
I think not.
---
(GL) Groklaw Lurker
End the tyranny, abolish software patents.[ Reply to This | # ]
|
| |
| Authored by: codswallop on Wednesday, October 20 2004 @ 12:16 PM EDT |
| The FTC produced a set of proposed changes to current patent law with supporting
analysis that is vastly superior to what is proposed in any of the 3 recent
patent articles. For each proposal it exlains the nature and history of the
problem discusses the various proposed solutions and the reason for its choice.
The FTC got opinions from hundreds of people and easily spent millions, if you
include the salaries of the staff that prepared it.
I think this should be
the place to start. We're spending time duplicating analysis that the FTC has
already done. Most of the arguments made in the articles and comments are in the
report, since it includes all viewpoints that had significant support.
I'd
like to see an article on it, but it's up to PJ. a good starting link is:
FTC report on patents
The proposals are:
Recommendation 1:
As the PTO
Recommends, Enact
Legislation to Create A New
Administrative Procedure to
Allow
Post-Grant Review of and
Opposition to Patents.
Recommendation
2:
Enact Legislation to Specify that
Challenges to the Validity of a
Patent
Are To Be Determined
Based on a “Preponderance of
the
Evidence.”
Recommendation 3:
Tighten Certain Legal Standards
Used to
Evaluate Whether A
Patent Is “Obvious.”
a. In applying the
“commercial
success” test, 1) evaluate on a caseby-
case basis whether
commercial
success is a valid indicator that the
claimed invention is not
obvious,
and 2) place the burden on the
patent holder to prove the
claimed
invention caused the commercial
success.
b. In applying the
“suggestion” test,
assume an ability to combine or
modify prior art references
that is
consistent with the creativity and
problem-solving skills that in
fact
are characteristic of those having
ordinary skill in the
art.
Recommendation 4:
Provide Adequate Funding for
the
PTO.
Recommendation 5:
Modify Certain PTO Rules and
Implement Portions
of the PTO’s
21st Century Strategic Plan.
a. Amend PTO regulations to
require
that, upon the request of the
examiner, applicants submit
statements of
relevance regarding
their prior art references.
b. Encourage the use of
examiner
inquiries under Rule 105 to obtain
more complete information,
and
reformulate Rule 105 to permit
reasonable follow-up.
c. Implement the
PTO’s
recommendation in its 21st Century
Strategic Plan that it expand
its
“second-pair-of-eyes” review to
selected areas.
d. Continue to implement
the
recognition that the PTO “forges a
balance between the public’s
interest in
intellectual property and
each customer’s interest in his/her
patent and
trademark.”
Recommendation 6:
Consider Possible Harm to
Competition –
Along with Other
Possible Benefits and Costs –
Before Extending the Scope
of
Patentable Subject Matter.
Recommendation 7:
Enact Legislation to
Require
Publication of All Patent
Applications 18 Months
After
Filing.
Recommendation 8:
Enact Legislation to Create
Intervening or
Prior User Rights
to Protect Parties from
Infringement Allegations That
Rely on
Certain Patent Claims
First Introduced in a Continuing
or Other Similar
Application.
Recommendation 9:
Enact Legislation to Require, As
a
Predicate for Liability for Willful
Infringement, Either Actual,
Written
Notice of Infringement
from the Patentee, or Deliberate
Copying of the
Patentee’s
Invention, Knowing It to Be
Patented.
Recommendation
10:
Expand Consideration of
Economic Learning and
Competition Policy
Concerns in
Patent Law Decisionmaking.
I estimate that under
these rules over 90% of software patents would either never have been granted or
would have been quickly and relatively inexpensively rescinded. Proposals 6 and
10 would probably have stopped the patenting of business
methods.
---
IANAL This is not a legal opinion.
SCO is not a party to the APA.
Discovery relevance is to claims, not to sanity. [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 12:20 PM EDT |
| If the Open Source people start playing this game? Patents ultimately are
created by individuals. If there are creative enough individuals out there
to
come up with ideas, supportive of OSS, and with the (modest) amount
of cash to
patent the ideas, what's to keep them from assigning the
patent rights to the
FSF or a similar organization? It's probably tax
deductable even. Furthermore,
it ought to scare the big guys sensible.
The patents war often compares
patents to nukes. To extend that
wretched metaphor to be even worse, IBM,
Microsoft, and the other big
patent holders are like the nuclear superpowers;
while it would be nice
to be the only one, they don't really want to start a
war. Many Open
Source people, on the other hand, want to bring the current
system
crashing down... and are the scary crazy types who will use them.
[ Reply to This | # ]
|
| |
| Authored by: tknarr on Wednesday, October 20 2004 @ 12:30 PM EDT |
My thoughts on software patents would be to modify the system like
this:
- Patents are only allowed on implementations, not concepts. This
is exactly parallel to physical patents: you can patent a specific process for
smelting iron into steel, but not the concept of smelting independent of any
process.
- If there's by neccesity only one way of implementing
something, that implementation isn't patentable. I'd submit that this situation
is so rare it's unlikely to be an issue, but I also think it deserves being
explicit: you can own a monopoly on a piece of the road, but not the entire
road.
- Before granting the patent, the desired end result (not a
detailed description of the implementation in the patent, only a summary of what
the result is, eg. not a detailed description of how the smelting process works
but "We have a pile of iron ore. We want to turn it into steel. Describe how to
do this.") is submitted to a number of programmers or others working in the
field (perhaps a dozen). If within 24 hours of being given the problem one or
more of them outlines the method described in the patent, the patent is deemed
"obvious" and denied. I can't think of a better definition of obvious than
someone working in the field being able to come up with it literally overnight
from a standing start. And if there are multiple implementations that don't
infringe on the patent, then there's no insurmountable problem granting the
patent.
- If the patentholder claims something infringes his patent, we
take him at his word. If it can be shown that what he's claiming infringes
existed more than a year before he applied for his patent, then on the basis of
the patentholder's claim it's accepted as valid prior art and voids the patent
(since you can't patent something that already existed independent of your
work).
Thoughts?[ Reply to This | # ]
|
| |
| Authored by: shareme on Wednesday, October 20 2004 @ 01:11 PM EDT |
as a noteable FOSS supporter, where is IBM in this conversation given its large
Patents on Sofware treasury?
---
Sharing and thinking is only a crime in those societies where freedom doesn't
exist.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 01:51 PM EDT |
The patent system does a better job of dampening creativity than promoting it.
It allows the rich (patents are costly) to remain rich (no one is allowed to
compete by offering a lower price for the same product) and become richer (they
can sue for patent infringement). It does offer a form of protection, but not
for society.
Because of the patent system, _creating_ new products costs too much for medium
and smaller companies (if they try, they lose in court or get swallowed by a
bigger company).
For a simple example, let's say I invented some tool and I did it all on my own
from my own creativity. Whether I sell it or not, one day I'm approached by
authorities and am told that I just infringed on a patent for that tool. What
are the chances that I will want to invent another tool? Why would I risk losing
my life's savings? Next time I want to invent, I will have to hire a lawyer to
make sure it doesn't already have a patent. It doesn't matter whether patent
owner created the tool or just came up with the idea, he still wins because he
has a patent for it.
If two people independently create the same product without knowing about other,
even at different times, then both should benefit from it. The core of the
patent system says that he who first patents the product wins, which sounds more
like a marketing ploy to me..."Order now while there's still time"
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 02:15 PM EDT |
Just because you put software on a chip, doesn't mean you can patent it.
Rather, it lends credence to the argument that firmware and such should be
copyrighted, like code. Now, if you invent a new way to etch that code onto a
chip, that etching process would be patentable. We need to get code out of the
patent system as soon as possible, before any more damage is done.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 02:24 PM EDT |
"""
If you invent a new way of binding paper into a book that is
better/cheaper/faster than existing methods, then you should be able to patent
that way of producing the book - but I do not believe that you should be able
extend patent protection to the contents of the book.
"""
I have to stop you there. Why should you be able to patent a
better/cheaper/faster method of binding paper into a book? If you were the first
person to discover this method then your reward is the competitive advantage you
have to sell a higher quality / less expensive product faster than the other
companies. Why should the society go one step further and let you sue someone
else who is able to implement your exact process?
If you say it's so that you can be protected from having your ideas and
design/manufacturing methods copied, thus taking future sales away from you,
then I say that's a cop out excuse.
If you manage to do the better/cheaper/faster binding then all the power to you.
If I want to step in and become your competition, then I guess I better put my
thinking cap on and out-do you. If I succeed then all the power to me!
The part of the patent system that says companies must publish their methods to
get a patent does indeed force them to share the secrets...but that's not
necessarily a good thing. A naive example would be permitting a student to see
the test answers before taking the test. It enables them to pass the test to
compete with other students but it stifles the thinking process.
The part of the patent system that says no other company may make or sell the
product using the now published methods is hugely flawed. If two people come up
with the same product & method independently the patent system says the
first one to "patent" the thing wins, while the other has to pay a
fine. At the start of the system, people could say "well, just over look
that bit because all you have to do is look at the list of patented things so
you don't re-invent them"...but as lousy of an excuse as that was back
then, it's impossible to do today. If you invent anything, even without copying
or stealing, you have to hire a patent lawyer to search through that long list
and hope you didn't re-invent something. If you did, too bad for you because
you'll have to forget the idea or pay them or at least ask for their permission
to continue using your product, even though they didn't even help you design or
manufacture your product.
Even putting all the history and "idea vs method" stuff aside, today's
patent system widens the gap between the rich (who can afford the patent lawyer
and the patent) and the poor (who can't afford it). The end result is always a
few large monopolies who are legally entitled to money without working for it.
[ Reply to This | # ]
|
| |
| Authored by: darkonc on Wednesday, October 20 2004 @ 02:28 PM EDT |
| Patents are meant to encourage the spreading of innovation -- in other words, in
return for publicly documenting some innovation (which would have otherwise
not been made public), the pantentor is granted monopoly rights over that
invention. If that discovery is widely used in spite of it being
patented, rather than because of the patent, then that patent should be
retroactively ruled to have failed the 'obviousness' test.
Patents were
meant for things like 'patent leather shoes' -- a process which, even when shown
the solution, most experts in the field went "How did you do that?", or the
RSA patent (a software patent) where people went "oh, wow! I'd never have
thought of that!).
When you've got 25 implementations of a solution by the
time it makes it out of the patent office, It's not novel and
innovative. ---
Powerful, committed communication. Touching the jewel
within each person and bringing it to life.. [ Reply to This | # ]
|
| |
| Authored by: GaryD on Wednesday, October 20 2004 @ 02:28 PM EDT |
One of the objections to calls to ban software patents is that it is hard to
tell where to draw the line between software and things where patents have
traditionally been accepted.
This may be overly simplistic, but why not
define a software patent as any patent where infringement would necessarily
involve producing a copyrightable work? In other words, if something is
copyrightable, it isn't patentable.
This doesn't stop someone patenting
the hardware part of an invention, it just prevents them from including the
software part of the invention as well.
Comments? :) ---
Gary
Duke [ Reply to This | # ]
|
| |
| Authored by: Fourmyle on Wednesday, October 20 2004 @ 05:05 PM EDT |
| There seems to be a thread in these postings that "well, we need hardware
patents, but software is bad" I'd like to point out that ALL patents have real
problems. We all know the people involved in FOSS , I like to mention another
valuable resource . Don Lancaster , author and all round hardware guru. He wrote
"The Case Against Patents" back in Nov. of 1990 and I've yet to find a major
point in it I can disagree with.
For anybody that wants to look at this here's
Don's Patent Avoidance Library
, at the least check out the first four papers.
[ Reply to This | # ]
|
| |
| Authored by: jkeltner on Wednesday, October 20 2004 @ 05:48 PM EDT |
| I think that Jonathan makes a good point about new technology blurring the line
between software and hardware. In a traditional view, software would be covered
by copyright law, not patent law, just as would other expressive works.
Hardware, on the other hand, would be covered by patent law, which covers
inventions. But as the line becomes blurred between these two areas, how do we
respond?
What is the intersection of patent law and copyright law. Certainly
copyright law is not at odds with the GPL or any other open source license, as
each of the relies on the copyright of the author to impose its restrictions on
people building on that code (namely that they contribute their additions back
to the public). How does this area of law properly intersect with patent law?
And more importantly, how does that intersection change with today's
technology?
The only way to resolve this issue is to fundamentally rething
the way we look at intellectual property laws in our country; and that means
affecting the decisions made in Congress and the people there to make them. To
that end, I would like to recommend that everyone head over and check out iPac a newly-formed political action
committee "dedicated to preserving individual freedom through balanced
intellectual property policy." It is great for this community to discuss these
issues, but to have a real impace, we must take an active role in shaping the
policy that will truly define this space in the years to come. [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, October 20 2004 @ 05:59 PM EDT |
My oh my, PJ has outdone herself this time. This is the clearest thing I have
read on the patent issue. [ Reply to This | # ]
|
| |
| Authored by: ansible on Wednesday, October 20 2004 @ 06:56 PM EDT |
Schwartz seems to think that settling with Kodak (or perhaps we should call it
KodACK) will "settle the issue" and "put it all behind
them". (paraphrasing)
Hah. I give you two points:
A) Kodak now has $92 million USD to spend on researching and acquiring new
patents. Then it will start suing more companies.
B) Sun has just painted a target on their own back. Now all the shysters will
talk to their clients to dig through their patent portfolios, and see if there
is anything they can sue Sun for, knowing that they will settle over even a
bogus patent.
It would have been smarter for them to try harder to get the patent overturned,
in my opinion. But they don't listen to me, they only just fixed the issue with
int vs. Integer.
Oh well. Good luck with all that. People have predicted Sun's demise over and
over again. And here I'm doing it too now. I'm probably wrong, but I don't
think Sun's future will be very pretty.[ Reply to This | # ]
|
| |
| Authored by: phirephly on Wednesday, October 20 2004 @ 07:34 PM EDT |
I really don't see why Sun would care about FOSS.
They already got their "Java" Desktop from us. I doubt even the Patent
Cold War could force the world to purge FOSS code. Perhaps their bet is that
FOSS code would be placed in the public domain subsequent to a ruling to
dismantle the FOSS community. Having the code, and being a big enough fish to
play the Patent Game, they could continue with their "Java" Desktop,
and let their (true) Linux competitors, Novell, Red Hat, Mandrake, etc be
obliterated one at a time through MS/Sun proxies. What would it take? 6? Twelve
law suits before no one sees value in being a distro? They've already expressed
interest in acquiring their own group of volunteer coders... Maybe it wouldn't
be incredibly easy to round up enough former Linux people; they would probably
bet on a few programmers that want to stay up to their elbows in the code
they've labored for years. Sun uses GPL code, but if I were them, I wouldn't
expect to see it taken away with any ruling against the FOSS. If anything, they
would get more rights to it, I would think. I bet if you checked now, SCO still
distributes nmap in their software. Not that they have more rights to it, but
that the nmap people probably don't have the means to go after SCO. Who knows.
I'm probably completely off base, but I don't see where it would make a bit of
difference to Sun if FOSS was disbanded. I think the assumption is there that
even if FOSS was destroyed and that open source was outlawed, they'd still get
free code. We write code for the beauty, because we like to, right? sigh. [ Reply to This | # ]
|
| |
| Authored by: blacklight on Wednesday, October 20 2004 @ 09:57 PM EDT |
The concept of a patent applied to a monkey wrench that can be used in tight
corners is legitimate to the extent that no one else independently came up with
the same or similar design. The concept of patents applied to software
algorithms is a huge, reckless stretch. This is because the design of the
innovative monkey wrench can be readily copuied by a larger competitor, whereas
reverse engineering a complex algorithm that has not been disclosed is an
altogether more complex and murkier challenge.
One question that just came to my mind is: had the German Enigma machine been
invented by an American, would that machine have been patentable? On one hand,
one could argue that the Enigma is a device and devices are patentable. On the
other hand, the Enigma machine could be considered a device whose sole reason
for existence is the implementation of a fiendishly complex encryption and
decryption algorithm, and algorithms are not patentable. One historical note:
even though Polish intelligence had secured an Enigma machine, the Enigma
machine did not yield its secrets without a huge, sustained effort by British
intelligence.[ Reply to This | # ]
|
| |
| Authored by: ikocher on Thursday, October 21 2004 @ 12:26 AM EDT |
There has been some talk about since when the patent office started accepting
software patents... The mess started when they accepted extremely broad ideas
of how systems might work. Not what the patent _really_ tries to patent.
I wonder if the mess can be stopped. What if the patent office starts to ask
for real implementations, something that the company patenting software must
show, a product, _source_ that works, an then _that_ functionallity is patented,
not more, not less.
Lawyers can even win on this, if a company wants to block competition, they will
have to fill many patents, on all possible functions. at $10k..$20k per
patents... wow!
Also, if a company comes up with a better idea than the one patented, it can
also be patented, because it will be a working example of a better idea that the
previous one.
Just like the original paper clip, and a patent that made an upgrade to it.
The problem I see is basically that the patent office doesn't ask anymore for an
implentation of the patented idea. On a hardware thing, I know it takes time
and money to get a product out of a patent. Well, it doesn't take a rocket
scientist to figure out the one can apply for patents on every possible idea out
there, because there is no need to show that it works, no source needed. :(
In software, if an idea is created there is _source_ because the creator must
have tried it before saying he has an idea, so why not tell the patent office to
accept only applications with _source_, real source, not pseudocode. The code
will be copyrighted also, so others cannot even just copy the code
"easily".
I know, there are ideas, new ideas that can be created, but are so big to
implement that the source will be extremely large, but then... what are they
trying to pantent? everything is this planet?
A patented software with a real implementation will stop patents that try to
patent life in another planet saying they are a data processing algorithm.
I am _not_ a pro-patent guy, and I think the world will be better without
patents anyway, but the US has a mess and I think a bit like Jonathan Schwartz
from Sun, you just cannot take that out from night to day. Something in between
is needed.
Take a look of m$ alleged patent on sender-id. If they file it with real
source, the patent office and everybody else will have a __clear__ idea of how
broad it is, not what m$ wants to say it is. I think corps will hate this idea,
but maybe even they can also think about it as an option.
Ivan
[ Reply to This | # ]
|
| |
| Authored by: Wesley_Parish on Thursday, October 21 2004 @ 06:58 AM EDT |
There's this
great
irony: spend time with Sun's microprocessor team, and
guess what - they're all writing software. Spend time with
our data
center switching team - they're all writing
software, too. Viewed
simplistically, computing hardware
is software burned into and onto
physical
things.
Let's carry on the
question of "software" and
what it is, just a little longer, shall
we?
When I was somewhat younger, I learned the basics of
something called Computerized Numerical Control, or CNC
for short. It's
a way of controlling machines through
entering control codes giving XYZ
locations for cutting
heads, hot wires, etc, to cut and shape materials.
For
what it's worth, it's also used to cut clothing material,
not
just the steel and plastics and suchlike.
CNC can be used as
a simple "conversation" with
the machine, entered interactively, it
can be bundled up
and used batchlike, or it can be used with pre-rolled
and
custom routines with all the complexity of more
"normal"
software.
Most people writing CNC aren't aware that they're
"writing software", "programming" in the
same sense that
they
would
regard software such as Operating Systems or desktop
applications. But it's "software" all right. Does
a routine that
cuts a perfect 90 degree corner with a
slight curve on the corner's point
itself blunting it,
then deserve a patent to protect the originator
thereof?
- Most competent CNC tutors will teach how
to do
the maths to get precisely that effect;
- it comes a
standard with practically every manufacturer's milling
machines' and
lathes' postprocessors;
- and if
anyone ever did get such a patent,
it'd bankrupt the
CNC/CAM manufacturers and kill off the
industry.
Then we get to the interesting thing
that precisely the
same methods of outlining 3D paths in XYZ notation are
the
same methods used in making printed circuit boards and
integrated circuits. (Ditto for using machines for
"populating" said
circuit boards with transistors,
resistors, capacitors, etc...) The
difference is of scale
and the
specific materials used. The concepts
are the same -
anyone who's used a CAD (Computer-Aided Drafting) package
can be retrained to use a CAM (Computer-Aided
Manufacturing) package
without too much trouble - the
major difference is in the concept of
tool paths,
which
don't exist in CAD.
At which
point does this celebrated necessity for
Patent protection kick
in? So, where do you
draw the line on patents? Firmware?
FPGA's? Silicon?
Systems?
I do believe, Jonathan
Schwartz, you've proved our point
on the absurdity of software patents.
If tool paths are a
trivial detail of postprocessing
CNC-to-VHDL(-and-Verilog-and-SystemC-etc), and every
patentable software
development can be boiled down to a
tool path, which can be
trivially reproduced by any
half-way competent CNC programmer
(and CNC's
merely the
computerized version of Numerical
Control, with the which
Boeing for example, built the celebrated B-17
Flying
Fortress, and Packard built the equally celebrated Packard
Merlin)
then since the basics are so
publically
available and so equally basically reproducible, why are
they considered
worthy of any patent protection? ---
finagement: The Vampire's
veins and Pacific torturers stretching back through his own season. Well,
cutting like a child on one of these states of view, I duck [ Reply to This | # ]
|
|
|
|
|
