Implications of copyright
Greg Hulands
the copyright and patents of Apple? I ask this question because I saw a
theme for Enlightenment for Linux and it looked identical to OS8. Because it
is open-source does it stop us from getting lawsuits slapped against us. I
think we should be immuned from any law suits because we know that companies
copy open-source code and do not give the credit where it is due.
These are my opinions.
Greg
Ronald C.F. Antony
> while an option might be nice I would prefer to see the effort going
> to finishing the remaining work needed on GNUstep (with the NeXT UI)
> first.
>
> besides apple tends to sue whenever someone sneezes so who knows what
> they would do...
>
> I suppose it depends on how much GNUstep `annoys' apple.
I can only agree here. The Mac interface is ok given it's age, but the
graphics are somewhat dated. But matters of taste aside, there is little
poing in potentially provoking Apple, they have shown to be lawsuit
happy in the past, and if we annoy them with the GUI, then they might
start to get paranoid and try some tricks like suing people for using
API that's not part of the OpenStep spec, or even for using the NeXT
look, which by virtue of the merger in theory also belongs to Apple.
Just since the NeXT look doesn't really compete with any existing
Apple product, and since there's a precedent of NeXT not enforcing
this, as they were shifting their focus to WinNT, chances that we see
trouble on that front are small.
We have this saying in German: You give the little finger to the
devil, and he takes the entire hand. So if we have this status of
mutual tolerance between GNUStep and Apple, then there is no point
in playing devil and going after Apple's pride...
If you really want Mac look&feel and the OpenStep/Cocoa API, then
go and buy OSX, it's not going to be that expensive...
Ronald
==============================================================================
"The reasonable man adapts himself to the world; the unreasonable one persists
in trying to adapt the world to himself. Therefore all progress depends on the
unreasonable man." G.B. Shaw | rc...@cubiculum.com | NeXT-mail welcome
Adam Fedor
You can sue anybody for anything (and in the US, people frequently do
this). However, regarding the issue of look&feel copyright infringement,
this issue has already been decided in at least one court case (in our
favor). The patent issue is different, however, and we have to be very
careful in implementing the behavior of certain parts of our system (The
IM, for instance).
--
Adam Fedor | Those who can't do, simulate
Digital Optics Co. |
fe...@doc.com (MIME) |------------------------------
fe...@gnu.org (GNUstep!)
Ronald C.F. Antony
> this). However, regarding the issue of look&feel copyright infringement,
> this issue has already been decided in at least one court case (in our
> favor).
Not quite. Look&feel in such wide guidelines such as when Apple tried to
sue MS for Windows looking too much like MacOS is one thing, but doing a
complete clone of the look, i.e. same look of all widgets, etc. that's another
issue, and I don't think it has been decided yet. Further, even it it's
decided, who'd be able to pay for it? While it might be a bad publicity
stint for Apple, we might have to give in merely for lack of funds for a
high profile trial. I don't think Apple will defend the NeXT look heavily,
particularly given how close Win* looks (although there are rumors that
MS is paying a license fee for that), but there's no need to copy the
MacOS look given that it's know how protective Apple is about it.
> The patent issue is different, however, and we have to be very
> careful in implementing the behavior of certain parts of our system (The
> IM, for instance).
Yup.
Pascal J. Bourguignon
Date: Fri, 1 Oct 1999 11:32:33 -0400
X-UIDL: 7b8686074f13bf723cc3e8f4e0f7b65d
> You can sue anybody for anything (and in the US, people frequently do
> this). However, regarding the issue of look&feel copyright infringement,
> this issue has already been decided in at least one court case (in our
> favor).
Not quite. Look&feel in such wide guidelines such as when Apple tried to
sue MS for Windows looking too much like MacOS is one thing, but doing a
complete clone of the look, i.e. same look of all widgets, etc. that's another
issue, and I don't think it has been decided yet. Further, even it it's
decided, who'd be able to pay for it? While it might be a bad publicity
What I wonder is who they'd sue?
Ok, actually there's a clear target: the GNU organization, since the
whole development process is done under its "copyright" umbrela. But
the development, the sources and the installation (one day) are
disseminated all around the world. Nobody knows who has it and what is
done with it (bare for the participants of these mail-list). Would
they sue each and every developer or user of GNUstep? I've made a
installation of GNUstep for a friend that's not on Internet and who's
enhancing it and installing it on all the computers of his
company. How Apple Computer Inc would know how about him and sue him?
May be we should try to improve the resistence of our way of
programming, by removing such a highlighted target as the gnu
organization or a more distributed development and attribution (well
no, don't do that until it realy proves that these big company really
sue for such silly look-and-feel copyright or algorithm patent
affairs).
> The patent issue is different, however, and we have to be very
> careful in implementing the behavior of certain parts of our system (The
> IM, for instance).
Legaly, it may be different. Technically, (in the source), I don't see
that it's different, and it could probably be handled the same way
(ignoring the whole affair). Note that for the GIF patent, it has not
made a difference, any web browser is still able to display it, and
any graphic application is still able to generate it.
You can step on an ant, but you cannot step on the whole colony.
__Pascal Bourguignon__
Kai Henningsen
> Many people do not remember the discussions between Apple Computer
> and the late Gary Kildall's Digital Research Inc. regarding a work
> called GEM, a windowing environment used on both IBM PC systems as
> well as the Atari ST. The net result of that conversation was that DR
> appeared to limit the capabilities of GEM. You don't see a lot of GEM
> out there on any platform today.
Somehow, I think this has much more to do with M$ marketing
than with Apple lawyers.
Incidentally, people seem to be forgetting that Apple is currently
trying to make friends with the Open Source community. It would
be a *really* bad PR move for them to try a look&feel lawsuit right
now.
Furthermore, remember the Apple boycott by the FSF *because of*
their earlier l&f lawsuits? I don't remember the details, but I seem
to remember the boycott was ended because Apple promised not
to do this any more. (Note I don't claim the boycott was the cause
of Apple's change of mind. I assume it was mostly that the tactic
proved to hurt more than it helped. Especially given that Apple
stole^Wborrowed the important ideas from Xerox in the first place.)
Regards - Kai Henningsen
--
http://www.cats.ms
Spuentrup CTI Fon: +49 251 322311 0
Windbreede 12 Fax: +49 251 322311 99
D-48157 Münster Mob: +49 161 3223111
Germany GSM: +49 171 7755060
Richard Frith-Macdonald
> On 4 Oct 99, at 0:20, Michael Brian Bentley wrote:
>
> > Many people do not remember the discussions between Apple Computer
> > and the late Gary Kildall's Digital Research Inc. regarding a work
> > called GEM, a windowing environment used on both IBM PC systems as
> > well as the Atari ST. The net result of that conversation was that DR
> > appeared to limit the capabilities of GEM. You don't see a lot of GEM
> > out there on any platform today.
>
> Somehow, I think this has much more to do with M$ marketing
> than with Apple lawyers.
>
> Incidentally, people seem to be forgetting that Apple is currently
> trying to make friends with the Open Source community. It would
> be a *really* bad PR move for them to try a look&feel lawsuit right
> now.
I wonder if there might be any mileage in seeing if Apple would be prepared
to explicitly say that GNUstep has their permission to duplicate the NeXTstep
look and feel? Perhaps for a nominal one-off fee we could license it?
If they really DO want to look good in the open-source community, they might
even let us license some of their patents (I'm thinking of EOF/web-objects
stuff here).
H.-R. Oberhage
: I wonder if there might be any mileage in seeing if Apple would be prepared
: to explicitly say that GNUstep has their permission to duplicate the
: NeXTstep look and feel? Perhaps for a nominal one-off fee we could license
: it? If they really DO want to look good in the open-source community, they
: might even let us license some of their patents (I'm thinking of
: EOF/web-objects stuff here).
Well, that sounds like a really good idea to me. Do it as Phil Katz
formerly did in the (SEA) ARC battle. Grab hold of a legal construction
(he, then, "invented" the ZIP format himself, copyrighted it and the
filename extension, all just to put it into the public domain; here the
look&feel would have to be acquired by a one-time fee)
and make the look&feel either PD or GPL.
I'm sure Apple doesn't give much about NeXT's look&feel, as you can witness.
If they won't "sell" it, it will only show (again), how they really feel
about open-source; but I'm not that pessimistic, here.
With their 'patents' (EOF/WOF etc), it's certainly "licensing only",
though!
Again, great idea I think, greetings,
Ruediger Oberhage
--
H.-R. Oberhage
Mail: Univ.-GH Essen E-Mail: phy...@sp2.power.Uni-Essen.DE
Fachbereich 7 (Physik) rued...@Theo-Phys.Uni-Essen.DE
S05 V07 E88
Universitaetsstrasse 5 Phone: (+49) 201 / 183-2493
D-45117 Essen, Germany FAX: (+49) 201 / 183-2120
Helge Hess
>
> With their 'patents' (EOF/WOF etc), it's certainly "licensing only",
> though!
How do you mean this ? In which way can a patent be licensing only ?
Helge
--
MDlink online service center GmbH
http://www.mdlink.de/
H.-R. Oberhage
: "H.-R. Oberhage" wrote:
: >
: > With their 'patents' (EOF/WOF etc), it's certainly "licensing only",
: > though!
:
: How do you mean this ? In which way can a patent be licensing only ?
Well, what I mean is this: The look&feel itself isn't patentable,
at least not in Germany (as you probably know :-)), just like algorithms
aren't (here), only the "product"/"tool" is. So once you come to an
agreement about patentable issues of a(nother) look&feel, in order to
build "your" look&feel, you're then free to make yours PD or GPL or whatever.
With patented products, like WebObjects or EOF etc., chances are slim
that you can do anything else, but "use" that license in your product.
You can not freely put a "public license" on patent-dependant things
or where others' licenses are involved, as long as you don't have the
consent of all those involved.
So what I wanted to express previously is, that I see a chance in
getting "NeXT's" look&feel into PD or under GPL, but see no chance
in getting (e.g.) GNUstep's EOF/WOF's implementation truely open-source
(at all and/or additionally under GPL), with all building blocks
uncovered. This I shortened, maybe misunderstandable, to "licensing
only".
Greetings,
Ruediger
Nicola Pero
> to explicitly say that GNUstep has their permission to duplicate the NeXTstep
> look and feel? Perhaps for a nominal one-off fee we could license it?
which looks 'similar' to the NeXTstep. We don't need to ask for that.
Of course, they can think differently and sue us.
But, if we ask for the permission, we are already accepting
the fact that we don't have this permission -- while we have it.
And, we are *not* cloning NeXTstep. We are writing an alternative,
similar but hopefully better, conforming to the OPENstep standards.
The look&feel is similar, for now; we'll add themes support one day
and then all the look&feel question will become marginal.
I think that Apple will not sue us, mainly because if they do,
that's probably the worst move, from a marketing point of view,
they can do. Suing the customers is not a brilliant idea.
And, even if they win such a trial against the FSF, they would loose
so many customers that I don't think they would be very happy
about it. Times are changed -- a trial against the FSF/GNU Project
would be very very very bad advertising today.
We are trying to make things extremely portable between MacOS-X and
GNUstep. That's only good for them, because when we have apps, they'll be
able to port our free apps to their environment.
And they need apps -- and everyone knows that free software communities
produce a huge amount of apps.
So, I would be glad if Apple encouraged our project,
and officially said it likes that we recreate something similar
to the NeXT.
I don't think we should ask for it, and we should *not* pay
a single penny.
If they sue the FSF and win, what will happen is that we will have to
(slightly) change the look&feel, and we will start to actively break
compatibility with MacOS-X.
They would obtain nothing, and loose much.
Nobody buys NeXTstep today, and nobody will buy it after they have
won the trial. But MacOS-X will have less apps, an ugly advertising,
no compatibility with whatever system...
I think they will not do that.
Ronald C.F. Antony
lookalike GNUStep. AFAIK the UI Guidelines were supposed to be part
of the OpenStep standard, and SUN's version copied the look, too.
In addition considering the marketing implications for Apple and
the fact that NeXTSTEP/OPENSTEP-MACH are no longer products they
actively market, I think we're pretty safe here. (Not even to talk about
the somewhat dubious merit of all look&feel lawsuits.)
However, the discussion started around the question if we should/could
copy the MacOS/MacOSX look, and I think, besides not liking it as much
as the NeXT look, that's an unnecessary provocation. We don't need the
Mac look, and we should, if we expect this to become a give and take
two-sided exchange, show some goodwill, too.
Particularly if we get in the murky areas of cloning EOF and WO, we
should not attract too much negative attention to us by copying something
that represents the Mac *marketing image* to *consumers*, which is what
Apple cares most. The worst thing that could happen, from Apple's point
of view is not that there is a free EOF version around running on Linux,
but that someone accidentally takes a GNUStep system for a MacOS X system.
If we stay away from that, I think we can count on continued goodwill
from Apple. They profit by establishing their APIs as a standard, they
gain apps from the GNUStep effort and get brain share, yet we're not
eating away at their core consumer/iMac markets. Everyone is happy.
However if we start messing up Apple's brand identity, by copying the
look, y'all can bet we'll draw Steve's wrath, and as things go with
legal matters: it often doesn't matter who's right, but who has the
deeper pockets and can sue the other guy out of business.
And I don't think anyone here can compete with Apple's war chest.
Regards,
Tiago Ribeiro
>We are trying to make things extremely portable between MacOS-X and
>GNUstep. That's only good for them, because when we have apps, they'll be
>able to port our free apps to their environment.
>And they need apps -- and everyone knows that free software communities
>produce a huge amount of apps.
You got the point.
Apple is not only aware of what are we doing, but also they (at least,
the technical guys) are encouraging us to keep doing it. On a
presentation at WWDC, Wilfredo Sanchez from Apple (OSX Core OS) told
about the contributions that could be done by the developers outside of
Apple for their Darwin operating system. One of the suggested tasks shown
on the slide show was 'to port GNUstep for Darwin'.
Now, regarding the lawyers... well, who knows what they think 'bout us...
;-)
Best regards,
-------------------------------------------------------------------
Tiago Ribeiro Faculty of Architecture and Urban Planning
tia...@uol.com.br University of Sao Paulo - FAU/USP
Voice: +55 11 8503512 Sao Paulo, SP
Fax: +55 11 39834578 Brazil
-------------------------------------------------------------------
Tiago Ribeiro
>However, the discussion started around the question if we should/could
>copy the MacOS/MacOSX look, and I think, besides not liking it as much
>as the NeXT look, that's an unnecessary provocation.
Agreed.
Gregory John Casamento
---
Gregory John Casamento OPENSTEP/Objective-C Developer
"Two things are infinite the universe and human
stupidity, and I'm not sure about the universe"
-- Albert Einstein
On 4 Oct 1999 15:58:47 GMT H.-R. Oberhage wrote:
>Helge Hess (helge...@mdlink.de) wrote:
>: "H.-R. Oberhage" wrote:
>: >
>: > With their 'patents' (EOF/WOF etc), it's certainly "licensing only",
>: > though!
>:
>: How do you mean this ? In which way can a patent be licensing only ?
>
>Well, what I mean is this: The look&feel itself isn't patentable,
>at least not in Germany (as you probably know :-)), just like algorithms
>aren't (here), only the "product"/"tool" is. So once you come to an
>agreement about patentable issues of a(nother) look&feel, in order to
>build "your" look&feel, you're then free to make yours PD or GPL or whatever.
There is nothing wrong with imitating the look and
feel of NeXTSTEP in GNUstep. The whole look and feel
argument was settled about 12 years ago when
Apple attempted to sue MS and lost. Besides we
already have an environment which looks very much
like NeXTSTEP called WindowMaker and Apple has
yet to do anything about it.
In addition MS Windows 95 is a blatent copy
of the look and feel of NS.
>With patented products, like WebObjects or EOF etc., chances are slim
>that you can do anything else, but "use" that license in your product.
WOF is not patented in any form or fashion. I have
been searching the patent databases at www.uspto.gov
and www.patents.ibm.com for one and have not
been able to find it.
Portions of EOF *are* patented. Specifically the
patent addresses "Dynamic Object Communication
Protocol" (US5898871). It is possible, however,
to get around the patent since it's claims are
very specific about what the algorithm does to
achieve this. Patents are applicable only
to a given process of doing a certain thing,
not to ALL possible processes for doing that
thing. All we must do is find another way to
achieve this and the patent is not infringed.
>You can not freely put a "public license" on patent-dependant things
>or where others' licenses are involved, as long as you don't have the
>consent of all those involved.
>
>So what I wanted to express previously is, that I see a chance in
>getting "NeXT's" look&feel into PD or under GPL, but see no chance
>in getting (e.g.) GNUstep's EOF/WOF's implementation truely open-source
>(at all and/or additionally under GPL), with all building blocks
>uncovered. This I shortened, maybe misunderstandable, to "licensing
>only".
Check out gnustep-db and gnustep-web. They
are open-source implementations of EOF and WO
respectively.
>Greetings,
> Ruediger
>
>
Later,
Gregory Casamento
--== Sent via Deja.com http://www.deja.com/ ==--
Share what you know. Learn what you don't.
Ronald C.F. Antony
> feel of NeXTSTEP in GNUstep. The whole look and feel
> argument was settled about 12 years ago when
> Apple attempted to sue MS and lost.
AFAIK this is a common misconception. That law suit was about things
like "uses windows, a mouse, has a desktop metaphor, etc."
Not about "has a close button with an X that's in the upper right corner
and that measures N pixel's square, is light gray pseudo 3D with a light
coming from the upper left corner."
Graphic design is copyrightable, otherwise you could rip of every poster,
etc. GUI elements are graphic design. Thus it's one thing to do something
similar (e.g. Win vs. MacOS) and to rip something off pixel by pixel,
(e.g. GNUStep vs. NeXTSTEP).
For all we know, Apple could even protect key parts of their GUI
graphic design as a trademark, sort of like the Nike swoosh, and then
we'd not be infringing on the copyright, but on trademark rights.
So no, I don't think that issue ever was settled 12 years ago. However,
I do believe that the specific look was supposed to be part of the
OpenStep spec, and I also don't think that Apple would care about the
look of an essentially dead product. Further more, if one doesn't
actively enforce one's IP rights, one looses them.
Since GNUStep has had a rather high profile, and so did Windowmaker,
it is fair to assume that Apple/NeXT by not having taken action probably
lost their chance to enforce their rights to this particular look.
However, that does not make it automatically allowable to copy the
MacOS GUI pixel-by-pixel. These things were never at stake in the
look and feel suits, because nobody had the audacity to ever try that
as commercial venture.
> Besides we
> already have an environment which looks very much
> like NeXTSTEP called WindowMaker and Apple has
> yet to do anything about it.
See above. Again, this debate ensued not about whether or not we should
change the current NeXTSTEP rip-off look (we shouldn't), but about if
we should make a MacOS look clone, which I consider a useless provocation.
H.-R. Oberhage
: [...]
:
: There is nothing wrong with imitating the look and
: feel of NeXTSTEP in GNUstep. The whole look and feel
: argument was settled about 12 years ago when
I'm not quite sure, if there is (legally) nothing wrong
with the imitation, see also other replies here, but I
do concede, that it would be difficult and probably
devastating to Apple to sue, here.
I still think, consensus would be the best, and tend to
have some agreement with Apple (for the NeXT look, not the
MacOS X one) to be safe, and would even consider paying a
moderate nominal fee for that, but if the focus is on "don't
do it, otherwise you risk losing what you already had", it's
ok with me, too. It would just be fine to be on the safe
side, that's all.
: [...]
: WOF is not patented in any form or fashion. I have
: been searching the patent databases at www.uspto.gov
: and www.patents.ibm.com for one and have not
: been able to find it.
:
: Portions of EOF *are* patented. Specifically the
: patent addresses "Dynamic Object Communication
: Protocol" (US5898871). It is possible, however,
: to get around the patent since it's claims are
: very specific about what the algorithm does to
: achieve this. Patents are applicable only
: to a given process of doing a certain thing,
: not to ALL possible processes for doing that
: thing. All we must do is find another way to
: achieve this and the patent is not infringed.
Thank you very much for investigating here. This might
come in handy at a later time. Me, I didn't know about
WOF not being patenten. With late NeXT's and then Apple's
focus on WebObjects, I would've thought they had taken
some/more protective measures, here.
: [...]
: Check out gnustep-db and gnustep-web. They
: are open-source implementations of EOF and WO
: respectively.
I didn't (much) until now, and it's rather clear that
a clean-room implementation of open standards is hard
to sue; from the original posting I was under the
impression, that "we" (=GNUstep community) don't have
such a thing as WebObjects and that is why there should
be some licensing with Apple, which would/could be ok
as an intermediate solution. We could still always make
a clean-room implementation of it at a later time and
re-evaluate the agreement with Apple on this, as long
as the implementation doesn't violate copyrights (as with
using code from the "original" implemenation, as clean-room
specifies).
But as said, I'm not into WebObjects, thus others should
have their say, here.
Thanks again for your work and trouble,
greetings,
Ruediger
PS: To many paragraphs starting with "I", bad style, but
don't mind :-).
Helge Hess
>
> >With patented products, like WebObjects or EOF etc., chances are slim
> >that you can do anything else, but "use" that license in your product.
>
> been searching the patent databases at www.uspto.gov
> and www.patents.ibm.com for one and have not
> been able to find it.
I'm not sure how patents work, does this mean that Apple also can't
*make* new patents on this topic (because there already are WO clones,
eg GNUstep-Web) ?
If not, wouldn't it be wise to make a patent on relevant stuff ASAP, to
guard against later commercial ones ?
(Is there are patent-howto file ?)
> Portions of EOF *are* patented. Specifically the
> patent addresses "Dynamic Object Communication
> Protocol" (US5898871). It is possible, however,
> to get around the patent since it's claims are
> very specific about what the algorithm does to
> achieve this.
Are patents valid in the whole world or does this require extra efford ?
Eg you write it is patent US589..., so is this a US only patent ?
Is the 'Dynamic Object Communication Protocol' key value coding ? If so,
it would probably also apply to WO.
> Check out gnustep-db and gnustep-web. They
> are open-source implementations of EOF and WO
> respectively.
That these exist doesn't mean in turn that they are legally 100% correct
and therefore it's a bit difficult to base development on them without
knowing about possible legal side effects.
I still wonder how Apple would behave if a free EOF and WO clone in the
same quality as the original would be made available. Currently they
probably don't care that much since GDL and gstep-web are no competitors
right now.
Greetings
Kai Henningsen
> look of an essentially dead product. Further more, if one doesn't
> actively enforce one's IP rights, one looses them.
This is false in general. It is, AFAIK, only true for trademarks (stuff
like using "to xerox" fro general photocopying).
IANAL, but this comes up far too often.
Ronald C.F. Antony
> : been searching the patent databases at www.uspto.gov
> : and www.patents.ibm.com for one and have not
> : been able to find it.
> : Portions of EOF *are* patented. Specifically the
> : patent addresses "Dynamic Object Communication
> : Protocol" (US5898871). It is possible, however,
> : to get around the patent since it's claims are
> : very specific about what the algorithm does to
> : to a given process of doing a certain thing,
> : not to ALL possible processes for doing that
> : thing. All we must do is find another way to
> : achieve this and the patent is not infringed.
>
> Thank you very much for investigating here. This might
> come in handy at a later time. Me, I didn't know about
> WOF not being patenten. With late NeXT's and then Apple's
> focus on WebObjects, I would've thought they had taken
> some/more protective measures, here.
AFAIK patents are only published once they are granted, however
they are applicable from whence they were applied for. So we don't
know for how many patents NeXT/Apple applied for in terms of WO,
since these patents may not have been granted yet. Sometimes patents
can take years before they get through the system. It would be
worth investigating if e.g. anywhere in the WO documentation,
data sheet, packaging, CD-ROM, etc. it mentions anything about
"patent(s) pending", which is the term one can use to mark the
turf after a patent has been applied for but before it has been
granted. Other things to look for: there are some companies that
have patents that seem to cover what WO is doing, and so Apple
may not have patents of their own, but may use some other company's
patents under license, at which point that still wouldn't buy us
a free ticket.
(Also, a clean room approach is no protection against a patent infringement
suit, only against a copyright suit...)
Ronald C.F. Antony
> *make* new patents on this topic (because there already are WO clones,
> eg GNUstep-Web) ?
It doesn't matter if there are clones or not. Once a process is accessible
to the public, it's no longer patentable. Even if you invent process X,
and you tell it e.g. in class room (a public forum), then you can no longer
patent it. (In effect you still can because chances are slim the patent
office would know about such an incident, but if someone in that class room
wanted to, they could challenge the patent, and derail it).
Only things that are trade secrets inaccessible to the public can be
patented. The entire reason for a patent (from latin "patere" "to open up")
is that an inventor gets rewarded for opening his discovery or technique
to the public rather than taking it with him to the grave and as a reward
gets a temporary monopoly on the use thereof. It is a tradeoff that society
made, to avoid that relevant knowhow disappears with the inventor's death.
(If Stradivari had patented how to make violins, his originals wouldn't
fetch such astronomical prices today, cause we could make similar ones
by the thousands...)
The idea that intellectual property right is something like a natural
right, etc. is rather modern, and cost the industry hundreds of millions
in lobbying money to spread. Naturally knowledge is free, because it
can be arbitrarily duplicated w/o deterioration.
> If not, wouldn't it be wise to make a patent on relevant stuff ASAP, to
> guard against later commercial ones ?
> (Is there are patent-howto file ?)
Patenting things is a rather tedious process, plus you need to do it for
each country separately. At least you'd need a US, Japanese and EU patent.
That costs a shitload of money and time.
Again, anything that's already out, can't be patented. Only new additions,
new algorithms replacing existing ones, etc. would be candidates.
It would be good, though, for the FSF to start creating a patent portfolio,
if only to have something to bargain with when it comes to cross-licensing
agreements. If the FSF can block industry, then industry has to pay or
stop blocking us. (e.g. compress, GIF, etc. come to mind)
The FSF could license all their patents for free as long as they are
used under the LGPL or GPL, and otherwise use them as bargaining chips
to erode the patent walls elsewhere.
> Are patents valid in the whole world or does this require extra efford ?
> Eg you write it is patent US589..., so is this a US only patent ?
Patents are not valid worldwide, they are country specific. That's why
e.g. the RSA crypto algorithm can be used everywhere freely, except in
the USA, where it's patented. Thus OpenSSL is possible with a native
RSA implementation. Within the US, people are forced to use the suboptimal
RSARef implementation, that has all sorts of strings attached to it.
Ronald C.F. Antony
> On 5 Oct 99, at 2:33, Ronald C.F. Antony wrote:
>
> > look of an essentially dead product. Further more, if one doesn't
> > actively enforce one's IP rights, one looses them.
>
> This is false in general. It is, AFAIK, only true for trademarks (stuff
> like using "to xerox" fro general photocopying).
>
> IANAL, but this comes up far too often.
I don't remember the details, but by father having been both a patent
attorney and a patent judge, I seem to remember that a reasonable effort
has to be made to enforce the rights, or else they are void.
The idea here is, that e.g. a company can't hide their patent claims and
wait for another company to make a lot of money with it, just to then
ask for much higher damages.
To take the landmark Polaroid vs. Kodak suit: If Polaroid hadn't instantly
sued Kodak when they came out with their instant film stuff things might
have turned out differently. In any case, negligent enforcements of rights
will have a significant effect on the sort of damages that can be claimed.
Patents are not supposed to support predatory behavior, i.e. waiting for
people to violate them, just that one can sue them later. The point is
to generate revenue through licensing. Thus if someone waits to enforce
the rights despite better knowledge, one would have to assume predatory
behavior.
Anyway, patent law is rather complex, and I don't claim to be an expert,
but I picked up a few things over the years from being exposed to it.
And there are, at least in some countries, certainly rules that negatively
affect a patent holders rights if he is not making a good faith effort
to enforce his patent.
Kai Henningsen
> you wrote:
> > On 5 Oct 99, at 2:33, Ronald C.F. Antony wrote:
> >
> > > look of an essentially dead product. Further more, if one doesn't
> > > actively enforce one's IP rights, one looses them.
> >
> > This is false in general. It is, AFAIK, only true for trademarks (stuff
> > like using "to xerox" fro general photocopying).
> >
> > IANAL, but this comes up far too often.
>
> I don't remember the details, but by father having been both a patent
> attorney and a patent judge, I seem to remember that a reasonable effort
> has to be made to enforce the rights, or else they are void.
Definitely true for trademarks.
Definitely false for copyright.
I have no idea how it works for patents.
> have turned out differently. In any case, negligent enforcements of rights
> will have a significant effect on the sort of damages that can be claimed.
Maybe, with patents. Probably not, with copyright.
Kai Henningsen
> Patents are not valid worldwide, they are country specific. That's why
> e.g. the RSA crypto algorithm can be used everywhere freely, except in
> the USA, where it's patented. Thus OpenSSL is possible with a native
That, and the fact that patents on algorithms are often frowned
upon.
Ronald C.F. Antony
> wouldn't infrange the EOF patent(s)?
If Apple/NeXT didn't apply and receive a European or Italian patent, then
yes. However, it's not a question of just programming it legally, it's
also a question of using it legally. In other words, even if it's
programmed in Italy, we in the US might not be able to use it.
On the other hand, if they don't have a Europan or Italian patent, then
people there can rejoice.
> I am not saying that tomorrow morning I will start coding, but I can start
> thinking about it if I am not violating any patent.
Yup. As I said, patent searches are difficult, because in many countries
the process of getting a patent can take years, and in some of them patents
aren't published before they are granted. (In some countries however the
patents are published once they are applied for, such as to give people a
chance to object based on prior art, etc.)
Also, not all countries offer an online listing of the patents, so it
may involve costly patent searches through agencies or the patent office.
All that just shows how difficult these things are. I could be writing
an original piece of software, and unless I spend large chunks of money
to do a patent search, I might be violating someone's patent w/o knowing.
Worse, I could pay for the patent search, and the other patent may have
been applied for, but not yet granted. So after I put years of work into
software, it might just be to be told to trash it all, and possibly to
pay fines. In other words, software patents are horrible in the way they
stiffle innovation, if you really care about them.
The big companies have pretty much all cross licensing agreements and
don't have to worry much about these things. They have bargaining chips.
Thus patents have been perverted: instead of helping small inventors,
and preventing them from taking their secrets to the grave, they only
help the big companies, who use it as a barrier of entry, who'd have
to tell so many people about their procedures anyway, that things
would leak to the public sooner or later, and who have deep enough
pockets to sue even legitimate owners of patents into oblivion.
(Don't even think you stand a chance defending your patent against
e.g. a company like MS: they can appeal, twist and bend the legal
process for so long, that you are bankrupt before you win, and in the
end, they can use your patent without worries, because you have to fold
or walk away with some minor compensation, which if you're lucky will
just about cover your legal expenses...)
Ronald C.F. Antony
> On 5 Oct 99, at 6:38, Ronald C.F. Antony wrote:
>
> > Patents are not valid worldwide, they are country specific. That's why
> > e.g. the RSA crypto algorithm can be used everywhere freely, except in
> > the USA, where it's patented. Thus OpenSSL is possible with a native
>
> That, and the fact that patents on algorithms are often frowned
> upon.
For the longest time patents on algorithms and mathematical formulas were
plain excluded by the law, as were patents on business ideas. We owe it to
lobbying and backdoors in the law that these ever became patentable.
I think the first construct that allowed software patents, was when a
hitherto purely machanical machine was equipped with an computer controller,
and thus the entire combination, hardware, computer, software was patented.
It was then argued, that if the same machine is patentable if it's purely
mechanical, then it would be unfair not to patent the computer and associated
software that takes over mechanics and improves the process.
So from there software was no longer considered "algorithm or formula" but
something that turns an general purpose computer into special purpose
hardware replacing traditional implements.
The next step was when (I think Citibank) got granted a patent on a business
idea, i.e. the linking of savings and checking accounts, such that a customer
could have the system automatically transfer funds from checking to savings
and vice versa, as soon as there was too much money on the checking account
or the danger of bouncing a check. Offering this service would be classified
as a business idea, however by patenting the software that implements it,
even that barrier was knocked down.
The entire IP law has become very disgusting the moment lobbies managed to
seat the idea in lawmakers' heads, that IP rights are rights, rather than
privileges granted by society, for the general good (like e.g. preventing
knowhow to be taken to the grave, or to protect the extremely risky and
expensive publishing of paper books.) How much so it was not considered a
right, but a privilege, is obvious by the fact that type font design was
never granted copyright status in the US, because way back when, all relevant
type foundaries were German, and there was no incentive for the US to
protect German interests at the expense of the american public.
(Someone tell the US that Taiwan has no interest to protect the US software
industry at the expense of the taiwanese public... Oops, did I just assume
equal yard sticks?...)
Ack. :-(
Gregory John Casamento
On Tue, 5 Oct 1999 11:29:55 Kai Henningsen wrote:
>On 5 Oct 99, at 2:33, Ronald C.F. Antony wrote:
>
>> look of an essentially dead product. Further more, if one doesn't
>> actively enforce one's IP rights, one looses them.
>
>This is false in general. It is, AFAIK, only true for trademarks (stuff
>like using "to xerox" fro general photocopying).
Actually it is true. If a company A patents a
paticular process. Company B then starts using
a process exactly like company A's, but company
A does nothing about it. Then along comes company
C which starts using another process exactly like
company A's. Company A cannot, at that time,
sue company C since it let company B get away with
it. It is called "sitting on ones rights".
Therefore company A has effectively surrendered
it's rights to the patent.
>IANAL, but this comes up far too often.
>
>
>Regards - Kai Henningsen
>
>--
>http://www.cats.ms
>Spuentrup CTI Fon: +49 251 322311 0
>Windbreede 12 Fax: +49 251 322311 99
>D-48157 M|nster Mob: +49 161 3223111
>Germany GSM: +49 171 7755060
>
Later,
---
Gregory John Casamento OPENSTEP/Objective-C Developer
"Two things are infinite the universe and human
stupidity, and I'm not sure about the universe"
-- Albert Einstein
--== Sent via Deja.com http://www.deja.com/ ==--
Giulio Cesare Solaroli
>[...]
> > Are patents valid in the whole world or does this require extra efford ?
> > Eg you write it is patent US589..., so is this a US only patent ?
>
>e.g. the RSA crypto algorithm can be used everywhere freely, except in
>the USA, where it's patented. Thus OpenSSL is possible with a native
>RSARef implementation, that has all sorts of strings attached to it.
Does this mean that if I start a EOF 3 cloning effort here in Italy I
wouldn't infrange the EOF patent(s)?
In this last tread, the EOF patent was cited as US5898871 (I have not check
this one, yet), but there is another patent I have checked and it is
US5873093 (Stephen MacDougall did found it).
I am not saying that tomorrow morning I will start coding, but I can start
thinking about it if I am not violating any patent.
Giulio Cesare Solaroli
Helge Hess
> I for myself would prefer a design right from scratch. Since the basic idea is very trivial (and already implemented many times), i.e. mapping relational rows to objects, I don't think patents can apply.
> The only thing which may be patented (and which belongs to the superior part of EOF) is the way faults are handled, which is a very elegant way of solving the object context problem in OODBMS.
Personally I don't think that faults are that good in this context.
Smart pointers as used in OODBMS can carry much more information, eg
modification state and editing context, which needs to be held elsewhere
otherwise (either in a special superclass or in a keyed info table).
Smart pointers have less problems with RC as well, especially in
reflexive relationships, since the object itself isn't affected by
references pointing to it (no -forgetAllObjects anymore ;-)
Giulio Cesare Solaroli
>Giulio Cesare Solaroli wrote: > At 06:38 10/5/99 -0400, Ronald C.F. Antony
>wrote: > [...] > > I am not saying that tomorrow morning I will start
>coding, but I can start > thinking about it if I am not violating any
>patent. > I'm too interested in an EOF equivalent for GNUstep. While I
>think the GUI level is done well, I don't believe it's a priority to clone
>the control/access layer of EOF2/3.
>I for myself would prefer a design right from scratch. Since the basic
>idea is very trivial (and already implemented many times), i.e. mapping
>relational rows to objects, I don't think patents can apply.
I am sorry, but I have bad news here.
The United States Patent 5,873,093
(http://164.195.100.11/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1
&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='5873093'.WKU.&OS=PN/5873093&RS=PN/
5873093) is all about how to encode the mapping of relational entities
(rows, relationships, fields, etc...) to objects entities (objects, object
references, object fields, etc...) into an external file (the model) and
how to use this file effectively.
Stephen MacDougall has analized it quite extensively and his opinion (as I
get it) was that it was almost impossible to implement an EOF compatible
library encoding the mapping information into an external file without
violating the patent.
One option left was to find other products using the same solution without
licencing the patent in order to be able to consider the patent void.
I have found a couple of them: the first one is ROF (a EOF clone 100%
java), from Watershed. Stephen told me that Watershed was an ex-NeXT shop
and so it was highly propable that they have licenced the patent.
Another one is, TopLink from ObjectPeople. TopLink is quite different from
EOF (as long as two relational-object mapping libraries can be); the
interesting point is that they are using a file to store the same kind of
information that are found inside an EOModeld file (folder).
Now, I am not a lawyer, but what comes out is that the basic idea patented
by NeXT has been used in another product. The implementation is different,
but the idea is just the same.
I don't know if this is enough to call this patent void.
>The only thing which may be patented (and which belongs to the superior
>part of EOF) is the way faults are handled, which is a very elegant way of
>solving the object context problem in OODBMS.
The key-value protocol too seams to be patented ("Dynamic object
communication protocol", United States Patent 5,898,871), but I have not
analized it deeply.
Hope this helps.
Giulio Cesare
Tim Bissell
>The key-value protocol too seams to be patented ("Dynamic object
>communication protocol", United States Patent 5,898,871), but I
I'm no lawyer, but the getX/setX stuff is central to Java beans,
so is Apple gearing up for a fight with Sun? Or is the concept of
"property" vs. "Ivar" sufficiently distinct for them to get away
with it?
Time to send more money to the FSF, I suppose!
Tim
=====
home: t...@ursidae.demon.co.uk
work: tim.b...@dresdnerkb.com
alt: tim_b...@yahoo.co.uk
____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie
Gerrit Van Dyk
>The United States Patent 5,873,093
>(http://164.195.100.11/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=
1
>&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='5873093'.WKU.&OS=PN/5873093&RS=PN
/
>5873093) is all about how to encode the mapping of relational entities
>(rows, relationships, fields, etc...) to objects entities (objects, object
>references, object fields, etc...) into an external file (the model) and
>how to use this file effectively.
I have used a similar concept in my smalltalk days. I just didn't store the
model in an external file, but rather in a SystemTable in the database. I
used the concept for mapping columns in a table to fields in a
comma,delimited ascii file. Now this is the same concept so I don't think
this can hold. I have done this long before EOF came out.
>The key-value protocol too seams to be patented ("Dynamic object
>communication protocol", United States Patent 5,898,871), but I have not
>analized it deeply.
This is a well documented Pattern from Martin Fowler called dynamic object
properties. (At the conceptual level). I have also used this type of
protocol in my early smalltalk days to let tableviews get their data from an
object in a dynamic way. I defined the mapping in a table and the tableview
read this mapping and asked the object for its valueForAttributeNamed:. The
implemented valueForAttributeNamed: method then queried the runtime
environment in smalltalk for the proper method to call.
This I have also done long before EOF's days, so how can this be a patent.
It seems that the problem with patent registering in the software world is
very easy, because nobody in the patent office knows where to look for
similar things.
I really don't think that these patents can be enforced as I am sure that
any type of dynamic system would have used similar types of methods and
protocols. It seems that everytime an idea is "stolen" out of another
environment and modified a bit then "Next/Apple" registers a patent for it.
Sometimes you think you are very clever with certain unique ideas, just to
find out later that somebody else has done the same thing before you. These
are basic "dynamic data driven" programming styles, IMHO I can't see how
this could ever have been patented. We might call it something different but
in essence its the same thing.
The problem though is, that our software were written in the sanction days
in South Africa, so they will not hold much water against a Patent that's
registered in the US.
Regards
Gerrit van Dyk
Robert J. Slover
On Wed, 6 Oct 1999, Giulio Cesare Solaroli wrote:
::Stephen MacDougall has analized it quite extensively and his opinion (as I
::get it) was that it was almost impossible to implement an EOF compatible
::library encoding the mapping information into an external file without
::violating the patent.
So, encode it into a database. Make the database gdbm if need be.
I worked at a company that implemented objects this way. We had
an object database of our own called NRM living on top of Oracle.
It implemented the 'model-model', a generic model which could
encode any other object model. One of the GUI's built on top of
this even had something similar to a web browser (built just for
interactive help) called HyperWeb that instantiated objects from
NRM to build the on-screen content. This was *years* before EOF.
--Robert
Robert J. Slover
On Wed, 6 Oct 1999, Tim Bissell wrote:
::
::>The key-value protocol too seams to be patented ("Dynamic object
::>communication protocol", United States Patent 5,898,871), but I
::>have not analyzed it deeply.
::
::I'm no lawyer, but the getX/setX stuff is central to Java beans,
::so is Apple gearing up for a fight with Sun? Or is the concept of
::"property" vs. "Ivar" sufficiently distinct for them to get away
::with it?
It could be argued that getX/setX is central to SNMP as well.
So what if the object I am communicating with is a hybrid of
firmware and hardware?
--Robert
::
::Time to send more money to the FSF, I suppose!
::
Joshua Marker
> model in an external file, but rather in a SystemTable in the database. I
> used the concept for mapping columns in a table to fields in a
> comma,delimited ascii file. Now this is the same concept so I don't think
> this can hold. I have done this long before EOF came out.
. . .
> This I have also done long before EOF's days, so how can this be a patent.
. . .
> It seems that the problem with patent registering in the software world is
> very easy, because nobody in the patent office knows where to look for
> similar things.
You've hit it exactly. They're several clue-by-fours short of a
house. The model just isn't capable of dealing with IP. Take a look at
some of the patents on elliptical encription; they're patenting basic
properties of geometry. Not even _really_ the process, any more than
'solve this equation'.
Sigh.
--- Joshua * San Francisco: 415 / 505.9406 (37.753083N -122.430501W)
Sorting out my life. . . .in O(n lg n).
Love is like pi: natural, irrational, and very important.
-- Lisa Hoffman
---
Ronald C.F. Antony
> in South Africa, so they will not hold much water against a Patent that's
> registered in the US.
That doesn't matter. Prior art is prior art. Period.
The issue is however to finance a potential patent law suit, if ever it
should be started. As I wrote before, being right, and getting justice
are two different things. If one has to thow the towel for lack of funds,
then the mere threat of a lawsuit works, even if otherwise they couldn't
win it.
Ronald C.F. Antony
> so is Apple gearing up for a fight with Sun? Or is the concept of
> "property" vs. "Ivar" sufficiently distinct for them to get away
> with it?
I think the difference between property and ivar is significant.
If you look at the Apple stuff, then values can be
- ivars
- contents of a dictionary
- dynamic results computed in a method
and on top of that, ivars can be accessed wither with a pre-programmed
accessor method or dynamically by the runtime system.
So this goes considerably further than just having accessor methods for ivars.
Ronald C.F. Antony
> control/access layer of EOF2/3. I for myself would prefer a design right
> from scratch.
That's like saying AppKit cloning isn't a priority. Compatibility is
central and essential to the GNUStep idea. The plan here is not to
become yet another OO toolkit, like there are already dozens, but to
be part of a standard that works over different platforms and implementations,
and that first and foremost includes what Apple does.
Now if Apple at some point starts butchering their toolkits, moves entirely
to Java, or does other nightmarish things, *then* it may be time to split,
because following Apple should not mean degrading what already exists,
but we are not at this junction yet, and we may never be there.
Robert J. Slover
On Wed, 6 Oct 1999, Gerrit Van Dyk wrote:
::
::I really don't think that these patents can be enforced as I am sure that
::any type of dynamic system would have used similar types of methods and
::protocols. It seems that everytime an idea is "stolen" out of another
::environment and modified a bit then "Next/Apple" registers a patent for it.
Sometimes as a business, patenting the obvious is seen as
self-protection. Oracle claims that this is the reason
they patent software concepts. It keeps someone else from
acquiring the patent and suing *you* for infringement.
::Sometimes you think you are very clever with certain unique ideas, just to
::
Giulio Cesare Solaroli
>Giulio Cesare Solaroli wrote:
>
> >The United States Patent 5,873,093
> >(http://164.195.100.11/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=
>1
> >&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1='5873093'.WKU.&OS=PN/5873093&RS=PN
>/
> >5873093) is all about how to encode the mapping of relational entities
> >(rows, relationships, fields, etc...) to objects entities (objects, object
> >references, object fields, etc...) into an external file (the model) and
> >how to use this file effectively.
>
>model in an external file, but rather in a SystemTable in the database. I
>used the concept for mapping columns in a table to fields in a
>comma,delimited ascii file. Now this is the same concept so I don't think
>this can hold. I have done this long before EOF came out.
The patent takes into consideration also relationships, flatten attributes,
flatten relationships and other details of EOF.
I am very pleased to be able to find out more and more evidence against the
patent, but I am not a lawer and I will not be able to decide on my own
whether it applies or not.
> >The key-value protocol too seams to be patented ("Dynamic object
> >communication protocol", United States Patent 5,898,871), but I have not
> >analized it deeply.
>
>This is a well documented Pattern from Martin Fowler called dynamic object
>properties. (At the conceptual level). I have also used this type of
>protocol in my early smalltalk days to let tableviews get their data from an
>object in a dynamic way. I defined the mapping in a table and the tableview
>read this mapping and asked the object for its valueForAttributeNamed:. The
>implemented valueForAttributeNamed: method then queried the runtime
>environment in smalltalk for the proper method to call.
>
>This I have also done long before EOF's days, so how can this be a patent.
I am arguing for the patents. I was very upset when we started thinking
about cloning EOF 3 and we stopped before starting because of these patent
issues.
If somebody on this comunity feels the right to declare these patents void,
I will be more than welcome to collaborate with the EOF cloning project.
I had many time free in the last months, but now I am very busy again, so I
need to wait for another opportunity
>[...]
>The problem though is, that our software were written in the sanction days
>in South Africa, so they will not hold much water against a Patent that's
>registered in the US.
Is someone still beliving that we all are the same? :-[
Regard,
Giulio Cesare Solaroli
Ronald C.F. Antony
> patent, but I am not a lawer and I will not be able to decide on my own
> whether it applies or not.
Chances are, you won't find a lawyer either. There are preciously few
that really understand software, and something special like EOF in
particular. Most will give you the advice, to stay away from a patent
if you have any doubts it may be valid, simply because they have the
opinion of "agree, for the law is costly" ;-)
Also lawyers never give a guarantee, only opinions. So they may tell
you a rubbery sentence like: "I believe the probability that the patent
would hold up in court is relatively small, but consider how much a
potential test of this costs in court..." or something similar.
In the end, you can get by claiming in court that black is white and
white is black, if your lawyer is good.
Essentially, there is only several ways of dealing with this:
a) play the probability, and hoping that Apple only uses the patent as
self-protection, or won't risk the bad publicity generated by suing
the FSF.
b) finding a country where the patent isn't valid, and develop the software
there, and let the users decide if they want to risk infringement by
using the product in a country where the patent is valid.
c) program the stuff under a pseudonym, and put the source into the PD
or under GNU license, w/o assigning the copyright to the FSF, such as
to prevent Apple from having any target to sue.
d) getting Apple's consent.
(If d) fails, the other options remain possible...)
> >The problem though is, that our software were written in the sanction days
> >in South Africa, so they will not hold much water against a Patent that's
> >registered in the US.
>
> Is someone still beliving that we all are the same? :-[
No, but facts are documentable. A sworn statement by witnesses testifying
that a certain technique was used, particularly if the product was publically
accessible (e.g. for sale), cannot be ignored by a court, regardless where it
happened. While equality in front of the law is an illusion, it must also be
said, that in reasonably civilized countries, a certain base line of justice
can be taken for granted, and a blunt disregard for documented prior art is
something I would not expect to happen.
Ronald C.F. Antony
I hate to follow up on my own posts, but a little birdie tweeted a message
into my ear. The gist of it relates to the following paragraphs of mine:
I wrote:
> Essentially, there is only several ways of dealing with this:
> a) play the probability, and hoping that Apple only uses the patent as
> self-protection, or won't risk the bad publicity generated by suing
> the FSF.
> b) finding a country where the patent isn't valid, and develop the software
> there, and let the users decide if they want to risk infringement by
> using the product in a country where the patent is valid.
> c) program the stuff under a pseudonym, and put the source into the PD
> or under GNU license, w/o assigning the copyright to the FSF, such as
> to prevent Apple from having any target to sue.
> d) getting Apple's consent.
I was told that chances of d) are about 0.0000001%, or something in that
order of magnitude, since some people supposedly already tried this.
Thus if one wants to deal with the patent issue directly, one would either
have to sue Apple in an effort to void the patents, or try best a
combination of b) and c). As long as c) is the case, Apple has no target
to sue, and as long as b) is the case, they have no right to sue.
A combination of these prevents e.g. the operators of ftp sites to be
sued (as long as they are located in countries where b) applies),
and it prevents the programmer from being sued for "enticing people
to infringe on Apple's patents".
Ideally, if one had the money, one should simply sue to have the patents
thrown out on the basis or prior art. Unfortunately, I don't know anyone
with deep enough pockets, and pro-bono patent cases are essentially
unheard of.
Option a) is possible, but rather risky. I mean, who'd like to open
themselves up as a target.
In consequnce, at this point, the best approach is to find countries in
which Apple's patents don't apply, and take it from there.
(Which excludes me as a contributor... just in case someone at Apple gets
some false idea here... ;-) )
Anyway, hope that helps. So doing a patent search in various countries
(Canada, EU, etc.) would be worthwhile.
Regards,
Pascal J. Bourguignon
Just an idea to stop all this software patent bullshit:
What if we code a program generator that would generate the whole
combination of algorithms?
Of course, it could never generate all past present and future programs, but
potentially, eventually, all would be generated. This kind of program generator
would contain virtually all the other programs, and therefore would be
prior art for any program than even a human could generate. That would end
this patent mess wouldn't it? Perhaps we should setup a computer to write
piles of CDWROMs of programs like tibetan monks...
Thinking about it, it would start by small program sources first, perhaps
it would generate fastly enough all the algorithms that a human being could.
__Pascal Bourguignon__
--
__Pascal Bourguignon__ http://www.imaginet.fr/~pjb/
mailto:p...@imaginet.fr mailto:6490...@activajoven.tsm.es
+--- C: a bool is a bool ; a pointer is a pointer!----------------+
void PleaseWrite(bool isBetter,void* pointer,int value)
{if((!isBetter)&&(pointer!=NULL)||(value==0)){ print("Right!\n");}}
void DontBeSilly(bool isSilly,void* pointer,int value)
{if((isSilly!=true)&&(pointer)||(!value)){ print("Wrong!!\n"); }}
+------- Hi! I'm a signature virus. Copy me into your sign -------+
Kai Henningsen
> Just an idea to stop all this software patent bullshit:
>
> What if we code a program generator that would generate the whole
> combination of algorithms?
Such as a human mind does, you mean?
> Of course, it could never generate all past present and future programs, but
> potentially, eventually, all would be generated. This kind of program generator
> would contain virtually all the other programs, and therefore would be
> prior art for any program than even a human could generate. That would end
> this patent mess wouldn't it? Perhaps we should setup a computer to write
> piles of CDWROMs of programs like tibetan monks...
Don't think so. It's only prior art once it has been published. You
could wait several million years for a construct like this to publish
the algorithm you need.
> Thinking about it, it would start by small program sources first, perhaps
> it would generate fastly enough all the algorithms that a human being could.
Do you know the story about the chess board and the rice? (Or
was it some other grain? I don't remember.)
Some king wanted to reward some guy for some of these not-really-
payable things.
"Ok", says that guy, "take this chess board. Put one grain into the
first field, two into the next, then four, and so on, always doubling
the amount. Then give me the contents of the last field".
'What a ridiculously small rewad', thinks the king ... until someone
calculates that the result is a lot more than the world rice harvest
and, thus, a lot more than the king _can_ pay.
Producing all possible algorithms is much, much worse.
Just to give an idea of the problem space:
All possible (ASCII-only) program sources of length 1KB are
128^1024 = 2^7168 > 10^2157.
All possible (ASCII-only) program sources of length 10KB are
128^10240 = 2^71680 > 10^21577.
Suppose you process one of those every microsecond. You thus
need about 10^21564 years for processing all of them. The universe
is currently about 10^10 years old, so you only wait 10^21554
times the age of the universe ...
Regards - Kai Henningsen
--
http://www.cats.ms
Spuentrup CTI Fon: +49 251 322311 0
Windbreede 12 Fax: +49 251 322311 99
D-48157 Münster Mob: +49 161 3223111
Philippe C.D. Robert
> Just an idea to stop all this software patent bullshit:
>
> What if we code a program generator that would generate the whole
> combination of algorithms?
>
> potentially, eventually, all would be generated. This kind of program generator
> would contain virtually all the other programs, and therefore would be
> prior art for any program than even a human could generate. That would end
> this patent mess wouldn't it? Perhaps we should setup a computer to write
> piles of CDWROMs of programs like tibetan monks...
No need to do that, since every 'program' (in 'theoretical computer science' slang) can be
encoded into one specific 'Goedel number', so theoretically, every 'computable problem'(is this
the right word in English?) is already written, when you count to infinity, resp. when you find
the right number...;-)
..of course, this is not very exact nor correct, but I couldn't resist...
sweet dreams, Phil
---
Philippe C.D. Robert
Developer - StudioSendai.com