Now, they can say they would have
broken it if if wasn't so scary to try.
Otherwise, they would have to admit it would
have taken them forty-leven trillion centons
just to try out the easy combinations like
"haxorz begone", before moving on to the
non-alphabetic "hard" ones.
lirkbald wrote, "if you don't know what what I
just said means, then you shouldn't be commenting on
the existence of 'important' theorems, should
you?"
If you don't know the answer, just say so, and let
somebody else speak up.
The examples quoted seem to me to answer the
opposite to the question I asked. For example, all
the standard electromagnetic results may be
expressed with quaternions, which do not support
commutativity.
I asked about examples of results for which
commutativity is an essential quality. I don't
doubt there are some, or plenty, but what are
a few of them?
Please do. I was hoping to learn something.
It seems like most extensions from scalars into
multiple dimensions (matrices, quaternions)
shed commutativity like last year's skin.
Can somebody please give an example of an
important and comprehensible theorem that depends
on commutativity? That is, what is a consequence
that would be meaningful to a non-mathematician?
I read the license, and it looks (from a cursory
reading) very good. Some of the language maybe
ought to be slurped into the GPL, in fact,
particularly the severability bits.
An interesting paragraph reads:
12.1 Termination. This License and the rights granted hereunder will terminate:...
(c) automatically without notice if You, at any time during the term of this License, commence an action for patent infringement (including as a cross claim or counterclaim) against Sybase or any Contributor.
So, contribute to a released version and get free
patent licenses from anybody who uses it, or they
have to stop using it. If this were added to the
GPL it would mean more.
I wrote:
A program that works only when a GPLed library is present is derivative of that library.
sholden wrote: Since when did your opinion trump the court's?...
A program is a derivative of a library if a court rules it so (and that may change on appeal...). Copyright law determines what a derivative is, and the courts have the job of interpreting that law.
No court has ruled. Legal scholars disagree.
My remark expressed the opinion the FSF obtained
from its legal experts, to provide a concise
counterpoint to the article. Rosen should have
expressed it himself, and would have done were
he interested in a fair exposition of the issue.
Concealing such a legal risk (i.e. that the court
might agree with the FSF) would harm his clients.
Is he telling his clients one thing and us
something else? He's either cheating them or
lying to us. Which is worse?
Memerot writes: "The GPL only applies at all if the work is a derivative work".
Memerot has understood the topic of the article
(and parrots its claim), but not of the
discussion. Suffice to say that other legal scholars disagree with him (and Rosen)
about what counts as a derivative work, and that
Rosen is disingenuous in pretending otherwise.
If Rosen were correct, then the GPL would in fact
be equivalent to the LGPL. That's a fact, not
under dispute. The debatable question is whether
he is correct. The ethical issue is whether
Rosen is being scrupulous in failing to present
the opposite argument.
Also at issue are Rosen's motives: is he
encouraging people to violate library authors' expressed wishes in an attempt to weaken GPL
protections from libraries?
jmorris42 wrote: Perhaps you missed the original poster's parenthetical comment,
"(Incidentally, it would knock the teeth out of
the GPL.) ".
I wrote that myself.
Copyright gives the author
only limited control. Control of distribution
and of public performance are two of the things it
does grant to the author. The author exercises
that control via licensing. The control is limited
by Fair Use (except where that conflicts with the
DMCA).
The Uniform Commercial Code (UCC) gives the
purchaser extra rights, beyond Fair Use. It does
not grant the right to distribute derived works
or make public performances in violation of the
license extended by the author.
jmorris42 writes further, "the clean flicks
people buy a number of copies, replace the disc in
the package with an edited version and rent them".
Renting counts as public performance, which is
also restricted under copyright. If they're
allowed to make public performances of private
versions without permission, then exploiters
will also be allowed to redistribute private
versions of GPLed programs without permission
(i.e. without sources).
Under the UCC the video stores are free to do
whatever they need to to make the physical disc
they bought do what the seller promised. (That's
what makes shrink-wrap licenses wastepaper.)
Selling a modified copy of it, however,
is not something the UCC allows. Play the disc,
decrypt the disc, hang it from your rear-view
mirror, yes. Sell a copy, no. Not even if you
pulp the original. Rent out an unauthorized hack
job, no. Stores pay for a license to public
performance, and the terms are right there.
jmorris42 writes,
I actually do understand the issues fairly well...
Wishful thinking makes a poor substitute for real
thinking.
It appears to me that jeremy_hogan has entirely
misunderstood the the topic.
The article claims that the extra permissions
offered in the LGPL are implicit in the GPL, and
that any library offered under the GPL may be used
as if it were released under the LGPL. Right or
wrong, it's the topic under discussion here.
I don't mind if Slashdot sells out. They're a
business, after all.
But promoting plastic sugar-water cans is pretty
far afield from "News for Nerds". I don't see any
way to turn these things into a microwave antenna.
What's next, pearlescent eye-shadow colors?
Britney's new album? (I'd better be careful, maybe
Katz has already done those!)
It looks like Slashdot has discovered a new source
of income: turning press releases into headlines.
It's far from original, but it works -- for a while,
anyway.
At least I hope Slashdot collected some
money for the headline. If not, they're bigger
suckers than those of us who read the article.
I see this article as an attack on the GPL, trying to turn it, essentially, into the Lesser GPL. While this would be convenient for commercial exploiters of GPLed code, it would also make businesses more reluctant to release their work under the GPL, for fear of having their work stolen by rivals.
The article disingenuously ignores cases where works were held to be derived even when they did not include any part of the original work, but only (e.g.) recycled characters from the original. When the second work makes sense only in light of the original, it's derivative. A program that works only when a GPLed library is present is derivative of that library.
Claiming that dependency other than copying doesn't count is a disservice to Mr. Rosen's clients, and to readers at large. An honest article would have at least acknowledged that serious legal scholars disagree with him.
... trying to create a new [right] that not even Schlafly would want if she thought about it
for a moment. (Incidentally,it would knock the
teeth out of the GPL.)
jmorris42 wrote:
If RedHat decided to pull every occurance of 'fuck' from the comments of the Linux source, this would violate the GPL how?
I'm going to assume this isn't meant as a troll,
despite its obtuseness.
The GPL explicitly allows you to redistribute
a bowdlerized Red Hat, as long as you send along
complete sources. Most copyright licenses are
not so liberal. If you were allowed to make any
alterations you like and redistribute, regardless
of license or lack of it, then you would also be
free to redistribute Red Hat with private features
but without sources, and ignore your obligations
under the GPL.
Or is jmorris42 imagining that the law would be
changed to allow copyright to be violated
with impunity only where the violations involve
erasing naughty words? How would you imagine
wording such a law? (It's possible he doesn't
really have any idea what he means.)
People's actions have good or evil consequences (often both); ideas have good or evil implementations. Get beyond the juvenile labels and to the root causes and you'll be a lot better off.
Well, he's right, and then again, he's right, but
Phyllis Schlafly, for heaven's sake?
Who will pop up next on "our side", Ollie North
and John Poindexter and their "Felons Fighting
for Fair Use (FFFFU)"?
More seriously, you can see a hint in her
editorial of how having a whacko on your side
can dirty the debate. Along with all the sound
arguments culled from our literature, she argues
for a right of video stores to distribute
their own bowdlerized versions of movies,
without permission. (Blockbuster does it
routinely, but is big enough to demand and get
permission!)
This confuses the issue, because it's not
defending an existing right, it's trying to
create a new one that not even Schlafly would
want if she thought about it for a moment.
(Incidentally, it would knock the teeth out of
the GPL.)
If you read the specs on the CPU carefully (which
I did, a couple of years back) you discover that
it is not really a RISC at all, by any definition
of the term. The architecture is very similar to
the VAX, in fact. They just call it a RISC
because that once sounded more advanced. Today, of
course, calling an architecture RISC makes it seem
kind of backward, but they've been saying it long
enough that it's probably too late to change.
This is not to say that the designation means
much any more... people have discovered how to
make the most horrendous instruction sets (read:
x86) go fast with only a million (!) extra
transistors or so. This CPU doesn't have those,
but what matters is that it's fast enough.
Still, it's amusing because half the complexity of
the instruction set (and a substantial parcel of
the chip) will never be exercised by any compiler.
It's there as a sort of homage or shrine to
machines from the days when programs were written
in assembly language, and machines were marketed
on how fancy the instruction set was, regardless
of how it slowed the machine down.
The CDC machines were exceptional: Seymour Cray really understood. Also, in the '60s, some
people at IBM built the 801, which evolved into
the PowerPC. The rest of the industry didn't
catch on until the Stanford RISC people made their
big splash.
The still-prevalent myth that lead was good for
engines was one of their big PR coups. Why do you
think engines last so much longer now than when we
used to put leaded fuel in them?
The other myth is that there were no good alternatives. In fact alcohol worked as well
then as now. (It just wasn't patentable.)
They managed to suppress the evidence for just
how toxic was the lead they were scattering around
for many decades. The suppression was deliberate
and criminally fraudulent.
Leaded gasoline was a disaster and a crime on a
scale similar to the asbestos deception of the
same era, but one that has still not been
prosecuted, largely for political reasons. It is almost a miracle that leaded gas got banned at all. The ban certainly wouldn't happen in today's political climate, even though lead was killing a World Trade Center's worth of Americans every
week. Killing Americans is a corporate privilege.
This is such an easy one, I'm astonished nobody
else suggested a good solution.
Just run User-mode Linux or VMWare, according to
what OSes you want to boot natively and from the
removable drive. They will happily boot off any
medium you can plug in.
Or, better, just put your private data in an
encrypted file system, and unmount it when your
employer might get to it. It doesn't even need
its own partition, it can be in a regular file.
Of course this assumes you're running Linux or
a BSD, but you can run VMWare under that to
load any cheesy employer-favored OS.
At my first engineering job, twenty years ago,
the head programmer liked to say, "Sounds like
bullshit to me!" It is only when I read about
things like this that I miss him...
The judges wrote lots that wasn't germane to the
question of jurisdiction. Much of it was false
or misleading, revealing confusion in the minds
of the judges. The dissenting judges were even
more confused than the deciding ones.
The false assumption that the
reverse-engineering was illegal, tainting DeCSS,
matters a lot. If there was no wrongdoing, then
there is no remaining trade secret, and DVDCCA
has no standing.
The claim that posting DeCSS was meant to harm
companies in California, rather than (as in fact)
to help people in other places, matters too.
If California companies are only incidentally
affected, then they cannot claim jurisdiction in
California. That DeCSS was already posted in
hundreds of other places means the notion
of any one posting being "aimed at" anyone is
ridiculous.
Remarkably, the judges never mentioned any of the
following key facts that should have completely
changed the nature of the (original) proceedings:
The DeCSS was already posted on hundreds of
web sites all over the world long before it was
posted on the LiViD site. Posting it on the
LiViD site cannot reasonably be claimed to have
had any substantial effect on anybody, beyond
minor convenience for LiViD developers.
Nobody has shown that the CSS was illegally
reverse-engineered. It's specifically allowed
to reverse-engineer in Norway despite any
contractual agreement. In any case there was no
contractual agreement, just a click-wrap button
which there is no evidence anybody clicked.
Once a trade secret is out, it's out, and anybody
is free to use it. DVDCCA likes to pretend, and
seems to have confused the judges into believing,
that something illegal occurred in Norway. The
worst anybody has come up with is that nobody
knows (despite what Norwegian law says)
what a Norwegian court would actually decide.
These judges insist that LiViD was "aimed at"
the movie and electronic industries, despite that
it has been explained that it was,rather, aimed
at benefiting legitimate owners of DVDs, who
have a Uniform Commercial Code right to watch
the movie they have bought. That some movie or
electronics companies might have been affected
was of no interest to the LiViD project.
Nobody has shown that these companies have been
affected in any way. Certainly lots of movies
are being released on DVDs, and lots of DVD
players are being sold.
I don't know whether to chalk up these omissions
by the judges to malfeasance or incompetence.
(Might as well assume both.:-)
Slashdot Mirror
Don't we have any nice Free alternatives yet, already ported throughout the known universe? (I thought we had dozens, but I don't use that stuff.)
Now, they can say they would have broken it if if wasn't so scary to try.
Otherwise, they would have to admit it would have taken them forty-leven trillion centons just to try out the easy combinations like "haxorz begone", before moving on to the non-alphabetic "hard" ones.
2. What is Cryptome doing on Verio anyway? It's a filthy spammer host.
If you don't know the answer, just say so, and let somebody else speak up.
I asked about examples of results for which commutativity is an essential quality. I don't doubt there are some, or plenty, but what are a few of them?
Please do. I was hoping to learn something. It seems like most extensions from scalars into multiple dimensions (matrices, quaternions) shed commutativity like last year's skin.
Can somebody please give an example of an important and comprehensible theorem that depends on commutativity? That is, what is a consequence that would be meaningful to a non-mathematician?
I don't know of any important theorems that depend on commutativity. Do you?
An interesting paragraph reads:
So, contribute to a released version and get free patent licenses from anybody who uses it, or they have to stop using it. If this were added to the GPL it would mean more.If you bother to read the GPL, you find that it specifically allows linking with OS-supplied libraries.
Try again, troll.
sholden wrote: Since when did your opinion trump the court's? ...
A program is a derivative of a library if a court rules it so (and that may change on appeal...). Copyright law determines what a derivative is, and the courts have the job of interpreting that law.
No court has ruled. Legal scholars disagree.
My remark expressed the opinion the FSF obtained from its legal experts, to provide a concise counterpoint to the article. Rosen should have expressed it himself, and would have done were he interested in a fair exposition of the issue.
Concealing such a legal risk (i.e. that the court might agree with the FSF) would harm his clients. Is he telling his clients one thing and us something else? He's either cheating them or lying to us. Which is worse?
Memerot has understood the topic of the article (and parrots its claim), but not of the discussion. Suffice to say that other legal scholars disagree with him (and Rosen) about what counts as a derivative work, and that Rosen is disingenuous in pretending otherwise.
If Rosen were correct, then the GPL would in fact be equivalent to the LGPL. That's a fact, not under dispute. The debatable question is whether he is correct. The ethical issue is whether Rosen is being scrupulous in failing to present the opposite argument.
Also at issue are Rosen's motives: is he encouraging people to violate library authors' expressed wishes in an attempt to weaken GPL protections from libraries?
I wrote that myself.
Copyright gives the author only limited control. Control of distribution and of public performance are two of the things it does grant to the author. The author exercises that control via licensing. The control is limited by Fair Use (except where that conflicts with the DMCA).
The Uniform Commercial Code (UCC) gives the purchaser extra rights, beyond Fair Use. It does not grant the right to distribute derived works or make public performances in violation of the license extended by the author.
jmorris42 writes further, "the clean flicks people buy a number of copies, replace the disc in the package with an edited version and rent them".
Renting counts as public performance, which is also restricted under copyright. If they're allowed to make public performances of private versions without permission, then exploiters will also be allowed to redistribute private versions of GPLed programs without permission (i.e. without sources).
Under the UCC the video stores are free to do whatever they need to to make the physical disc they bought do what the seller promised. (That's what makes shrink-wrap licenses wastepaper.) Selling a modified copy of it, however, is not something the UCC allows. Play the disc, decrypt the disc, hang it from your rear-view mirror, yes. Sell a copy, no. Not even if you pulp the original. Rent out an unauthorized hack job, no. Stores pay for a license to public performance, and the terms are right there.
jmorris42 writes, I actually do understand the issues fairly well ...
Wishful thinking makes a poor substitute for real thinking.
The article claims that the extra permissions offered in the LGPL are implicit in the GPL, and that any library offered under the GPL may be used as if it were released under the LGPL. Right or wrong, it's the topic under discussion here.
But promoting plastic sugar-water cans is pretty far afield from "News for Nerds". I don't see any way to turn these things into a microwave antenna. What's next, pearlescent eye-shadow colors? Britney's new album? (I'd better be careful, maybe Katz has already done those!)
At least I hope Slashdot collected some money for the headline. If not, they're bigger suckers than those of us who read the article.
The article disingenuously ignores cases where works were held to be derived even when they did not include any part of the original work, but only (e.g.) recycled characters from the original. When the second work makes sense only in light of the original, it's derivative. A program that works only when a GPLed library is present is derivative of that library.
Claiming that dependency other than copying doesn't count is a disservice to Mr. Rosen's clients, and to readers at large. An honest article would have at least acknowledged that serious legal scholars disagree with him.
The GPL explicitly allows you to redistribute a bowdlerized Red Hat, as long as you send along complete sources. Most copyright licenses are not so liberal. If you were allowed to make any alterations you like and redistribute, regardless of license or lack of it, then you would also be free to redistribute Red Hat with private features but without sources, and ignore your obligations under the GPL.
Or is jmorris42 imagining that the law would be changed to allow copyright to be violated with impunity only where the violations involve erasing naughty words? How would you imagine wording such a law? (It's possible he doesn't really have any idea what he means.)
Who will pop up next on "our side", Ollie North and John Poindexter and their "Felons Fighting for Fair Use (FFFFU)"?
More seriously, you can see a hint in her editorial of how having a whacko on your side can dirty the debate. Along with all the sound arguments culled from our literature, she argues for a right of video stores to distribute their own bowdlerized versions of movies, without permission. (Blockbuster does it routinely, but is big enough to demand and get permission!) This confuses the issue, because it's not defending an existing right, it's trying to create a new one that not even Schlafly would want if she thought about it for a moment. (Incidentally, it would knock the teeth out of the GPL.)
With friends like Phyllis Schlafly, who needs enemies?
This is not to say that the designation means much any more... people have discovered how to make the most horrendous instruction sets (read: x86) go fast with only a million (!) extra transistors or so. This CPU doesn't have those, but what matters is that it's fast enough.
Still, it's amusing because half the complexity of the instruction set (and a substantial parcel of the chip) will never be exercised by any compiler. It's there as a sort of homage or shrine to machines from the days when programs were written in assembly language, and machines were marketed on how fancy the instruction set was, regardless of how it slowed the machine down.
The CDC machines were exceptional: Seymour Cray really understood. Also, in the '60s, some people at IBM built the 801, which evolved into the PowerPC. The rest of the industry didn't catch on until the Stanford RISC people made their big splash.
The other myth is that there were no good alternatives. In fact alcohol worked as well then as now. (It just wasn't patentable.)
They managed to suppress the evidence for just how toxic was the lead they were scattering around for many decades. The suppression was deliberate and criminally fraudulent.
Leaded gasoline was a disaster and a crime on a scale similar to the asbestos deception of the same era, but one that has still not been prosecuted, largely for political reasons. It is almost a miracle that leaded gas got banned at all. The ban certainly wouldn't happen in today's political climate, even though lead was killing a World Trade Center's worth of Americans every week. Killing Americans is a corporate privilege.
Just run User-mode Linux or VMWare, according to what OSes you want to boot natively and from the removable drive. They will happily boot off any medium you can plug in.
Or, better, just put your private data in an encrypted file system, and unmount it when your employer might get to it. It doesn't even need its own partition, it can be in a regular file. Of course this assumes you're running Linux or a BSD, but you can run VMWare under that to load any cheesy employer-favored OS.
At my first engineering job, twenty years ago, the head programmer liked to say, "Sounds like bullshit to me!" It is only when I read about things like this that I miss him...
The false assumption that the reverse-engineering was illegal, tainting DeCSS, matters a lot. If there was no wrongdoing, then there is no remaining trade secret, and DVDCCA has no standing.
The claim that posting DeCSS was meant to harm companies in California, rather than (as in fact) to help people in other places, matters too. If California companies are only incidentally affected, then they cannot claim jurisdiction in California. That DeCSS was already posted in hundreds of other places means the notion of any one posting being "aimed at" anyone is ridiculous.
Remarkably, the judges never mentioned any of the following key facts that should have completely changed the nature of the (original) proceedings:
- The DeCSS was already posted on hundreds of
web sites all over the world long before it was
posted on the LiViD site. Posting it on the
LiViD site cannot reasonably be claimed to have
had any substantial effect on anybody, beyond
minor convenience for LiViD developers.
-
Nobody has shown that the CSS was illegally
reverse-engineered. It's specifically allowed
to reverse-engineer in Norway despite any
contractual agreement. In any case there was no
contractual agreement, just a click-wrap button
which there is no evidence anybody clicked.
Once a trade secret is out, it's out, and anybody
is free to use it. DVDCCA likes to pretend, and
seems to have confused the judges into believing,
that something illegal occurred in Norway. The
worst anybody has come up with is that nobody
knows (despite what Norwegian law says)
what a Norwegian court would actually decide.
-
These judges insist that LiViD was "aimed at"
the movie and electronic industries, despite that
it has been explained that it was,rather, aimed
at benefiting legitimate owners of DVDs, who
have a Uniform Commercial Code right to watch
the movie they have bought. That some movie or
electronics companies might have been affected
was of no interest to the LiViD project.
-
Nobody has shown that these companies have been
affected in any way. Certainly lots of movies
are being released on DVDs, and lots of DVD
players are being sold.
I don't know whether to chalk up these omissions by the judges to malfeasance or incompetence. (Might as well assume both.:-)