It's relatively easy to write a content filter for audio which can detect reasonably close copies (ie, is tolerant of different MP3 bitrates and encoder variants)-- companies like GraceNote and Shazam have services available which can be used to ID music files, but they are at the mercy of having up-to-date signatures to catch the latest music going around.
Depending on the service, they may have a web-based POST mechanism which returns an XML result, which can accept either raw PCM/WAV or sometimes other formats like WMA or MP3. They also tend to have a "fingerprinting" utility that will process incoming raw PCM/WAV audio and come up with the sonic "checksum" they compare with their signature database of all of the recognizable music their service knowns about. False positives are extremely rare ( 1%), and for known-recognizable songs, they could be 99+% reliably recognized given about 1-2 MB of PCM sample data from anywhere in the song, corresponding to between 10 - 30 seconds of play time from a clip (modulo sample rate and lots of other stuff).
On the other hand, the recognition can be easily fooled by significant noise or hiss in the recording: ie, a live version of a song taped from a show would likely not be recognized especially if there was audience noise present.
I have no direct knowledge of how the recognition side works for video content, but I suspect the situation is similar. (Ie, it's easy to match pretty reliably from even a few seconds of a copyrighted sample once it is in the database.)
The moon's average temperature is somewhat warmer than that of Mars as Luna is closer to the sun, but because Luna is tidally locked with the Earth, it has a ~28-earth-day long "day", which means the rovers would have to deal with 14-earth-day periods of darkness and cold, which is not ideal for solar-powered units.
RTG-powered rovers would probably work out better....
Actually, the Africans who were sold as slaves to the United States in the 1700s were enslaved by their brother Africans as a more profitable alternative than simply killing the losing side of a tribal conflict en masse, which previously was a somewhat common outcome of inter-tribal warfare in Africa.
Sadly, that kind of tribal warfare-- including mass-murder of civilians-- is still going on today (cf. Sudan)....
1500 amps/phase sounds like a building feeder here in NYC (it'd be about a residential block's worth in more rural areas). Of course, that's the designed sustained load; when you manage to do a phase short you'll experience extremely high transient current loads-- we're talking tens to possibly hundreds of thousands of amps-- for a fraction of one 60Hz cycle (ie, on the close order of a few milliseconds) until the fuses blow.
Be glad it was the fuses which blew and not the equipment being powered by it.:-)
Shorts like that are generally a result of frayed insulation, but human error (ie, cutting into a wall and cutting into a power line behind it) can cause them as well.
I beg Volton's pardon-- never argue with a giant mecha robot that'll kick your butt for being impolite-- but the 6502 predates him by several years: 6502 V flag
Even the scientists claim expertise when they do not really know.
This is true of all sorts of people, but it is less true of genuine experts (in whatever field).
What we need are people to be honest and say that they really do not know what is causing global warming or disease, etc. We need open research minds that are objective. The global warming campaign is not very objective. The fact is, no one understands earth's climate completely. We may have some understanding of small aspects of the climate but we have yet to see how these aspects affect the larger picture.
We don't have to understand all of the details of a complex system like the world's climate to make useful and reliable predictions. Your local weatherman can predict the weather over the next week with a lot more accuracy than someone could 100 years ago. And, while we can't predict whether it will rain on one specific day a year from now, and likely never will be able to due to Lorentz and the "butterfly effect", we're getting pretty decent at predicting how much rainfall will be seen over a month a year from now
And we don't have to understand and be able to cure every single disease to demonstrate that the basics of "germs cause disease" works very well. Simple things like boiling water before using it to clean wounds or sterilize dressings makes an amazing difference; similar effects occur with the prevention of "childbed fever". You do realize that before Semmelweis & Pasteur in the 19th century, something like 15-20 % of pregnant women died after giving birth?
There has been definitive evidence that homeopathic and herbal treatments can be effective without the side effects.
True. But then, there has been definitive evidence that giving people sugar pills (as a placebo) will result in a ~30% effective treatment for pain or for things like a viral flu which has no current effective treatment.
Agreed. To respond to the GP post: I qualify as "American", and I would write today's date as "2006-12-14" or "20061214" or maybe "2006/12/14", depending on whether I was just writing it out on paper, using the date stamp in a DNS zone file or some other computer representation, or setting up a directory hierarchy for something like daily webserver logfiles.
Most US-based computer companies commonly use the YYYY-MM-DD format now, but YMMV. (That's "Your Mileage May Vary", not "YEAR+MONTH+overflow bit"....:-)
No one likes a paedophile. It's hard wired at an instinctual level as a biological protection of our own offspring. There is no "moral question" -- it's so important it's a physically hardwired response to view such activities with disgust and contempt.
Whether paedophilia is considered "wrong" or whether people feel "disgust and contempt" towards it is a learned response based upon your particular upbringing, and your reaction is about as common as hating someone because they are white [or black, or hispanic, or in the shudra or dalit ("untouchable") Indian castes, etc].
Note that paedophilia used to be a normal part of many ancient cultures-- go read Homer's Illiad and the Odysseus again, carefully, preferably in the original Greek. What was the relationship between Achilles and Patroclus? Lets see what this says:
The relationship between Achilles and Patroclus is a key element of the myths associated with the Trojan War. Its exact nature has been a subject of dispute in both the classical period and modern times. In the Iliad, it is clear that the two heroes (who are also first cousins once removed) have a deep and extremely meaningful friendship, but the evidence of a romantic or sexual element is equivocal. Commentators from the classical period to today have tended to interpret the relationship through the lens of their own cultures. Thus, in 5th century BC Athens the relationship was commonly interpreted as pederastic. Contemporary readers are more likely to interpret the two heroes either as non-sexual "war buddies" or as an egalitarian homosexual couple.
The Ancient Mediterranean world had vastly different attitudes toward gender and sexuality than those found in 21st century North America or Europe. Although sexual relations between men are well attested, there was no term for, or concept of, homosexuality as such.
PS: Note carefully that I haven't expressed an opinion about paedophilia. I think child abuse is wrong, and I don't believe that children should be involved in sexual activity until they are old enough to make their own decisions and form mutually consensual relationships. In most of the world and for most of human history this happens in the early teens; only in first-world countries do people live long enough that procreation can be delayed to the late teens or early twenties.
Usenet was a cesspool of spam when web forums started to get popular.
Actually, the first major spam to Usenet was around April, 1994-- the green-card lottery mass-posted everywhere by a pair of bottom-feeding lawyers. Before then, Usenet was almost entirely free of spam:
If you've run a honeynet, you'll find that you tend to see between ~300 and ~1500 or so "attacks" per IP address per day-- about 80% TCP-based, about 15% UDP-based, and about 5% ICMP-based. I'm not sure a simple ICMP ECHO_REQUEST qualifies as an "attack" (although there are plenty of security vendors who will claim it is, simply to inflate their numbers), but ICMP redirects which try to tell a host to send local traffic to a remote IP surely does qualify as a hostile attack.
Assuming that there's about 1000 attacks per day on average, or 30K per month per IP, suggests that Microsoft only has three or four Internet-routable machines, which clearly isn't the case-- perhaps they are only counting attacks which make it through the front line of their existing firewalls, or they are aggregating a single source IP which launches the same viral payload against many destination IPs as a single "attack"...?
If I'm testing a DNS server, I try pinging it before doing anything else, and then graduate to using "dig" to see whether the DNS server can look up other hostnames...
I assumed you were talking about glibc, sorry about that.
You deserve credit for acknowledging at least the most basic point.:-)
1. If you link your code to glibc, you are making a derived work of glibc.
This claim is still very much open to question and is, as yet, an untested matter of law, when the library in question is a standard part of the operating system.
2. If you distribute your source under a restrictive license without linking it to glibc and you only use apis that were documented elsewhere, then you're not distributing a derived work of glibc.
Yes, agreed. This is one of my main points.
In fact, you could dynamicly link to a shared version of libc.so, and distribute your binary without including libc.so, and your program would still not be a derivative work, so long as it was using publicly documented APIs found elsewhere.
Let's skip ahead to the GNU readline case:
2. If you distribute your source under a restrictive license without linking it to readline and you only use apis that were documented elsewhere, then you're not distributing a derived work of readline. Problem is, there's no "elsewhere" for GNU readline. It was invented there.
There are actually at least two libraries similar to GNU readline which are not under the GPL; one was written by Christos Zoulas and is called editline (or libedit.a) and ships with FreeBSD, NetBSD, and OpenBSD, I believe; and there was another implementation called "BSD libreadline" implemented by Jaromir Dolecek which appears in NetBSD and MacOS X 10.4 (aka "Tiger").
3. This is moot as you can't legally distribute the source, let alone a binary.
4. And yeah, moot again because your program clearly isn't independant. You could remove the readline specific parts and distribute that under a restrictive license, but that's irrelevant.
There are proprietary programs which will dynamically load libreadline, if it exists, or use their own internal readline/line-editting capabilities otherwise. So long as the program works in either case, and works with the BSD licensed implementation of readline, the program "can be reasonably considered independent" of the GNU version of readline. Per the GPL clause 2, that program is not considered a derivative work of the GPL'ed library and "this License, and its terms, do not apply to those sections when you distribute them as separate works."
So getting back to what we were talking about.. NVIDIA have written a kernel module which contains some source portions and some binary portions.
[... ] The whole thing is under a restrictive license.
Agreed.
They're using apis that are a part of linux and are documented nowhere but in linux.
You sure about that?
A review of the driver sources and the use of nm or otool to look at the symbols of the binary portion suggests that the driver is using primarily nVidia's internal APIs and then those for MESA/OpenGL's GLX extension to X11, not Linux specific APIs. Go look for yourself.
They didn't exist before linux. [... ] NVIDIA has deliberately set out to create a restricted work that links to GPL code. A blind chipmuck with a pencil in its teeth could win this court case.
Actually, X11 and I believe the GLX extension predates the existence of Linux.
nVidia itself doesn't quite, but nVidia was releasing proprietary drivers for Windows for nearly a decade before they released a Linux driver. The binary portion of the nVidia Linux driver is not a derived work of Linux or the Linux kernel, and you can check for yourself that the binary portion uses no symbols specific to the Linux kernel-- only the shim they've published in source code form does. The end-user of a system is free to compile that shim to produce a Linux kernel module which can be loaded, but, because the GPL and the nVidia proprietary license conflict, the end-user cannot redistribute the combination, which is a derivative work of both.
I've never had to agree to the terms of a library simply in order to use it. I might have to agree to those terms to redistribute the library itself, but, per the GPL clause #2:
"These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works."
If the same source code can be compiled using a BSD libc and a GNU libc, then that source code "can be reasonably considered independent" and the GPL does not apply to the program itself.
That's why the standard C library is under the LGPL, not the GPL.
Actually, no, the standard C library was defined by Kernighan and Richie. It was originally implemented by them back in 1971 as part of AT&T V1 Unix, which was later made available in the mid-70's in what is now called BSD Unix. GNU Libc is a GPL'ed re-implementation of Kernigahn & Richie's published APIs, created by Roland McGrath around 1992.
Jesus, it's like copyright 101 in here.
Yep, and I've got news for you: you're failing to pass.
If you think that relicensing the GNU libc under the GPL rather than the LGPL would make every C program which does a #include <stdio> a derived work of GNU libc, then you don't understand either the history of the C language or copyright law.
You don't have to be a lawyer to understand that GNU libc is released under the LGPL, and if it were released under the GPL instead then any program using printf in glibc would be required to be GPL as well.
Kernighan and Richie would certainly disagree with that claim.
I'll repeat an observation I made to someone else: if the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
Which APIs would those be? The whole point of them writing a wrapper is that there are no standard published APIs.
Sure there are. Check "man 9 intro" and the output of "ls/usr/share/man/man9".
Or look at Linux Kernel API....
They are #including header files from the kernel. It's clearly a derived work.
Does a userland C program which does a "#include " become a derived work of the standard C library?
If the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
kernel modules are derived works of the kernel. As such, they are required to be GPL licensed
There is substantial reason to believe that kernel modules are not derived works of the kernel if they use the standard published kernel APIs, just as normal userland programs are not considered derivative works of the standard system header files and libraries, even if the userland program #include's or links against libc.so.
It's also true that the Linux kernel includes a fair amount of BSD-licensed source code, so it's obviously not true that all Linux kernel modules need to be under the GPL to be redistributed. A quick check against the Linux-2.6.0 kernel suggests 274 examples:
% find linux-2.6.0 -print0 | xargs -0 grep -l 'Redistribution and use in source and binary forms, with or without' | wc -l
274
These files start with linux-2.6.0/arch/m68k/mac/iop.c, include the ACPI & scsi/aic7xxx drivers, NFSd (linux-2.6.0/fs/nfs/idmap.c) & SunRPC (linux-2.6.0/net/sunrpc/auth_gss/auth_gss.c), and others.
1. are distributing a derived work
While Eben Moglen may have other opinions, most people on the OSI mailing lists did not consider the nVidia driver to be a derived work of the Linux kernel.
2. is bound by the license on the kernel if they are not distributing the kernel
The answer to this is obviously no.
3. is satisfying the terms of the GPL by distributing partial source code to their driver
The answer to this is obviously no, but the nVidia driver isn't and never has been under the terms of the GPL, because nVidia wrote it themselves.
4. would only be violating copyright if they were distributing binaries with no source code wrapper
The answer to this is "probably no", going by the "published API"s standard I mentioned above for deciding whether source code constitutes a derivative work or not.
5. are guilty of contributory copyright infringement by facilitating others to distribute binaries if 4 were the case
This suggestion comes straight from Eben Moglen, and is unlikely to gain much traction in a court of law. (a) nVidia has done nothing to encourage people to redistribute a Linux kernel plus the proprietary nVidia driver. (b) nVidia releasing a free (as in "no cost to download") driver for someone to use their nVidia video card with Linux would almost certainly qualify as "fair use" and be protected under 17 USC 107.
And if not, just who would you sue? Linux users who want GLX hardware acceleration? SCO sued it's userbase, and look how far that got them...
You need to hit some better bars.:-) Start with 7B (aka Vasacs, the bar in Crocodile Dundee), or the Patriot bar (long live the Village Idiot!), or Ace bar between 5A & 5B. Or head a bit uptown and visit Yogi's off of 75th? & Broadway. Several of the above will serve a 64oz pitcher of Guiness for $9-10. Say hi to Tommy and Lilly...
And driving along a highway in a game that has billboards would effect gameplay how? If it is some mountain path that doesn't have billboards...yes, I can understand how that would ruin the immersion...but on a typical highway, there are billboards.
Depends on where you are. Apparently, there are four states which have laws against billboards-- Maine & Vermont, are two, I believe.
I understand you hate ads and will probably spontaneously combust if you see them in a game.
I don't hate all ads, but I dislike annoying ads. I can recall from way back when, that ads seemed to be a lot less annoying ten or twenty years ago than they are today.
Well, tough, they are going to be there so it is best advertisers understand how they can do it without pissing off the majority of gamers. Some small number of people may boycott games with ads...but they aren't going to be significant enough to matter (e.g. bnetd and blizzard).
Welcome to Slashdot; this is where geeks sometimes solve problems, not just talk about them. In the case of games using Massive's ad-network, I suggest setting up your own nameserver on your LAN, and making it authoritative for madserver.net, via a zone file like:
$TTL 86400
@ IN SOA ns1.example.com. hostmaster.example.com. ( 2006111900 ; serial 3h ; Refresh 3 hours 1h ; Retry 1 hour 30d ; Expire 30 days 1d ) ; Minimum 24 hours
@ NS ns1.example.com.
localhost A 127.0.0.1
madsever.net. A 192.168.1.2 ad A 192.168.1.2 imp A 192.168.1.2 media A 192.168.1.2
...where, in my case, IP 192.168.1.2 is the IP of the nameserver, as I wanted to do some packet sniffing and see what kind of traffic a game using Massive's ads creates. And yes, those should be tabs, not spaces...<ecode> doesn't behave quite like <pre>, unfortunately.
If you like, sure. I don't consider a manufacturer's logo on or in a product to be the same as displaying third-party ads, but if you want to call it advertising, that's a reasonable position.
"Broadband households have become even harder to reach: some 81% of those with high-speed Internet access employ pop-up blockers and spam filters."
It's not surprising, either. At one point, it was commonly recognized that computers belonged to the people that owned them, and that it was the responsibility of people writing software to make sure that the software was well-behaved and did what the user told the software to do-- except for deliberately malicious software. While I do not claim that all forms of advertising are malicious, it's becoming the case that websites using lots of pop-up or pop-under ads, or software like games using Massive's technology or other in-game ad-delivery mechanisms operate under the assumption that they are free to do things with the user's computer and consume networking resources to fetch and display content that the user didn't ask for and does not want.
I can tolerate ad-bars appearing on the right-hand side of a page, so long as most of the screen real-estate shows the actual content I want, but some sites do obnoxious and deceptive things like displaying an interstital ad first. My response to that is to copy the ad link into an email, and send a complaint off to both the webmaster of the site I was on, and the site holding the advertising, indicating that their ad was so annoying that I won't be returning to the offending site for at least one week, and that obviously they will be losing my eyeballs and ad impression revenue for that period of time.
It seems to have an effect, too. At least two of the newspapers I visit (the Boston Globe & the LA Times) have toyed with interstitial ads and have dropped them soon afterwards....
And you would enjoy those games just as much if there was a billboard every now and then on the side of the road
Speak for yourself, not for me, OK?
Look, in case you don't realize, you can drive between San Francisco and San Jose in California on I-280, and find a half-dozen real-world references like Pacifica, Mountain View, Alpine Rd, Portorola, and others which are scenic drives you can take-- twisty, windy roads up the side of small mountain ridges-- which EA games based the tracks in much of the NeedForSpeed series on. They're fun to drive in a decent sporty car, and might be even better if you want to ride a motorbike over 'em; I go driving to enjoy the view and the outstanding scenery of the region, especially in mid-fall.
I don't drive them to watch billboards.
NASCAR/F1 racing has ads everywhere in reality. So having the ads in the game change nothing.
Of course NASCAR/F1 has ads everywhere in reality. Having the ads in real NASCAR/F1 racing means I choose not to watch or participate in NASCAR/F1 racing. Having the ads in a racing game means I won't purchase the game. But I might purchase a F1-style racer if it offered non-circuit tracks.
Slashdot Mirror
It's relatively easy to write a content filter for audio which can detect reasonably close copies (ie, is tolerant of different MP3 bitrates and encoder variants)-- companies like GraceNote and Shazam have services available which can be used to ID music files, but they are at the mercy of having up-to-date signatures to catch the latest music going around.
Depending on the service, they may have a web-based POST mechanism which returns an XML result, which can accept either raw PCM/WAV or sometimes other formats like WMA or MP3. They also tend to have a "fingerprinting" utility that will process incoming raw PCM/WAV audio and come up with the sonic "checksum" they compare with their signature database of all of the recognizable music their service knowns about. False positives are extremely rare ( 1%), and for known-recognizable songs, they could be 99+% reliably recognized given about 1-2 MB of PCM sample data from anywhere in the song, corresponding to between 10 - 30 seconds of play time from a clip (modulo sample rate and lots of other stuff).
On the other hand, the recognition can be easily fooled by significant noise or hiss in the recording: ie, a live version of a song taped from a show would likely not be recognized especially if there was audience noise present.
I have no direct knowledge of how the recognition side works for video content, but I suspect the situation is similar. (Ie, it's easy to match pretty reliably from even a few seconds of a copyrighted sample once it is in the database.)
The moon's average temperature is somewhat warmer than that of Mars as Luna is closer to the sun, but because Luna is tidally locked with the Earth, it has a ~28-earth-day long "day", which means the rovers would have to deal with 14-earth-day periods of darkness and cold, which is not ideal for solar-powered units.
RTG-powered rovers would probably work out better....
Actually, the Africans who were sold as slaves to the United States in the 1700s were enslaved by their brother Africans as a more profitable alternative than simply killing the losing side of a tribal conflict en masse, which previously was a somewhat common outcome of inter-tribal warfare in Africa.
Sadly, that kind of tribal warfare-- including mass-murder of civilians-- is still going on today (cf. Sudan)....
1500 amps/phase sounds like a building feeder here in NYC (it'd be about a residential block's worth in more rural areas). Of course, that's the designed sustained load; when you manage to do a phase short you'll experience extremely high transient current loads-- we're talking tens to possibly hundreds of thousands of amps-- for a fraction of one 60Hz cycle (ie, on the close order of a few milliseconds) until the fuses blow.
:-)
Be glad it was the fuses which blew and not the equipment being powered by it.
Shorts like that are generally a result of frayed insulation, but human error (ie, cutting into a wall and cutting into a power line behind it) can cause them as well.
You need the 65816 upgrade.
(I don't think it will let you play Voltron, but it would let you play Repton or Rescue Raiders...)
I beg Volton's pardon-- never argue with a giant mecha robot that'll kick your butt for being impolite-- but the 6502 predates him by several years: 6502 V flag
This is true of all sorts of people, but it is less true of genuine experts (in whatever field).
What we need are people to be honest and say that they really do not know what is causing global warming or disease, etc. We need open research minds that are objective. The global warming campaign is not very objective. The fact is, no one understands earth's climate completely. We may have some understanding of small aspects of the climate but we have yet to see how these aspects affect the larger picture.
We don't have to understand all of the details of a complex system like the world's climate to make useful and reliable predictions. Your local weatherman can predict the weather over the next week with a lot more accuracy than someone could 100 years ago. And, while we can't predict whether it will rain on one specific day a year from now, and likely never will be able to due to Lorentz and the "butterfly effect", we're getting pretty decent at predicting how much rainfall will be seen over a month a year from now
And we don't have to understand and be able to cure every single disease to demonstrate that the basics of "germs cause disease" works very well. Simple things like boiling water before using it to clean wounds or sterilize dressings makes an amazing difference; similar effects occur with the prevention of "childbed fever". You do realize that before Semmelweis & Pasteur in the 19th century, something like 15-20 % of pregnant women died after giving birth?
There has been definitive evidence that homeopathic and herbal treatments can be effective without the side effects.
True. But then, there has been definitive evidence that giving people sugar pills (as a placebo) will result in a ~30% effective treatment for pain or for things like a viral flu which has no current effective treatment.
Agreed. To respond to the GP post: I qualify as "American", and I would write today's date as "2006-12-14" or "20061214" or maybe "2006/12/14", depending on whether I was just writing it out on paper, using the date stamp in a DNS zone file or some other computer representation, or setting up a directory hierarchy for something like daily webserver logfiles.
Most US-based computer companies commonly use the YYYY-MM-DD format now, but YMMV.
(That's "Your Mileage May Vary", not "YEAR+MONTH+overflow bit"....:-)
Whether paedophilia is considered "wrong" or whether people feel "disgust and contempt" towards it is a learned response based upon your particular upbringing, and your reaction is about as common as hating someone because they are white [or black, or hispanic, or in the shudra or dalit ("untouchable") Indian castes, etc].
Note that paedophilia used to be a normal part of many ancient cultures-- go read Homer's Illiad and the Odysseus again, carefully, preferably in the original Greek. What was the relationship between Achilles and Patroclus? Lets see what this says:
Wikipedia on Achilles
PS: Note carefully that I haven't expressed an opinion about paedophilia. I think child abuse is wrong, and I don't believe that children should be involved in sexual activity until they are old enough to make their own decisions and form mutually consensual relationships. In most of the world and for most of human history this happens in the early teens; only in first-world countries do people live long enough that procreation can be delayed to the late teens or early twenties.
Actually, the first major spam to Usenet was around April, 1994-- the green-card lottery mass-posted everywhere by a pair of bottom-feeding lawyers. Before then, Usenet was almost entirely free of spam:
Wikipedia on Newsgroup Spam
Michael Valentine Smith isn't really a space marine, he's a duly accredited Envoy from our new Martian overlords.
The Martians don't really need human marines; you didn't think that asteroid belt between Mars and Jupiter happened by accident, do you?
If you've run a honeynet, you'll find that you tend to see between ~300 and ~1500 or so "attacks" per IP address per day-- about 80% TCP-based, about 15% UDP-based, and about 5% ICMP-based. I'm not sure a simple ICMP ECHO_REQUEST qualifies as an "attack" (although there are plenty of security vendors who will claim it is, simply to inflate their numbers), but ICMP redirects which try to tell a host to send local traffic to a remote IP surely does qualify as a hostile attack.
Assuming that there's about 1000 attacks per day on average, or 30K per month per IP, suggests that Microsoft only has three or four Internet-routable machines, which clearly isn't the case-- perhaps they are only counting attacks which make it through the front line of their existing firewalls, or they are aggregating a single source IP which launches the same viral payload against many destination IPs as a single "attack"...?
If I'm testing a DNS server, I try pinging it before doing anything else, and then graduate to using "dig" to see whether the DNS server can look up other hostnames...
You deserve credit for acknowledging at least the most basic point. :-)
1. If you link your code to glibc, you are making a derived work of glibc.
This claim is still very much open to question and is, as yet, an untested matter of law, when the library in question is a standard part of the operating system.
2. If you distribute your source under a restrictive license without linking it to glibc and you only use apis that were documented elsewhere, then you're not distributing a derived work of glibc.
Yes, agreed. This is one of my main points.
In fact, you could dynamicly link to a shared version of libc.so, and distribute your binary without including libc.so, and your program would still not be a derivative work, so long as it was using publicly documented APIs found elsewhere.
Let's skip ahead to the GNU readline case:
2. If you distribute your source under a restrictive license without linking it to readline and you only use apis that were documented elsewhere, then you're not distributing a derived work of readline. Problem is, there's no "elsewhere" for GNU readline. It was invented there.
There are actually at least two libraries similar to GNU readline which are not under the GPL; one was written by Christos Zoulas and is called editline (or libedit.a) and ships with FreeBSD, NetBSD, and OpenBSD, I believe; and there was another implementation called "BSD libreadline" implemented by Jaromir Dolecek which appears in NetBSD and MacOS X 10.4 (aka "Tiger").
3. This is moot as you can't legally distribute the source, let alone a binary.
4. And yeah, moot again because your program clearly isn't independant. You could remove the readline specific parts and distribute that under a restrictive license, but that's irrelevant.
There are proprietary programs which will dynamically load libreadline, if it exists, or use their own internal readline/line-editting capabilities otherwise. So long as the program works in either case, and works with the BSD licensed implementation of readline, the program "can be reasonably considered independent" of the GNU version of readline. Per the GPL clause 2, that program is not considered a derivative work of the GPL'ed library and "this License, and its terms, do not apply to those sections when you distribute them as separate works."
So getting back to what we were talking about.. NVIDIA have written a kernel module which contains some source portions and some binary portions. ... ] The whole thing is under a restrictive license.
[
Agreed.
They're using apis that are a part of linux and are documented nowhere but in linux.
You sure about that?
A review of the driver sources and the use of nm or otool to look at the symbols of the binary portion suggests that the driver is using primarily nVidia's internal APIs and then those for MESA/OpenGL's GLX extension to X11, not Linux specific APIs. Go look for yourself.
They didn't exist before linux. [ ... ] NVIDIA has deliberately set out to create a restricted work that links to GPL code. A blind chipmuck with a pencil in its teeth could win this court case.
Actually, X11 and I believe the GLX extension predates the existence of Linux.
nVidia itself doesn't quite, but nVidia was releasing proprietary drivers for Windows for nearly a decade before they released a Linux driver. The binary portion of the nVidia Linux driver is not a derived work of Linux or the Linux kernel, and you can check for yourself that the binary portion uses no symbols specific to the Linux kernel-- only the shim they've published in source code form does. The end-user of a system is free to compile that shim to produce a Linux kernel module which can be loaded, but, because the GPL and the nVidia proprietary license conflict, the end-user cannot redistribute the combination, which is a derivative work of both.
I've never had to agree to the terms of a library simply in order to use it.
I might have to agree to those terms to redistribute the library itself, but, per the GPL clause #2:
"These requirements apply to the modified work as a whole. If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works."
If the same source code can be compiled using a BSD libc and a GNU libc, then that source code "can be reasonably considered independent" and the GPL does not apply to the program itself.
Actually, no, the standard C library was defined by Kernighan and Richie. It was originally implemented by them back in 1971 as part of AT&T V1 Unix, which was later made available in the mid-70's in what is now called BSD Unix. GNU Libc is a GPL'ed re-implementation of Kernigahn & Richie's published APIs, created by Roland McGrath around 1992.
Jesus, it's like copyright 101 in here.
Yep, and I've got news for you: you're failing to pass.
If you think that relicensing the GNU libc under the GPL rather than the LGPL would make every C program which does a #include <stdio> a derived work of GNU libc, then you don't understand either the history of the C language or copyright law.
People like being confused, I guess. :-)
You don't have to be a lawyer to understand that GNU libc is released under the LGPL, and if it were released under the GPL instead then any program using printf in glibc would be required to be GPL as well.
Kernighan and Richie would certainly disagree with that claim.
I'll repeat an observation I made to someone else: if the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
That's supposed to be "#include "...I forgot to escape the less-than & greater-than symbols in HTML posting mode.
Sure there are. Check "man 9 intro" and the output of "ls /usr/share/man/man9".
Or look at Linux Kernel API....
They are #including header files from the kernel. It's clearly a derived work.
Does a userland C program which does a "#include " become a derived work of the standard C library?
If the same exact source code works on both Linux, Windows, and BSD, does that C program somehow have to be a derivative work of Linux's GNU libc, the Windows/Visual-C libraries, *and* the traditional BSD libc, all at the same time...?
There is substantial reason to believe that kernel modules are not derived works of the kernel if they use the standard published kernel APIs, just as normal userland programs are not considered derivative works of the standard system header files and libraries, even if the userland program #include's or links against libc.so.
It's also true that the Linux kernel includes a fair amount of BSD-licensed source code, so it's obviously not true that all Linux kernel modules need to be under the GPL to be redistributed. A quick check against the Linux-2.6.0 kernel suggests 274 examples:
% find linux-2.6.0 -print0 | xargs -0 grep -l 'Redistribution and use in source and binary forms, with or without' | wc -l
274
These files start with linux-2.6.0/arch/m68k/mac/iop.c, include the ACPI & scsi/aic7xxx drivers, NFSd (linux-2.6.0/fs/nfs/idmap.c) & SunRPC (linux-2.6.0/net/sunrpc/auth_gss/auth_gss.c), and others.
1. are distributing a derived work
While Eben Moglen may have other opinions, most people on the OSI mailing lists did not consider the nVidia driver to be a derived work of the Linux kernel.
2. is bound by the license on the kernel if they are not distributing the kernel
The answer to this is obviously no.
3. is satisfying the terms of the GPL by distributing partial source code to their driver
The answer to this is obviously no, but the nVidia driver isn't and never has been under the terms of the GPL, because nVidia wrote it themselves.
4. would only be violating copyright if they were distributing binaries with no source code wrapper
The answer to this is "probably no", going by the "published API"s standard I mentioned above for deciding whether source code constitutes a derivative work or not.
5. are guilty of contributory copyright infringement by facilitating others to distribute binaries if 4 were the case
This suggestion comes straight from Eben Moglen, and is unlikely to gain much traction in a court of law. (a) nVidia has done nothing to encourage people to redistribute a Linux kernel plus the proprietary nVidia driver. (b) nVidia releasing a free (as in "no cost to download") driver for someone to use their nVidia video card with Linux would almost certainly qualify as "fair use" and be protected under 17 USC 107.
And if not, just who would you sue? Linux users who want GLX hardware acceleration?
SCO sued it's userbase, and look how far that got them...
You need to hit some better bars. :-) Start with 7B (aka Vasacs, the bar in Crocodile Dundee), or the Patriot bar (long live the Village Idiot!), or Ace bar between 5A & 5B. Or head a bit uptown and visit Yogi's off of 75th? & Broadway. Several of the above will serve a 64oz pitcher of Guiness for $9-10. Say hi to Tommy and Lilly...
Depends on where you are. Apparently, there are four states which have laws against billboards-- Maine & Vermont, are two, I believe.
I understand you hate ads and will probably spontaneously combust if you see them in a game.
I don't hate all ads, but I dislike annoying ads. I can recall from way back when, that ads seemed to be a lot less annoying ten or twenty years ago than they are today.
Well, tough, they are going to be there so it is best advertisers understand how they can do it without pissing off the majority of gamers. Some small number of people may boycott games with ads...but they aren't going to be significant enough to matter (e.g. bnetd and blizzard).
Welcome to Slashdot; this is where geeks sometimes solve problems, not just talk about them. In the case of games using Massive's ad-network, I suggest setting up your own nameserver on your LAN, and making it authoritative for madserver.net, via a zone file like:
If you like, sure. I don't consider a manufacturer's logo on or in a product to be the same as displaying third-party ads, but if you want to call it advertising, that's a reasonable position.
A better quote from the article would have been:
"Broadband households have become even harder to reach: some 81% of those with high-speed Internet access employ pop-up blockers and spam filters."
It's not surprising, either. At one point, it was commonly recognized that computers belonged to the people that owned them, and that it was the responsibility of people writing software to make sure that the software was well-behaved and did what the user told the software to do-- except for deliberately malicious software. While I do not claim that all forms of advertising are malicious, it's becoming the case that websites using lots of pop-up or pop-under ads, or software like games using Massive's technology or other in-game ad-delivery mechanisms operate under the assumption that they are free to do things with the user's computer and consume networking resources to fetch and display content that the user didn't ask for and does not want.
I can tolerate ad-bars appearing on the right-hand side of a page, so long as most of the screen real-estate shows the actual content I want, but some sites do obnoxious and deceptive things like displaying an interstital ad first. My response to that is to copy the ad link into an email, and send a complaint off to both the webmaster of the site I was on, and the site holding the advertising, indicating that their ad was so annoying that I won't be returning to the offending site for at least one week, and that obviously they will be losing my eyeballs and ad impression revenue for that period of time.
It seems to have an effect, too. At least two of the newspapers I visit (the Boston Globe & the LA Times) have toyed with interstitial ads and have dropped them soon afterwards....
Speak for yourself, not for me, OK?
Look, in case you don't realize, you can drive between San Francisco and San Jose in California on I-280, and find a half-dozen real-world references like Pacifica, Mountain View, Alpine Rd, Portorola, and others which are scenic drives you can take-- twisty, windy roads up the side of small mountain ridges-- which EA games based the tracks in much of the NeedForSpeed series on. They're fun to drive in a decent sporty car, and might be even better if you want to ride a motorbike over 'em; I go driving to enjoy the view and the outstanding scenery of the region, especially in mid-fall.
I don't drive them to watch billboards.
NASCAR/F1 racing has ads everywhere in reality. So having the ads in the game change nothing.
Of course NASCAR/F1 has ads everywhere in reality. Having the ads in real NASCAR/F1 racing means I choose not to watch or participate in NASCAR/F1 racing. Having the ads in a racing game means I won't purchase the game. But I might purchase a F1-style racer if it offered non-circuit tracks.