As described, assuming my reading of the GPL is accurate, this may not actually be a GPL violation...
It is perfectly legal under the GPL to distribute GPL binaries, as long as they are compiled from publically-available source code. So selling a box that runs on the Linux kernel and the Squid daemon is not a GPL violation, even if other non-GPL software runs on the box as well.
The requirement to produce source code only kicks in if it is modified by the developer. For example, if Viglen sold its box with a "hardened" Linux kernel, for example, and did not release a diff of the "hardening" source, then they would be violating the GPL.
From what I remember when I got my first PowerPC-based Mac, there was only one lage incompatibility, but it was a real humdinger - the 68K emulator did not support 68K FPU code. Any 68K app that required a FPU did not run (none of my 3D rendering apps worked). Someone wrote an extension to "render" 68K FPU instructions to 68K integer instructions, which allowed these apps to run, but the performance sucked hard.
Probably the most amusing incident from the migration was Connectix's SpeedDoubler application, which replaced Apple's 68K emulator with its own version that ran twice as fast as Apple's. Apple implemented Connectix's emulator, abanonding their own, in Mac OS 7.6, IIRC.
I'll agree with the shout-out to Andy Wallace - he mixed Front 242's "06:21:03:11 Up Evil" album, which, up until NIN's "The Downward Spiral" was released a year later, was probably the best-sounding industrial rock album to date (Fire up Kazaa and download "Crapage" and "Hymn" if you need an example).
Also, you are correct about Vig - he wasn't a "name" producer at the time "Nevermind" was released - it was that album that put him on the map.
If someone used these documents to hack P4, then the "other's economic gain" clause of the law will have been realized by the users of hacked smartcards (because they won't have to pay the monthly bill) and the people who sell hacked smartcards for profit.
So, there are potential economic beneficiaries of the theft of trade secrets. The big question is whether stealing for someone's *potential* economic benefit is good enough.
What's interesting is that the prosecutors may wind up having to prove that the documents can indeed be used to potentially devise a hack for P4. And by doing so, they could wind up putting enough information in the public record (testimony of expert witnesses, et al) to actually enable a crack to be devised. I wonder if the US Government is immune from its own law...
Re:$5 to anyone who proves this statement wrong-
on
The Economics of Spam
·
· Score: 1
Um, did you see the woman's photo on the page? Chances are she isn't going to be terribly successful at that endeavor...
That is incorrect - the slave computers will send the DoS traffic to themselves, not back to the master. The master will just sit there wondering why his target is still live.
"Service Location Restriction: Each circuit may only service a single location (residence, apartment, office, or place of business). If you connect your circuit to a neighboring premise without specific permission from Speakeasy, your service may be disconnected or terminated at our sole discretion."
I'm presuming this includes all forms of reselling, including noncommercial/coop connection sharing.
I'm wondering how they implemented smartcard authentication: Wishful thinking says that they were dumb enough to build the smartcard driver into the kernel, thereby legally obligating them to release the source code, thereby declaring open season for smartcard hackers...
Honestly, they most likely did it as a kernel module (which doesn't need to be GPL'ed; see Nvidia). Oh well.
Point #1 sucks, but is more or less a non-issue, because it happens to be the status quo in most areas.
Point #2 is the scariest from my perspective - if this is implemented, Covad, and probably most of the independent ISPs that use them, are out of business. Unless the LECs are willing to resell (don't bet on it), my company, a medium-size carrier-class ISP, will see one or more of its largest billing customers get regulated out of business. Lots of people will lose their jobs if this becomes reality.
Point #3 is scary, too, but unless the FCC takes back the 2.5 GHz band, unlikely, unless that band becomes too crowded to be usable for IP traffic.
The largest logical failure in the FCC's line of thinking is the assumption that all three platforms will be available in all areas. Cable has shown the most promise for wide-scale deployment, but is hampered by the lack of competition that begets "because we can" policies (no servers or VPNs allowed, port blocking, etc.). Why? because unless DSL is available where you live, the cable companies truly have a monopoly. If the same thing happens to DSL, something tells me that the idea of running a server on a residential broadband line will become a faint memory.
My hope the the future is wireless - there's no way that any one wireless cellphone provider will have a monopoly of any given area, and hopefully the same will be true with data. I just hope those providers will allow servers.:)
1. I'm inferring that anyone who can point a dish at one of the towers can get a wireless signal. How is IP allocation, et al handled? Can one connect the dish to an ethernet hub and connect to multiple machines, or would NAT be required?
2. How much does the CPE gear cost? What percentage of the homes in the area have a clear line of sight to a tower?
3. Do the users get public or private IP addresses?
4. How will abuse issues be handled? I didn't see an AUP at first glance...
As a side note, I recently went from a wired network connection to a WLAN PCI card (The DSL came to my phone jack in my upstairs office, but Verizon delivered the T1 to the basement). I normally run my machine "naked", and discovered that in order to get a decent wireless signal, I had no choice but to replace the tower case cover.
Yes, but hydrogen atoms are the simples atoms in existence (1 proton, 1 neutron, 1 electron). As such, it should be the easiest type of antimatter to create. Manufacturing larger anti-atoms might now prove to be worth the effort - an anti-helium atom reaction will generate only 2x the energy of an anti-hydrogen atom, and chances are it will cost more to generate a singel anti-helium atom than two anti-hydrogen atoms.
IT's "quick start" feature saves the contents of RAM to disk, just like XP's Hibernate function. When you start it up, the system is just how you left it, apps and all.
A good folloup test would be to do the same experiment, but with the addition of a speaker that emitted a "buzz" that sounds like the humming of a transformer, or some other stimuli that the subject should interpret as a sign of existence of the flowing current - nothing too overt; a "power on" light in the room might be a interpreted as easily spoofable. The key is to make the subject believe that the juice is flowing independent of whether or not it actually is.
Turn said stimuli on and off in a pattern completely unrelated to the actual activation/deactivation of the current. See if the subject reponds to the stimuli pattern instead of the actual current flow. I'll betcha they do.
At the consumer level, there's no need for Firewire audio devices as USB provides ample bandwidth. But if you look at pro gear, you'll need more than that...
Mark Of The Unicorn sells the 828 and 896, two Firewire/1394 devices for audio I/O. They support 8 uncompressed audio channels in and out simultaneously (USB's current data rate can't come close to the bandwidth necessary to do this), with room to add multiple units to the same Firewire bus.
The 828 is the sole reason why my next audio workstation will be a Powerbook and not a desktop G4.
You obviously had better luck than I did...I've got an Akamai streaming installation in my building, and my packet loss was in the 25 - 50% for most of the keynote, with the predictable loss of image quality making the video practically useless.
Slashdot Mirror
As described, assuming my reading of the GPL is accurate, this may not actually be a GPL violation...
It is perfectly legal under the GPL to distribute GPL binaries, as long as they are compiled from publically-available source code. So selling a box that runs on the Linux kernel and the Squid daemon is not a GPL violation, even if other non-GPL software runs on the box as well.
The requirement to produce source code only kicks in if it is modified by the developer. For example, if Viglen sold its box with a "hardened" Linux kernel, for example, and did not release a diff of the "hardening" source, then they would be violating the GPL.
Does that sound right?
From what I remember when I got my first PowerPC-based Mac, there was only one lage incompatibility, but it was a real humdinger - the 68K emulator did not support 68K FPU code. Any 68K app that required a FPU did not run (none of my 3D rendering apps worked). Someone wrote an extension to "render" 68K FPU instructions to 68K integer instructions, which allowed these apps to run, but the performance sucked hard.
Probably the most amusing incident from the migration was Connectix's SpeedDoubler application, which replaced Apple's 68K emulator with its own version that ran twice as fast as Apple's. Apple implemented Connectix's emulator, abanonding their own, in Mac OS 7.6, IIRC.
Something that should be mentioned is that the closest competition to Pro Tools, Logic Audio, was acquired by Apple last year.
I'll agree with the shout-out to Andy Wallace - he mixed Front 242's "06:21:03:11 Up Evil" album, which, up until NIN's "The Downward Spiral" was released a year later, was probably the best-sounding industrial rock album to date (Fire up Kazaa and download "Crapage" and "Hymn" if you need an example).
Also, you are correct about Vig - he wasn't a "name" producer at the time "Nevermind" was released - it was that album that put him on the map.
If someone used these documents to hack P4, then the "other's economic gain" clause of the law will have been realized by the users of hacked smartcards (because they won't have to pay the monthly bill) and the people who sell hacked smartcards for profit.
So, there are potential economic beneficiaries of the theft of trade secrets. The big question is whether stealing for someone's *potential* economic benefit is good enough.
What's interesting is that the prosecutors may wind up having to prove that the documents can indeed be used to potentially devise a hack for P4. And by doing so, they could wind up putting enough information in the public record (testimony of expert witnesses, et al) to actually enable a crack to be devised. I wonder if the US Government is immune from its own law...
Um, did you see the woman's photo on the page? Chances are she isn't going to be terribly successful at that endeavor...
That is incorrect - the slave computers will send the DoS traffic to themselves, not back to the master. The master will just sit there wondering why his target is still live.
To translate a film from American English to U.K English? What do they do, overdub "F*cking" with "bloddy" every time it's uttered?
Um, reread that TOS:
"Service Location Restriction: Each circuit may only service a single location (residence, apartment, office, or place of business). If you connect your circuit to a neighboring premise without specific permission from Speakeasy, your service may be disconnected or terminated at our sole discretion."
I'm presuming this includes all forms of reselling, including noncommercial/coop connection sharing.
-Chris
To be more accurate, you can either download the patch using Apple Software Update or download it manually, check the sig, and install.
I'm wondering how they implemented smartcard authentication: Wishful thinking says that they were dumb enough to build the smartcard driver into the kernel, thereby legally obligating them to release the source code, thereby declaring open season for smartcard hackers...
Honestly, they most likely did it as a kernel module (which doesn't need to be GPL'ed; see Nvidia). Oh well.
Point #1 sucks, but is more or less a non-issue, because it happens to be the status quo in most areas.
:)
Point #2 is the scariest from my perspective - if this is implemented, Covad, and probably most of the independent ISPs that use them, are out of business. Unless the LECs are willing to resell (don't bet on it), my company, a medium-size carrier-class ISP, will see one or more of its largest billing customers get regulated out of business. Lots of people will lose their jobs if this becomes reality.
Point #3 is scary, too, but unless the FCC takes back the 2.5 GHz band, unlikely, unless that band becomes too crowded to be usable for IP traffic.
The largest logical failure in the FCC's line of thinking is the assumption that all three platforms will be available in all areas. Cable has shown the most promise for wide-scale deployment, but is hampered by the lack of competition that begets "because we can" policies (no servers or VPNs allowed, port blocking, etc.). Why? because unless DSL is available where you live, the cable companies truly have a monopoly. If the same thing happens to DSL, something tells me that the idea of running a server on a residential broadband line will become a faint memory.
My hope the the future is wireless - there's no way that any one wireless cellphone provider will have a monopoly of any given area, and hopefully the same will be true with data. I just hope those providers will allow servers.
Large conglomerate shoots self in foot to hurt competition. Film at 11.
Luckily, all my credit cards are Visa and Amex.
Were these words used at all in his testimony?
1. I'm inferring that anyone who can point a dish at one of the towers can get a wireless signal. How is IP allocation, et al handled? Can one connect the dish to an ethernet hub and connect to multiple machines, or would NAT be required?
2. How much does the CPE gear cost? What percentage of the homes in the area have a clear line of sight to a tower?
3. Do the users get public or private IP addresses?
4. How will abuse issues be handled? I didn't see an AUP at first glance...
As a side note, I recently went from a wired network connection to a WLAN PCI card (The DSL came to my phone jack in my upstairs office, but Verizon delivered the T1 to the basement). I normally run my machine "naked", and discovered that in order to get a decent wireless signal, I had no choice but to replace the tower case cover.
Oops...I meant to say, "Manufacturing larger anti-atoms might NOT prove to be worth the effort." Must...slow...down...typing. :/
Yes, but hydrogen atoms are the simples atoms in existence (1 proton, 1 neutron, 1 electron). As such, it should be the easiest type of antimatter to create. Manufacturing larger anti-atoms might now prove to be worth the effort - an anti-helium atom reaction will generate only 2x the energy of an anti-hydrogen atom, and chances are it will cost more to generate a singel anti-helium atom than two anti-hydrogen atoms.
And this is different from every other type of public discourse HOW?
IT's "quick start" feature saves the contents of RAM to disk, just like XP's Hibernate function. When you start it up, the system is just how you left it, apps and all.
IN the case of Q3A, all Loki handled was distribution. Id did the Linux port themselves.
A good folloup test would be to do the same experiment, but with the addition of a speaker that emitted a "buzz" that sounds like the humming of a transformer, or some other stimuli that the subject should interpret as a sign of existence of the flowing current - nothing too overt; a "power on" light in the room might be a interpreted as easily spoofable. The key is to make the subject believe that the juice is flowing independent of whether or not it actually is.
Turn said stimuli on and off in a pattern completely unrelated to the actual activation/deactivation of the current. See if the subject reponds to the stimuli pattern instead of the actual current flow. I'll betcha they do.
Nothing substantial, just a bunch of graphics and a URL: http://www.tronkillerapp.com. Not sure if this is the sequel or a computer game...
At the consumer level, there's no need for Firewire audio devices as USB provides ample bandwidth. But if you look at pro gear, you'll need more than that...
Mark Of The Unicorn sells the 828 and 896, two Firewire/1394 devices for audio I/O. They support 8 uncompressed audio channels in and out simultaneously (USB's current data rate can't come close to the bandwidth necessary to do this), with room to add multiple units to the same Firewire bus.
The 828 is the sole reason why my next audio workstation will be a Powerbook and not a desktop G4.
You obviously had better luck than I did...I've got an Akamai streaming installation in my building, and my packet loss was in the 25 - 50% for most of the keynote, with the predictable loss of image quality making the video practically useless.