Eco misses the whole point. The great advantage of online content is searchability. He describes using an encyclopedia in an "advanced way" to find out if
Napoleon ever met Kant.
When we query Google for that question, we immediately discover that this 2003 talk by Eco is a rehash of a talk he gave in 1995, and a very similar talk he gave in 1996, and again in 1998, and yet again in 2000 . Each of those talks contains the Napoleon/Kant/encyclopedia example. So Eco has been giving much the same talk for almost a decade now.
A search at Amazon.com reveals that Bertrand Russell compared Napoleon and Kant back in 1935, and mentioned that Kant never travelled more than 10 miles from his home town of Konigsberg, Germany. Eco has presumably read Russell, one of the great philosophers and essayists, and may have lifted the Kant/Napoleon example from Russell.
So we've learned something important about Eco himself, something he didn't tell us. He's less creative and original than he would like us to think.
Before Internet searches, it would have taken considerable scholarly research to discover that. Now, anyone can do it in a few minutes.
Google hasn't published their investor agreement, so it's not clear what rights the VCs have. But the Google founders are said to be in an unusually strong position.
No one has mentioned this, but members of Congress are immune to lawsuits like this. That's why it's useful, on major public interest issues involving documents, to get a member of Congress to put them into the record, as Kucinich has done here.
"The Senators and Representatives shall... in all cases, except treason, felony and breach of the peace, be privileged from arrest during their attendance at the session of their respective Houses, and in going to and returning from the same; and for any speech or debate in either House, they shall not be questioned in any other place." -- U.S. Constitution
Any individual member of Congress has that privilege, and historically, it's been an important one. That keeps Ashcroft from harassing members of Congress, for example.
is that Google is not only profitable but has paid back its startup costs.
Few Internet companies can say that. AOL/etc. can't. Yahoo can't. The new telecoms can't. Microsoft can.
Google as a business has no real need to go public, because they don't need cash. Going public is a dumb financial decision for them, because they'd be overpaying for money they don't need.
The founders could buy out their initial investors in a leveraged buyout and go private. That's a tough deal to set up, and the VCs would have to agree, but it's another, and perhaps a better, option.
Spamcop is an ISP under this definition
on
Who Is An ISP?
·
· Score: 2, Interesting
It looks like anti-spam services like Spamcop can use this law. It could even be a big profit center for them.
Clearly SpamCop is a service provider. No problem there.
Services like Spamcop see many copies of the same spam, so they can aggregate them up to the $1 million limitation. They see enough addresses that they can detect dictionary attacks (an "aggrivated offense"). If they use spam trap addresses, they can detect mining, another aggrevated offense.
Because SpamCop routinely informs spammers that they're spammers, they can even meet the high standard under the "special definition of procure"
that places liability on advertisers only if
they have "actual knowlege, or by consciously avoiding knowing,whether such person is engaging [in spamming]".
But none of this will stop the coming flood of "legitimate" spam. That, though, can be automatically filtered.
So what we need now is a service like SpamCop that sues spammers aggressively.
This could reduce Google's valuation for their IPO.
Google will have to put "pending litigation" with a big dollar value in their prospectus. This affects the valuation. Perhaps by billions.
What an extortion racket.
On Monday, December 5, the discovery motions in the IBM/SCO case go before the judge. That's the first "put up or shut up" event in the case.
It's another of David Chaum's overly complicated cryptographic ideas. It's got one time pads. It's got multiple-copy receipts. It's got majority-vote master key recovery. It is so l33t. But explaining it is hopeless.
I've talked to David Dill about this, He considers cryptographic solutions to be too complex for the real world. I tend to agree.
Printers are more practical. If you have two printers, and the voting machine chooses one of them randomly, the receipt can be displayed to the voter behind a window, then wound onto a takeup roll, without compromising voting anonymity.
With receipts on a roll, a recount is possible either manually or with a suitable scanner.
If you're willing to accept a sequential vote log (which has some privacy implications), just videotaping the touch screen images would give a good log. Put a VGA splitter on the line to the touch screen panel, run the output through a VGA->NTSC converter, and pipe the output into a recorder. Preferably an analog VCR, one too dumb to do anything to the video. This can be recounted by hand (slowly), or by computer means
(checkable by viewing the tape). Also, because you get to see all the user interaction, you can find out if voters seem to be having problems.
The video solution gets rid of the paper handling problem. People are comfortable with VCR technology.
QNX is a great microkernel real time OS, but it has no security features beyond those of classic UNIX.
If all interprocess connection requests (which include file, network, and screen access) went through a security monitor, QNX would have real potential as a secure OS. But it's not set up that way. In QNX, each resource manager separately enforces its own security policy. Also, resource managers have to run as root, which really shouldn't be required unless they need to access hardware directly. (A resource manager takes over a portion of the pathname space, as with the/proc file system in Linux. This really should only require permissions comparable to "mkdir")
The opt-out provision only applies to the specific company being advertised.
No. It's even narrower than that. It only applies to the specific line of business of the specific company being advertised. So one spammer can send you a Viagra spam, a mortgage-refinancing spam, an inkjet cartridge spam, a long distance spam, a cigarettes-by-mail spam, an extend-your-warranty spam, an online greeting card spam, a dating service spam, a credit card spam, a debt-consolidation spam, and a wireless video camera spam. You then have to opt out of each one separately.
It's OK that there are lots of viruses and worms, they're not doing any harm.
If people patched and upgraded their embedded systems every week like they should, this wouldn't be a problem.
Even though the system was broken into, nobody stole any money because they didn't bother.
There are other systems a crook could break into that would allow them to steal more money.
We really need a secure microkernel OS for applications like this. There's nothing available. Windows CE is a mess. Linux and the UNIX variants are too bloated. QNX isn't designed to be secure. NSA Secure Linux has no applications. The Hurd crowd can't get their act together. And Multics is dead.
Here's an
old Slashdot page in the Internet archive. Decoding the URL http://web.archive.org/web/20000301205131/http://w ww.slashdot.org/ is straightforward. It's just the archiving site ("web.archive.org"), the medium being archived ("web"), the date and time, and the original URL being archived.
There's another copy of the archive at "archive.bibalex.org", in Egypt. Brewster Kale wants to have four copies worldwide; then, he thinks, the information will be safe.
One problem with the Internet Archive is that the server farm is unreliable. Sections of the archive drop offline for days at a time. It's built out of thousands of commodity PCs sitting on shelves in a building in San Francisco.
Another problem is that web sites that are too complex don't get archived properly. If there are links embedded in JavaScript, Java, or Flash, they won't be properly adjusted to the appropriate archive references. This becomes more of a problem as more pages are created with overly complex authoring tools.
At last, the intercontinental torpedo. This is going to go over big with terrorists. Or small countries that need some effective deterrent against US attack.
The next step in weaponization is a torpedo powerplant and seeker. This would be used only in the last stage, when wave motion has brought the thing to a harbor mouth, allowing a final attack run with power. The thing can be launched hundreds of miles offshore. Maybe thousands.
It's back to submarine nets, like WWII. SOSUS isn't going to pick this up; it's just drifting sea junk most of the time.
Re:What do you make it out of?
on
Son of Concorde
·
· Score: 1
I've seen that guy's name on SBIR applications before.
There was some work on doing something like that with liquid hydrogen in the early 1960s. It turned out not to be necessary for the SR-71, but it was considered.
The SR-71 circulates its fuel internally as a coolant before it burns it. This works because it has a fuel that has very low volatility.
What do you make it out of?
on
Son of Concorde
·
· Score: 2, Interesting
Remember Reagan's "National Aerospace Plane" from the 1980s? Same idea. Same problem.
Ben Rich, head of Lockheed's Skunk Works and propulsion designer on the SR-71, refused to bid on that idea. "We used titanium. You know anything stronger?" The SR-71 was speed-limited by the melting point of its skin. More power could have been added, but woudn't help. Just cooling the pilot was a major effort. Cooling a big passenger cabin would be really tough.
Ceramics? Maybe someday, but they're brittle, like the Space Shuttle tiles.
At last, the ultimate weapon against the RIAA
on
Synthesized Singers
·
· Score: 4, Informative
U.S. copyright law provides a "compulsory mechanical license" option, allowing anyone to record a "cover" version of a song for a fixed statutory royalty of $0.0155/minute per copy. The RIAA likes this, because they can re-record some oldie using a new band cheaply.
But this works for anybody. If you can synthesize music from MIDI and vocal models, you can use that deal. The RIAA can't stop you from doing this.
A synthesized music web site could even buy blanket ASCAP and BMI licenses, which aren't too expensive, and allow music downloads. The going rate seems to be about $5000 per million downloads, or about $0.005 per song.
This is a real threat to the RIAA. If the technology works.
Sun is going the way of DEC and SGI. Nobody can quite figure out why anybody really needs them any more. HP is basically a printer company with some sideline businesses. Cisco is competing with mountains of used Cisco gear left over from dead dot-coms. Intel doesn't have any hot new products in the pipe. (There was supposed to be a Pentium 5 this year, but due to lack of interest, it was cancelled.)
None of those companies are growing. They're all shrinking.
This "hoteling" stuff looks just like downsizing.
A few years back, at the Game Developer's Conference, there was a woman who plays Quake for a living. She had a one year contract with some game magazine to play Quake.
She was taking on all comers, fragging them in about 30 seconds each, and had a score of about 40-1 in early afternoon.
Some cops use the term "driving while stupid". Usually this means drugs or booze, but any officer who works traffic can tell you about wierd stuff they've found people doing while driving badly.
He was in California. If Congress hadn't legalized spam at 6 AM this morning, he could have sued the crap out of them after January 1. He could have sued them anyway, and tried out the "long arm" feature of California's old anti-spam law, even though the penalties were small.
Here's what's going to happen after February 21:
You're going to get tons of spam, and from major companies. This becomes legal, even in states where it used to be illegal.
The headers will be correct. There are penalties for forging headers.
The spam won't necessarily have the company name, just some unsubscribe URL and a P.O. box for written "opt-out" requests.
You can go through the motions of "opting out", but it won't do much.
"Opt-out" is interpreted narrowly, on a "per sender" basis. "Sender" is defined narrowly - "The term `sender', when used with respect to a commercial electronic mail message, means a person who initiates such a message and whose product, service, or Internet web site is advertised or promoted by the message." (from S.877) Note the "and"; it's not there by accident. Each combination of spammer and advertiser may be considered a different "sender". That clause could even be interpreted to completely let third-party spammers off the hook.
So advertisers get to throw away the opt-out list every time they change spamhauses. There's even a "separate line of business" exception to make this explicit - spammers with both "Viagra" and "refinancing" spams don't have to use the opt-out list from one with the other.
You can't sue. Only the FTC and the U.S. Justice Department can sue.
This was all carefully crafted by lobbyists for the Direct Marketing Organization, who will be celebrating as soon as they get some sleep, having been up all night getting this through the House.
As of 3:11 AM EDT, the spam bill, S.877, still hasn't come to a vote. The House is still in session.
There's a big fight going on over the Medicare prescription drug benefit bill, and the House is dealing with procedural votes on that.
Saturday ends the session. If it doesn't pass today, it's dead.
Slashdot Mirror
When we query Google for that question, we immediately discover that this 2003 talk by Eco is a rehash of a talk he gave in 1995, and a very similar talk he gave in 1996, and again in 1998, and yet again in 2000 . Each of those talks contains the Napoleon/Kant/encyclopedia example. So Eco has been giving much the same talk for almost a decade now.
A search at Amazon.com reveals that Bertrand Russell compared Napoleon and Kant back in 1935, and mentioned that Kant never travelled more than 10 miles from his home town of Konigsberg, Germany. Eco has presumably read Russell, one of the great philosophers and essayists, and may have lifted the Kant/Napoleon example from Russell.
So we've learned something important about Eco himself, something he didn't tell us. He's less creative and original than he would like us to think. Before Internet searches, it would have taken considerable scholarly research to discover that. Now, anyone can do it in a few minutes.
Google hasn't published their investor agreement, so it's not clear what rights the VCs have. But the Google founders are said to be in an unusually strong position.
Any individual member of Congress has that privilege, and historically, it's been an important one. That keeps Ashcroft from harassing members of Congress, for example.
Few Internet companies can say that. AOL/etc. can't. Yahoo can't. The new telecoms can't. Microsoft can.
Google as a business has no real need to go public, because they don't need cash. Going public is a dumb financial decision for them, because they'd be overpaying for money they don't need.
The founders could buy out their initial investors in a leveraged buyout and go private. That's a tough deal to set up, and the VCs would have to agree, but it's another, and perhaps a better, option.
Clearly SpamCop is a service provider. No problem there.
Services like Spamcop see many copies of the same spam, so they can aggregate them up to the $1 million limitation. They see enough addresses that they can detect dictionary attacks (an "aggrivated offense"). If they use spam trap addresses, they can detect mining, another aggrevated offense.
Because SpamCop routinely informs spammers that they're spammers, they can even meet the high standard under the "special definition of procure" that places liability on advertisers only if they have "actual knowlege, or by consciously avoiding knowing,whether such person is engaging [in spamming]".
But none of this will stop the coming flood of "legitimate" spam. That, though, can be automatically filtered.
So what we need now is a service like SpamCop that sues spammers aggressively.
What an extortion racket.
On Monday, December 5, the discovery motions in the IBM/SCO case go before the judge. That's the first "put up or shut up" event in the case.
They don't even pay him that much. He only got $14,500 from the whole entertainment industry. His big contributors are from the insurance industry.
I've talked to David Dill about this, He considers cryptographic solutions to be too complex for the real world. I tend to agree.
Printers are more practical. If you have two printers, and the voting machine chooses one of them randomly, the receipt can be displayed to the voter behind a window, then wound onto a takeup roll, without compromising voting anonymity. With receipts on a roll, a recount is possible either manually or with a suitable scanner.
If you're willing to accept a sequential vote log (which has some privacy implications), just videotaping the touch screen images would give a good log. Put a VGA splitter on the line to the touch screen panel, run the output through a VGA->NTSC converter, and pipe the output into a recorder. Preferably an analog VCR, one too dumb to do anything to the video. This can be recounted by hand (slowly), or by computer means (checkable by viewing the tape). Also, because you get to see all the user interaction, you can find out if voters seem to be having problems.
The video solution gets rid of the paper handling problem. People are comfortable with VCR technology.
QNX is a great microkernel real time OS, but it has no security features beyond those of classic UNIX. If all interprocess connection requests (which include file, network, and screen access) went through a security monitor, QNX would have real potential as a secure OS. But it's not set up that way. In QNX, each resource manager separately enforces its own security policy. Also, resource managers have to run as root, which really shouldn't be required unless they need to access hardware directly. (A resource manager takes over a portion of the pathname space, as with the /proc file system in Linux. This really should only require permissions comparable to "mkdir")
No. It's even narrower than that. It only applies to the specific line of business of the specific company being advertised. So one spammer can send you a Viagra spam, a mortgage-refinancing spam, an inkjet cartridge spam, a long distance spam, a cigarettes-by-mail spam, an extend-your-warranty spam, an online greeting card spam, a dating service spam, a credit card spam, a debt-consolidation spam, and a wireless video camera spam. You then have to opt out of each one separately.
We really need a secure microkernel OS for applications like this. There's nothing available. Windows CE is a mess. Linux and the UNIX variants are too bloated. QNX isn't designed to be secure. NSA Secure Linux has no applications. The Hurd crowd can't get their act together. And Multics is dead.
SCO is getting close to the first "put up or shut up" point. .
There's another copy of the archive at "archive.bibalex.org", in Egypt. Brewster Kale wants to have four copies worldwide; then, he thinks, the information will be safe.
One problem with the Internet Archive is that the server farm is unreliable. Sections of the archive drop offline for days at a time. It's built out of thousands of commodity PCs sitting on shelves in a building in San Francisco.
Another problem is that web sites that are too complex don't get archived properly. If there are links embedded in JavaScript, Java, or Flash, they won't be properly adjusted to the appropriate archive references. This becomes more of a problem as more pages are created with overly complex authoring tools.
The next step in weaponization is a torpedo powerplant and seeker. This would be used only in the last stage, when wave motion has brought the thing to a harbor mouth, allowing a final attack run with power. The thing can be launched hundreds of miles offshore. Maybe thousands.
It's back to submarine nets, like WWII. SOSUS isn't going to pick this up; it's just drifting sea junk most of the time.
There was some work on doing something like that with liquid hydrogen in the early 1960s. It turned out not to be necessary for the SR-71, but it was considered.
The SR-71 circulates its fuel internally as a coolant before it burns it. This works because it has a fuel that has very low volatility.
Ben Rich, head of Lockheed's Skunk Works and propulsion designer on the SR-71, refused to bid on that idea. "We used titanium. You know anything stronger?" The SR-71 was speed-limited by the melting point of its skin. More power could have been added, but woudn't help. Just cooling the pilot was a major effort. Cooling a big passenger cabin would be really tough.
Ceramics? Maybe someday, but they're brittle, like the Space Shuttle tiles.
But this works for anybody. If you can synthesize music from MIDI and vocal models, you can use that deal. The RIAA can't stop you from doing this.
A synthesized music web site could even buy blanket ASCAP and BMI licenses, which aren't too expensive, and allow music downloads. The going rate seems to be about $5000 per million downloads, or about $0.005 per song.
This is a real threat to the RIAA. If the technology works.
Just think. Back when that commercial was made, there was no nationwide wiretapping. No Patriot Act. No spam. No DRM.
None of those companies are growing. They're all shrinking. This "hoteling" stuff looks just like downsizing.
A few years back, at the Game Developer's Conference, there was a woman who plays Quake for a living. She had a one year contract with some game magazine to play Quake. She was taking on all comers, fragging them in about 30 seconds each, and had a score of about 40-1 in early afternoon.
Of course, that's a risk to Microsoft's customers, so that may not be considered as critical.
This definitely qualifies.
Here's what's going to happen after February 21:
-
You're going to get tons of spam, and from major companies. This becomes legal, even in states where it used to be illegal.
-
The headers will be correct. There are penalties for forging headers.
-
The spam won't necessarily have the company name, just some unsubscribe URL and a P.O. box for written "opt-out" requests.
-
You can go through the motions of "opting out", but it won't do much.
"Opt-out" is interpreted narrowly, on a "per sender" basis. "Sender" is defined narrowly - "The term `sender', when used with respect to a commercial electronic mail message, means a person who initiates such a message and whose product, service, or Internet web site is advertised or promoted by the message." (from S.877) Note the "and"; it's not there by accident. Each combination of spammer and advertiser may be considered a different "sender". That clause could even be interpreted to completely let third-party spammers off the hook.
So advertisers get to throw away the opt-out list every time they change spamhauses. There's even a "separate line of business" exception to make this explicit - spammers with both "Viagra" and "refinancing" spams don't have to use the opt-out list from one with the other.
-
You can't sue. Only the FTC and the U.S. Justice Department can sue.
This was all carefully crafted by lobbyists for the Direct Marketing Organization, who will be celebrating as soon as they get some sleep, having been up all night getting this through the House.A bit of good news, though. It doesn't take effect for 120 days. For almost three months, the California law will be in effect.
Saturday ends the session. If it doesn't pass today, it's dead.