"terminate immediately if the Licensee or his or her employer ceased to be an Omnicare customer."
"The licensee or his or her employer" is not and never was an "Omincare customer". Therefore, they never "ceased to be an Omnicare customer", and hence, the license will not terminate due to this condition.
Also, if the license terminates, doesn't that mean that it no longer applies; you are no longer bound by it?
There are many other flaws with them. In short, they're only 80-90% reliable under the best circumstances. See the webpage. It makes for an interesting read.
As far as I see it, anonymous checkout won't work in a free library. OTOH, if you require the library patron to deposit the value of the book before they can check out the book, that would work. You might be able to decrease the deposit amount to a percentage, depending on the honesty of the patrons, but I wouldn't count on it. Put money into the equation, and dishonesty increases.
Another method would be to give a "library card voucher" to every new resident, and allow them to obtain a card with the voucher, using some sort of random hat draw or something. Have some type of card trading system in the library, where people can trade cards, reliably knowing that each card has no books on it. (Scan cards, it tells you, "No books are checked out." You then randomly decide whether or not to exchange cards.) Of course, if you lose your card, you're screwed. This method would require a bit more honesty than today's libraries. IOW, it's susceptible to many of the same no-return attacks as modern libraries, (Borrow books, and never use the card again. Not much they can do about that either way.) But the fact that it can be traced back to you seems to encourage honesty, regardless of the library enforcing returns with external mechanisms. IOW, being anonymous increases dishonesty.
The best method seems to be to delete the records of a patron's borrowed books as soon as they are returned.
Applied Cryptography has some interesting notes on this. I don't have my copy with me at the moment, but the one way to prove you came up with something:
Write up a report, "digitize" your secret.
Hash the secret (using a large cryptographically secure hash), and publish the hash in the New York Times, or whatever.
When it comes time to show your proof, reveal your secret, and point to the widely published, hash as proof of when you came up with your secret
Or you could mail a letter to yourself, with the stamp on the openable portion of the envelope, so that the postmark "proves" when you sent the message. Or even send yourself a postcard with the secret, (or a hash of the secret). Once again, you can prove when you sent the postcard via the postmark.
These two solutions assume you trust the validity of a postmark. But then again, the first solution also implies some trust in the NYT not being forged and replaced en masse at some point in the future. (Highly unlikely, but still.)
This is a trivial problem. I think there is more of an issue as to "What do you do if you discover a proof that NP complete is O(n)?"
I believe that they can't remove you from the class just by sending you some bits of paper. In order for it to work, you would actually have to use the coupons, signifying your acceptance of the terms of their "settlement".
I use WinMX on this system (PIII 450, 128 MB, T1). I notice considerable slowdown, even on a secondary connection. (On a primary connection, I can't use my computer for anything else.) I notice, however, if I disconnect from the network, I get all of my speed back, and the "leaked" memory seems to get freed.
(I share several hundred files, totalling quite a few gigabytes.)
I may have misused the terms "nonpredictable deterministic", but I don't think so. I meant that without the key, it appears random, hence nonpredictable. But deterministic because given the same input, you get the same results.
Also, I suggested a cryptographicly secure hash, because you don't want people to be able to guess whether you're forwarding the data or not.
Hence, if you get thirty queries for the same file, you would either forward all thirty, or you would forward none of them, and statistics alone wouldn't be enough to disclose the location of a file. Traffic analysis or the attacker controlling all the nodes you communicate are the attacks I see.
Also, take the following attacks:
Assuming that you hash the query to a bit, without a key, then use that bit to determine if you forward the query or not. The attacker simply performs the same hash, and knows whether the query was forwarded. Security broken.
Assuming you have a 128 bit key, and you hash the query and the 128 bit key to a bit using an insecure hash, such as XORing every bit together. You reduced the security of that 128 bit key down to 1 bit. The attacker simply needs to know whether one query was forwarded, and then the attacker can precalculate whether any query would be forwarded. Security broken, (but with quite a bit more difficulty).
Assuming you have a 128 bit key, and you hash the query and the 128 bit key together to a bit using a cryptographically secure hash. I can't see any attacks against this system. Without the key, it would appear completely random whether or not the query is forwarded.
Of couse, more cryptographically secure bits would be needed if you want to set the chance of something being forwarded to something other than 50%. Modulo arithmetic would serve nicely here.
...the node owner can always claim perfect ignorance of what's on their node. It is all kept encrypted in local stores for exactly that reason.
Ah, but with the P2P hacking bill, the (RI|MP)AA could hack into your system if they suspect you of "sharing" their copyrighted material. Statistical evidence that their copyrighted material is on your Freenet node would be proof enough for them.
Yes, you *could*, but there's a setting in later versions of FreeNet that assigns a probability of forwarding TTL=0 (or is it a probability of decrement... I forget) queries such that you could never be *sure* that it was that node that had the file.
The way to work around that is with statistics. Send 300 identical TTL 0 queries, and if 150/300 of them return something, there is a strong probability that a node one hop away is hosting said material.
The fix to that is to compute the chance of forwarding based on a cryptographic hash of the query and of a local key. In essence, you get nonpredictable deterministic behaviour.
(Note: I don't know what Freenet has done about this.)
Actually, I think it's closer to publishing a book on how to kill someone. Under the first ammendment, you can do that, as long as you don't incite people to actually commit the crime. Therein lies the legal gray area.
I had an interesting thought. If an asteroid was to hit the Earth, what would the government do? Tell everyone, and have them all panic? Or launch a secret mission to try to divert it? I'd guess the latter. Anyway, a mission to change the course of an asteroid is extremely hard to hide. So why not just say that it's a "test run"?
I think part of the problem is that under the DMCA, you can reverse engineer something to provide compatibility. However, I recall there being a statement that said you need to contact the company first, to see if they will release the protocol/whatever to you without reverse engineering. If they do, you can no longer "reverse engineer the program to provide compatibility".
Obvious implications, although I doubt even Microsoft is that devious.
I've missed you so much! How's life in Asdf treating you these days? I remember all those times we strolled along Ghjkl, listening to Classical, Opera and Pop music, purchasing movie tickets three to fives times a month, and renting movies six to eight times a month.
I work in Telecommunications now. I've made it to the job title of Clergy, where I do Engineering/R&D. I got a raise last week. I now make $30K to $39,999. And I Read It Occasionally, But Don't Subscribe.
I remember reading the editors mentioning it some time ago on another article. They said that they posted it early in the morning, so that the site wouldn't be/.ed. But I can't seem to find a reference for that now.
For some weird reason, it's rather difficult to get any meaningful results on a search for "slashdot effect" on slashdot.
The editors waited until the dead of the night (in the US) to post this story. They did this because it will mitigate the slashdot effect to some degree.
Allow webpages to: [ ] Open unrequested windows [ ] Open a link in a new window (requires restart...) [ ] Move or resize existing windows [ ] Raise or lower windows [ ] Change status bar text [X] Change images [X] Create or change cookies [X] Read cookies
On a related note, due to the inherent instabilities of a ringworld, I would suggest looking for signs of jets (or other methods of in-space propulsion) around the peripherary of the disk. That should provide significant evidence as to whether it's really a ringworld, or "just" a belt of dust, as the article indicates.
Slashdot Mirror
IANAL, but, as a programmer,...
"terminate immediately if the Licensee or his or her employer ceased to be an Omnicare customer."
"The licensee or his or her employer" is not and never was an "Omincare customer". Therefore, they never "ceased to be an Omnicare customer", and hence, the license will not terminate due to this condition.
Also, if the license terminates, doesn't that mean that it no longer applies; you are no longer bound by it?
use the same software to highlight road signs and banner-block ugly billboards with pictures of trees or background patterns by day.
But how long until billboard makers start using "road sign codes" to make their billboards stand out and grab your attention?
Lie detecters (Polygraphs) aren't difficult to fool. See http://antipolygraph.org/ for more info.
There are many other flaws with them. In short, they're only 80-90% reliable under the best circumstances. See the webpage. It makes for an interesting read.
Silly rabbit, tricks are for hookers!
As far as I see it, anonymous checkout won't work in a free library. OTOH, if you require the library patron to deposit the value of the book before they can check out the book, that would work. You might be able to decrease the deposit amount to a percentage, depending on the honesty of the patrons, but I wouldn't count on it. Put money into the equation, and dishonesty increases.
Another method would be to give a "library card voucher" to every new resident, and allow them to obtain a card with the voucher, using some sort of random hat draw or something. Have some type of card trading system in the library, where people can trade cards, reliably knowing that each card has no books on it. (Scan cards, it tells you, "No books are checked out." You then randomly decide whether or not to exchange cards.) Of course, if you lose your card, you're screwed. This method would require a bit more honesty than today's libraries. IOW, it's susceptible to many of the same no-return attacks as modern libraries, (Borrow books, and never use the card again. Not much they can do about that either way.) But the fact that it can be traced back to you seems to encourage honesty, regardless of the library enforcing returns with external mechanisms. IOW, being anonymous increases dishonesty.
The best method seems to be to delete the records of a patron's borrowed books as soon as they are returned.
That's what the "pleasure" electrode is for...
Or you could mail a letter to yourself, with the stamp on the openable portion of the envelope, so that the postmark "proves" when you sent the message. Or even send yourself a postcard with the secret, (or a hash of the secret). Once again, you can prove when you sent the postcard via the postmark.
These two solutions assume you trust the validity of a postmark. But then again, the first solution also implies some trust in the NYT not being forged and replaced en masse at some point in the future. (Highly unlikely, but still.)
This is a trivial problem. I think there is more of an issue as to "What do you do if you discover a proof that NP complete is O(n)?"
I believe that they can't remove you from the class just by sending you some bits of paper. In order for it to work, you would actually have to use the coupons, signifying your acceptance of the terms of their "settlement".
I use WinMX on this system (PIII 450, 128 MB, T1). I notice considerable slowdown, even on a secondary connection. (On a primary connection, I can't use my computer for anything else.) I notice, however, if I disconnect from the network, I get all of my speed back, and the "leaked" memory seems to get freed.
(I share several hundred files, totalling quite a few gigabytes.)
Also, I suggested a cryptographicly secure hash, because you don't want people to be able to guess whether you're forwarding the data or not.
Hence, if you get thirty queries for the same file, you would either forward all thirty, or you would forward none of them, and statistics alone wouldn't be enough to disclose the location of a file. Traffic analysis or the attacker controlling all the nodes you communicate are the attacks I see.
Also, take the following attacks:
Of couse, more cryptographically secure bits would be needed if you want to set the chance of something being forwarded to something other than 50%. Modulo arithmetic would serve nicely here.
Ah, but with the P2P hacking bill, the (RI|MP)AA could hack into your system if they suspect you of "sharing" their copyrighted material. Statistical evidence that their copyrighted material is on your Freenet node would be proof enough for them.
Yes, you *could*, but there's a setting in later versions of FreeNet that assigns a probability of forwarding TTL=0 (or is it a probability of decrement... I forget) queries such that you could never be *sure* that it was that node that had the file.
The way to work around that is with statistics. Send 300 identical TTL 0 queries, and if 150/300 of them return something, there is a strong probability that a node one hop away is hosting said material.
The fix to that is to compute the chance of forwarding based on a cryptographic hash of the query and of a local key. In essence, you get nonpredictable deterministic behaviour.
(Note: I don't know what Freenet has done about this.)
Actually, I think it's closer to publishing a book on how to kill someone. Under the first ammendment, you can do that, as long as you don't incite people to actually commit the crime. Therein lies the legal gray area.
I had an interesting thought. If an asteroid was to hit the Earth, what would the government do? Tell everyone, and have them all panic? Or launch a secret mission to try to divert it? I'd guess the latter. Anyway, a mission to change the course of an asteroid is extremely hard to hide. So why not just say that it's a "test run"?
There is a P2P network layer called GNUNet. I've studied the papers on it, and the design looks extremely solid and resilient.
I think part of the problem is that under the DMCA, you can reverse engineer something to provide compatibility. However, I recall there being a statement that said you need to contact the company first, to see if they will release the protocol/whatever to you without reverse engineering. If they do, you can no longer "reverse engineer the program to provide compatibility".
Obvious implications, although I doubt even Microsoft is that devious.
I've created a user discussion at http://ask.slashdot.org/comments.pl?sid=37691, and posted some of my experiences with the issue there. Please join me there.
John? John Smith? Is that you?!
Qwetry? Qwertyy Uiop?
I've missed you so much! How's life in Asdf treating you these days? I remember all those times we strolled along Ghjkl, listening to Classical, Opera and Pop music, purchasing movie tickets three to fives times a month, and renting movies six to eight times a month.
I work in Telecommunications now. I've made it to the job title of Clergy, where I do Engineering/R&D. I got a raise last week. I now make $30K to $39,999. And I Read It Occasionally, But Don't Subscribe.
I too have seen the Department of Social Services go after a suspected child abuser. And I know for a fact that the accusations weren't true.
Anyway, they (DSS) said, "We'd rather wrongly take a thousand children away from their parents than let one child abuser go free."
Really? I was born on 1/1/70, and I've lived in New York, NY, (ZIP code 10001) for my entire life.
True, but web bugs are more indicative of a person actually viewing the message.
I remember reading the editors mentioning it some time ago on another article. They said that they posted it early in the morning, so that the site wouldn't be /.ed. But I can't seem to find a reference for that now.
For some weird reason, it's rather difficult to get any meaningful results on a search for "slashdot effect" on slashdot.
The editors waited until the dead of the night (in the US) to post this story. They did this because it will mitigate the slashdot effect to some degree.
They're not complete morons, you know.
OK, Edit->Preference
Advanced->Scripts & Windows
Enable JavaScript for:
[X] Navigator
Allow webpages to:
[ ] Open unrequested windows
[ ] Open a link in a new window (requires restart...)
[ ] Move or resize existing windows
[ ] Raise or lower windows
[ ] Change status bar text
[X] Change images
[X] Create or change cookies
[X] Read cookies
Slashdot.org target=_blank
Wow, it works! How the hell do "they" manage to open new windows then? I guess I'll have to check the HTML source next time it happens.
(And thanks.)
Larry Niven, author of the Ringworld books, among others is hardly obscure. He also coined the term flash crowd. Google, as always, turns up a wealth of information.
On a related note, due to the inherent instabilities of a ringworld, I would suggest looking for signs of jets (or other methods of in-space propulsion) around the peripherary of the disk. That should provide significant evidence as to whether it's really a ringworld, or "just" a belt of dust, as the article indicates.
If I'm reading this correctly, we've got a nearly guaranteed winner of the Nobel Prize here.