As other posters have pointed out, this product has all signs of snake oil. Similar announcements are made every other day. I fail to see how this justifies an article on Slashdot?!
IIRC the United Nations tried something similar a few years ago (jamming cell-phone signals in the conference rooms). Unfortunately, some of their members dug out a study which showed that the jam signals could cause cancer and infertility (unlike cell phone signals, of course ). So, they had to stop it.
So do I. I consider them "the usual spam", nothing more, nothing less. I also get fake "Reply to your question" spam. Does that mean we have to outlaw all Reply buttons in e-mail clients?
I agree with you in that the system - like so many others - can be exploited. The problem is that forbidding it does (in my opinion) more harm than good. What we have to do is go against spam and the spammers, not shut down the channels they (might) abuse.
Don't mix up the medium with the message. Don't shoot the messenger.
Okay, some people don't seem to "get" the problem stated in the article, so just for clarification:
The newsletter operator used the standard procedure: Subscribe on the website, get a confirmation mail, reply to the mail. In the court's opinion, the problem is: Someone signs up for you, you get a confirmation mail you didn't ask for, so this is spam, so this is illegal. The only way to circumvent this would probably be digital signatures used during subscription.
By the way: Yes, this decision is also considered crazy among German geeks.
After companies offering e-cards, now senders of online newsletter could face extinction. In the opinion of the Berlin regional court, the unsolicited sending of a newsletter subscription by e-mail is an illegal advertisement.
The applicant for the decision from September 19th, 2002, had received an e-mail, in which he was asked to click an activation link in order to be added to a newsletter mailing list. If he did not wish to be added, he should just delete the mail. The applicant considered this UCE and requested a cease & desist against the operator of the information service.
The court confirmed in its decision again the current public opinion that the unsolicited sending of an e-mail with commercial contents constitutes an illegal interference with the business of companies receiving them. Private persons also have a right to be spared from such mails as stated in 1004, 823 sect. 1 of German Civil Law.
The newsletter operator's objection that the applicant had signed in for the mailing list himself was not accepted by the court. In its opinion, the operator must prove that the applicant signed in personally. This couldn't be proved by the provider. The decision is seen controversially among jurists. The opt-in method for newsletters the decision is based on is used widely throughout the internet and was considered legally unobjectionable up to now.
If they know the exact bit to twiddle, it shouldn't be too hard to write a program which opens the file, flips the bit, and closes the file again (a one-liner, if you start sufficiently far to the left:-) ). Or am I missing something?
It should be anonymous. PayPal isn't. Credit cards aren't. So they can create a customer profile of you. Goodbye privacy.
It should be secure. Credit card numbers can be stolen or faked. An "ideal" system would guarantee security.
It should be flexible. Most existing anonymous and secure e-cash systems use "tokens" of a fixed value. A good system should be able to handle arbitrary cash values in a flexible way.
Systems that work under these premises do exist. Anonymity and security can be reached by using blind electronic signatures. Flexibility can be reached by using "divisible" e-cash systems. Unfortenately, such schemes are very expensive in terms of computing power (which means, of course, that they are also expensive in terms of transaction costs).
And then, of course, there's one important question: Who wants anonymous e-cash? Banks and credit companies probably don't (because they like to have your customer profile). Shops probably don't either (for the same reason). Customers? Well, face the facts: Most customers just don't appreciate the value of privacy. So, there's a simple conclusion: No market, so e-cash. It's as easy as that.
Many copy protection systems work by writing multisession CDs. The first session contains the usual audio data, so normal CD players (which don't support multiple sessions) will play the CDs correctly.
The second (third,...) session contains bogus data, TOC entries that are outside the physical CD, (insert your favorite way to irritate CD drives here). So drives trying to read those sessions will hang up, think the CD is damaged or whatever (depending on their firmware).
Now, when you black out the outer regions (containing the later sessions), the drive can't find them and only uses the data in the first session, making the CD appear perfectly normal. As for "how wide the marking has to be", this depends on the size of the sessions. It's a question of fractions of millimeters, so it's a little hard, but manageable. The German IT magazine c't demonstrated it about a year ago.
This description may be a little simplified, but at least it's the basic principle.
Here in Germany, junk mails (e-mails, that is) are considered the same as junk faxes legally. It doesn't help much, though: Either you can't trace the spammers, or they're sitting in some obscure Caribbean country where your legislation has no power.
They don't have (or need) dedicated telescope access, they just take the data the telescope collects anyway and process it. So they don't have to pay for telescope access, the only cost is for shipping the recorded tapes to Berkeley (that's right, no broadband internet connection in Arecibo...).
In fact, depending on the Latin nominative of "Google", this should be...
[from "Googlus" or "Googlum"] In Googlo non est (most probable solution is "Googlum", I think)
[from "Googla"] In Googla non est (if you prefer a female Google)
[from "Googlis" or "Googles"] In Google / Googli non est (if you're a fan of the more exotic declination classes)
Unless, of course, your version of the Internet contains several Googles... (Alright, there are google.com, google.de, google.fr,..., Google Groups...).
An opt-in policy will result in better privacy, but is it practical? The aim of the whole thing is to create a rather complete directory. If you consider how many student are too lazy to answer and don't care about whether they're included or not, an "opt-in directory" will probably only contain a very small percentage of the students. Opt-out, on the other hand, just removes the ones who are really concerned about their privacy, for the price of a little work on their side, resulting in a directory as complete as possible.
In fact, this has been tried already. It seems, however, that they get scratched while in use, and thus lose their water-repelling properties. It's kind of cool, though... I saw a video of honey dripping off a spoon; the spoon was completely clean afterwards.
Well, designing an algorithm for true randomness is impossible. An algorithm, by definition, must be deterministic, which means that if it doesn't take external input, it's not truly random (because either it will generate the same numbers over and over again or a series of pseudo-random numbers). If it takes external input, on the other hand, this must come from somewhere. You can't take another algorithm (see recursion), so it has to be something mechanical. QED.
Give your managers a copy of Tom DeMarco's Peopleware. And include his novel The Deadline for a bit of light reading.:-)
Some books on Extreme Programming might help, too. Even if you do not plan to use it, they show how to share responsibilities between management and programmers. It all boils down to:
Management decides what will be done (business value blah blah).
The programmers decide how long it's gonna take
Management sets the priorities.
This helps to avoid impossible deadlines and to keep up with high quality.
Try some books on Compiler Design, especially the first chapters about analysis. They should provide lots of information about parsing. One good book is Modern Compiler Design.
Book about Algorithms & Data Structures
on
General IT Books?
·
· Score: 1
What this list is missing is a good book about Algorithms & Data Structures - what most software development is all about. Several comments have suggested Knuth's The Art of Computer Programming; however, for getting started, a book like Introduction to Algorithms by Cormen, Leiserson and Rivest (MIT Press) seems more appropriate.
I disagree; take a look at other industries. Some of the highest-quality products are produced by the tiny, niche-market manufacturers.
True in the material world, but also true in the software industry; it just depends on the type of software. There isn't a market for high-quality word processing software (as long as you don't count DTP), but there's definitely a market for HQ vs. "consumer level" databases (read Access vs. Oracle / DB2), rendering/image editor (read Windows Paint vs.... vs. Maya) etc.
In those areas, there is a clear distinction between cheap, mass-market products and high-quality niche products. You get what you're ready to pay for.
On the other hand, on the OS market, the highest quality is available for free...
You could make some money offering training classes on how to avoid common viruses.
Could they? Think a moment about who decides whether employees will attend such classes - that's right, PHBs who (usually) don't have a single clue about computer security or the importance of those classes. Sigh...
Terry Pratchett, Ian Stewart and Jack Cohen just published the second part of their Science of Discworld science book / fantasy novel mix. Unlike the first part, it focuses on the evolution of the human species and on psychology. A worthwile read - and if you don't know the first part yet, get it, too, and read it first.
Slashdot Mirror
So what? Block google.com with NoScript, add mail.google.com (and, while you're at it, maps.google.com) to your whitelist, et voila.
So, how about MSN search viewed in Opera?
As other posters have pointed out, this product has all signs of snake oil. Similar announcements are made every other day. I fail to see how this justifies an article on Slashdot?!
So, the owner of a hypothetical site www.ogeefy.com could be sued by Disney, and they'd win?! I can't believe that.
IIRC the United Nations tried something similar a few years ago (jamming cell-phone signals in the conference rooms). Unfortunately, some of their members dug out a study which showed that the jam signals could cause cancer and infertility (unlike cell phone signals, of course ). So, they had to stop it.
So do I. I consider them "the usual spam", nothing more, nothing less. I also get fake "Reply to your question" spam. Does that mean we have to outlaw all Reply buttons in e-mail clients?
I agree with you in that the system - like so many others - can be exploited. The problem is that forbidding it does (in my opinion) more harm than good. What we have to do is go against spam and the spammers, not shut down the channels they (might) abuse.
Don't mix up the medium with the message. Don't shoot the messenger.
As was already mentioned in another thread, e-mail headers can be faked ad libitum, so none the better...
The newsletter operator used the standard procedure: Subscribe on the website, get a confirmation mail, reply to the mail. In the court's opinion, the problem is: Someone signs up for you, you get a confirmation mail you didn't ask for, so this is spam, so this is illegal. The only way to circumvent this would probably be digital signatures used during subscription.
By the way: Yes, this decision is also considered crazy among German geeks.
After companies offering e-cards, now senders of online newsletter could face extinction. In the opinion of the Berlin regional court, the unsolicited sending of a newsletter subscription by e-mail is an illegal advertisement.
The applicant for the decision from September 19th, 2002, had received an e-mail, in which he was asked to click an activation link in order to be added to a newsletter mailing list. If he did not wish to be added, he should just delete the mail. The applicant considered this UCE and requested a cease & desist against the operator of the information service.
The court confirmed in its decision again the current public opinion that the unsolicited sending of an e-mail with commercial contents constitutes an illegal interference with the business of companies receiving them. Private persons also have a right to be spared from such mails as stated in 1004, 823 sect. 1 of German Civil Law.
The newsletter operator's objection that the applicant had signed in for the mailing list himself was not accepted by the court. In its opinion, the operator must prove that the applicant signed in personally. This couldn't be proved by the provider. The decision is seen controversially among jurists. The opt-in method for newsletters the decision is based on is used widely throughout the internet and was considered legally unobjectionable up to now.
If they know the exact bit to twiddle, it shouldn't be too hard to write a program which opens the file, flips the bit, and closes the file again (a one-liner, if you start sufficiently far to the left :-) ). Or am I missing something?
- It should be anonymous. PayPal isn't. Credit cards aren't. So they can create a customer profile of you. Goodbye privacy.
- It should be secure. Credit card numbers can be stolen or faked. An "ideal" system would guarantee security.
- It should be flexible. Most existing anonymous and secure e-cash systems use "tokens" of a fixed value. A good system should be able to handle arbitrary cash values in a flexible way.
Systems that work under these premises do exist. Anonymity and security can be reached by using blind electronic signatures. Flexibility can be reached by using "divisible" e-cash systems. Unfortenately, such schemes are very expensive in terms of computing power (which means, of course, that they are also expensive in terms of transaction costs).And then, of course, there's one important question: Who wants anonymous e-cash? Banks and credit companies probably don't (because they like to have your customer profile). Shops probably don't either (for the same reason). Customers? Well, face the facts: Most customers just don't appreciate the value of privacy. So, there's a simple conclusion: No market, so e-cash. It's as easy as that.
The second (third, ...) session contains bogus data, TOC entries that are outside the physical CD, (insert your favorite way to irritate CD drives here). So drives trying to read those sessions will hang up, think the CD is damaged or whatever (depending on their firmware).
Now, when you black out the outer regions (containing the later sessions), the drive can't find them and only uses the data in the first session, making the CD appear perfectly normal. As for "how wide the marking has to be", this depends on the size of the sessions. It's a question of fractions of millimeters, so it's a little hard, but manageable. The German IT magazine c't demonstrated it about a year ago.
This description may be a little simplified, but at least it's the basic principle.
Here in Germany, junk mails (e-mails, that is) are considered the same as junk faxes legally. It doesn't help much, though: Either you can't trace the spammers, or they're sitting in some obscure Caribbean country where your legislation has no power.
They don't have (or need) dedicated telescope access, they just take the data the telescope collects anyway and process it. So they don't have to pay for telescope access, the only cost is for shipping the recorded tapes to Berkeley (that's right, no broadband internet connection in Arecibo...).
"The Bible - a review" is being prepared at the moment and will be posted to Slashdot in a few days - watch out for it!
In fact, depending on the Latin nominative of "Google", this should be...
- [from "Googlus" or "Googlum"] In Googlo non est (most probable solution is "Googlum", I think)
- [from "Googla"] In Googla non est (if you prefer a female Google)
- [from "Googlis" or "Googles"] In Google / Googli non est (if you're a fan of the more exotic declination classes)
Unless, of course, your version of the Internet contains several Googles... (Alright, there are google.com, google.de, google.fr,An opt-in policy will result in better privacy, but is it practical? The aim of the whole thing is to create a rather complete directory. If you consider how many student are too lazy to answer and don't care about whether they're included or not, an "opt-in directory" will probably only contain a very small percentage of the students. Opt-out, on the other hand, just removes the ones who are really concerned about their privacy, for the price of a little work on their side, resulting in a directory as complete as possible.
In fact, this has been tried already. It seems, however, that they get scratched while in use, and thus lose their water-repelling properties. It's kind of cool, though... I saw a video of honey dripping off a spoon; the spoon was completely clean afterwards.
Well, designing an algorithm for true randomness is impossible. An algorithm, by definition, must be deterministic, which means that if it doesn't take external input, it's not truly random (because either it will generate the same numbers over and over again or a series of pseudo-random numbers). If it takes external input, on the other hand, this must come from somewhere. You can't take another algorithm (see recursion), so it has to be something mechanical. QED.
Some books on Extreme Programming might help, too. Even if you do not plan to use it, they show how to share responsibilities between management and programmers. It all boils down to:
- Management decides what will be done (business value blah blah).
- The programmers decide how long it's gonna take
- Management sets the priorities.
This helps to avoid impossible deadlines and to keep up with high quality.Try some books on Compiler Design, especially the first chapters about analysis. They should provide lots of information about parsing. One good book is Modern Compiler Design .
What this list is missing is a good book about Algorithms & Data Structures - what most software development is all about. Several comments have suggested Knuth's The Art of Computer Programming; however, for getting started, a book like Introduction to Algorithms by Cormen, Leiserson and Rivest (MIT Press) seems more appropriate.
True in the material world, but also true in the software industry; it just depends on the type of software. There isn't a market for high-quality word processing software (as long as you don't count DTP), but there's definitely a market for HQ vs. "consumer level" databases (read Access vs. Oracle / DB2), rendering/image editor (read Windows Paint vs. ... vs. Maya) etc.
In those areas, there is a clear distinction between cheap, mass-market products and high-quality niche products. You get what you're ready to pay for.
On the other hand, on the OS market, the highest quality is available for free...
Could they? Think a moment about who decides whether employees will attend such classes - that's right, PHBs who (usually) don't have a single clue about computer security or the importance of those classes. Sigh...
Terry Pratchett, Ian Stewart and Jack Cohen just published the second part of their Science of Discworld science book / fantasy novel mix. Unlike the first part, it focuses on the evolution of the human species and on psychology. A worthwile read - and if you don't know the first part yet, get it, too, and read it first.