There are multiple problems though, hiding addresses is a problem for spammers who build target lists.
Using the cgi where the client can control the recipient address for any mail is bad, even if can't be used for high-volume spam, it can still be used by stalkers, crackers, etc.
The method I posted above does not allow you to send email to anyone who's address did not appear in a page which has been viewed (relatively) recently. The hidden field is just for the random key. There isn't a lot you can do with that key --- a govenor will trip when you've sent (or attempted to send) too much mail too quickly... unless you rotate source IP addresses... a lot.... and you still can't see or control the recipient.
...but I've just been too busy to finish... and some Perl nut could probably finish it in 2 seconds:
An Apache plugin which will activate whenever a mail address is going to be rendered, and will do the following:
Store semi-permanently (refresh daily) (in a DB, RAM or a file...) a hash keyed with the email address found on the requested page. The hash contains minimally, the last time the address recieved mail and a count of mail received.
Replace the address with only the first name of the account, with something like @nodomain.not
Generate a random number and use it as a key for another hash, which points to the email address. This can live for hours, it doesn't really matter.
Pop up a form with a hidden field containing a random number. The form has a Subject, a body, and a sender field. There is no recipient field.
When the correspondant hits 'send', the server uses the random number to look up the address,
The server then checks to see if the recipient has recieved mail from the source IP address within the past x minutes, and that the reply-to address is reasonably valid (e.g. passes a DNS query... nothing is sent to the reply-to address)
If everything is o.k., the server then sends the mail to the recipient, including the source IP address, along with the appropriate reply-to address and the random number
The fact that the message was sent is stored in the hash keyed by the email address, along with the random number (this is necessary to protect the server from being flooded by bogus undeliverable messagse)
The from address is set to the server... if a not found, undeliverable, or similar error appears, then the type of error is recorded in the field keyed by the mail address, the random number appears in the undeliverable message.
The sender gets a URL with the magic number, which they can click on to check the delivery status... Only the status is shown, not the message.
When the recipient receives the message, they're free to ignore it or reply.
The mail database can clear itself every arbitrary period... something to make undeliverable messages meaningful.
The system could also be expanded to allow for total refusal of receipt from that server by the recipient clicking on a URL.
It's a bit hackish, and not a thorough plan, there are abuse details I've omitted, but I think it is a relatively simple system. The important part is to make the databases transparent and the system as trival to configure as just saying "load module"... The other important part is that nobody ever sees the email address, and being a module, it even applies to dynamic pages.
Quite right, but it would surprise me if they could write a plot where the 'real world' was not such a bad place, and still keep some kind of dramatic tension.
Yeah, like it was all some big mistake caused by some computer taking Douglas Adams too seriously.
The power idea is too stupid... the machines would have less trouble harvesting plants.
IMHO the "Zion destroyed 6 times before" line, Neo gaining power over the squid-things, and that matrix-virus travelling to a human host in Zion all reek of a plot based on Zion being part of the matrix used to deal with weird people like Neo.
It beats the Rambo plot of defeating the machines through brute force, or the Richie-Rich plot of introducing a new plot device right at the climax of the plot to save the world.
Of course it can't end entirely with Neo realizing that he loves big brother... or the matrix, so ultimately some American values theme has to come out, whether or not it results in squashing the matrix.
What would be kind of cool is if at the end of the story, Neo needs to decide whether to free people into a real world of inevitable famine and suffering or to leave them in the Matrix, freeing only his Zion friends and others around the world who don't fit in.
Of course I'm being hopeful. My pessimistic prediction is that they'll use a combination of the the Richie Rich plot, with a little bit of G.I. Joe thrown in for good measure. I mean, despite a valiant showing in a hopeless special-effects laden battle, Zion will be overwhelmned, Lots of nameless people will die, and a few strategically named people will die, humanity will be on the cusp of anihilation, all down to one terrible terrrible moment, all the doubt will lay heavily upon Morpheus' head, the audience will be left hanging... "maybe Morpheus was wrong", and that's when Neo will pull out his machine-blasting Richie-Rich superpower crap.
45 minutes of special effects later, Morpheus will be dead, Trinity will be pregnant, Neo will have a minor injury, Neo and Trinity will kiss and everyone will leave the theatre and say that the movie sucked.
Users think they want a hammer, but if they had a hammer, they would have to know how to hold a hammer, how to swing a hammer, how to hold a nail, how to hit the nail... etc. The hammer is a beautiful example of a non-intuitive user friendly design. Most of us know how to use it simply because we've all learned how to use it.
Just imagine what would happen if you gave your end-users a hammer, a handful of nails and two peices of wood.
Other strong examples of non-intuitive user friendly design v.s. intuative non-friendly designs would be the qwerty keyboard v.s. alphabetaic keyboards, the mouse v.s. the touchscreen, vi and emacs v.s. notepad, the command line v.s. WIMP environments, etc.
It's the other way around. Correlation, causation and all that rot. "The loss of customers, stock value, and reputation" can be due to a percieved lack of ethics, but that doesn't mean that the corporation has motives which are ethical.
IMHO, the parent post is quite correct.
There are limits as to what a corporation can do unethically, but the only reason for that is convienient coincidence, the arbitrary ethics of the corporate leadership, the legal limits of unethical behavior, and as you say, the limit of the lack of ethics which customers and shareholders will bear.
You should throw out the frequency measurement for the same reason that the experiment threw out the speed of light measurement.
This thing isn't intended as an "experiment" anyways, it's just a cool thing to do with your microwave to demonstrate the nodes and troughs, and to understand how frequency, wavelength and the speed of light relate to one another.
Patents suck because they last too long in the software industry and they're being applied to things which really shouldn't be patentable.
Copyright sucks because it offers too little protection for the developer and there's a risk that many of the things which are copyrighted are fundamental mathematical algorithms in standard languages.
Bill Gates' solution sucks (Copyright + EULA) because it has little accountability for failed EULA's, offers unlimited powers to the developer and generally gives nothing back to society.
The FSF's solution, (Copyright + GPL) doesn't do anything to protect any model where a software industry profits from sale of a software 'product'
Throwing all of this aside, what is really unique about one particular solution for a problem vs. another? What makes the software industry valuable? What protections should they have? What horrors are we trying to prevent? What kinds of innovation would halt if we had no protections? What kind of innovation is being halted now because of the current system?
The right to control the usage of computer programs in the sense of executing or running them is, contrary to your implications, based on copyright, not on patents, at least in USA, Commonwealth, and the European Union and for two simple reasons...
I didn't imply it, I outright stated it and backed it up with a link to a U.S. government website which contains the full text of the laws involved.
Reasoning that copyright covers usage of software because to do otherwise in a world of patents would be legally awkward, is to miss a goodly portion of the whole point of the argument against software patents. Further, it completely ignores the reason for bizzare things like click-wrap EULA's. Yes, it is absurd. It's absolutely nuts. You can lock yourself in a box with no access to the outside world, write a program to create an mp3 and legally, you could be torn out of your box and dragged into court for violating Frauenhofer's patents.
That aside, if you read your own quote:
"According to GPL section 7, patent licences that 'permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly' are acceptable to the GPL."
Say it with me: redistributionredistributionredistributionredistributionredistributionredistributionredistribution.
There are two parts of the GPL which cover usage. The first is Section 8, which says that you may opt to restrict distribution from areas where software patents might make usage of your software incur royalties.
The other part of the GPL which covers usage is the preamble, which promises that the text below restricts software in a way similar to what you describe, but nothing I can find in the GPL actually does this. I've writtent the FSF for clarification, but I can't see how the GPL could do this and at the same time be protected from a malicious company claiming a patent on some key Linux software and using the GPL to freeze all distribution. After all, that is the point of such a clause isn't it? To prevent distribution if a royalty demanding patent is inserted?
If you quote the whole paragraph in Section 0, it actually says precisely the opposite of what you're implying:
Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does.
In the context, the GPL is talking about the GPL, not about placing restrictions on running a GPL'd program. It is saying quite clearly that the GPL does not place restrictions on running the program.
As for me pointing out laws "linking the execution of programs to patents and not to copyright", that's preposterous. I can't point you to a law which doesn't exist -- the "execution of programs" is not governed by copyright in any country.
In a nutshell, this is what copyright in the U.S. governs:
SCO can only "licence the usage" of code for which they own the copyright and only if the licencee inteded to use code with that kind of restrictions, i.e. no GPL code.
That has nothing to do with copyright. That's a private agreement made on the condition of receiving the product. It's all well and true that SCO has no such agreement with Linux users, big deal, that's not what I was talking about.
I think my point was quite clear. Groklaw emphasized that by releasing Linux under the GPL, SCO has no claim to the code for which they can claim a tithe.
Groklaw:
You have continued to distribute the Linux kernel, despite alleging that it contains infringing source code. Simultaneously, you are attempting to compel purchase of "Linux Intellectual Property" licenses for binary-only use, the terms of which are incompatible with freedoms granted under the GPL.
This is incorrect. The "Linux Intellectual Property" licenses are not incompatible with the GPL because the GPL doesn't touch usage, the GPL covers distribution. SCO is trying to license usage.
Point to the spot in the GPL where it restricts distributing code under the GPL, while charging money for usage? e.g. GPL'd code which implements patented algorithms?
Without the sort of private agreement you describe, as I said before, as far as I know, the only hold SCO can have over Linux users is in software patents.
Except that the GPL covers distribution, while SCO is licensing usage.
I don't think they've read SCO's license FAQ carefully enough, nor paid enough attention to paragraph 7 of the GPL.
If this weren't the case, then any patent-encumbured GPL'd software would fail under the GPL for distribution.
My question is how SCO can license usage for a product, even if there is copyrighted material in it. Patents cover usage. SCO has made no patent claims.
OLE has always been my greatest nightmare in Word and other Windows apps. Files bloat, software crashes, version upgrades do unpredictable things, ugh, to top it all off, for anything which is meant to leave your computer in binary form, OLE is almost useless.
I know how to use it, I know how to make it work, but I just haven't ever needed to use it... with the notable exception of putting spreadsheet cells into a document or presentation, or scripting OLE objects to perform certian actions... even then, I'm so worried about something crashing and corrupting the document, I wouldn't dream of using it for anything important.
Isn't this graphical search close to what would be in the domain of Corbis?
And isn't Corbis the oft-critisized Bill Gates' project to buy photographs and sell licenses for them?
I have to credit Bill Gates with two things, first, good or bad, he's a shrewd business man, second, unless he's crazy, he's doing his job because he enjoys it.
Any technology which is restricted by a power of 2 is best measured in powers of 2.
Any technology which is not restricted by a power of 2, is best measured in powers of 10.
RAM and Flash memory is addressed in powers of 2, it would be wasteful and technologically bizzare to count it in powers of 10.
Hard drives don't fill the storage capacity of their addressing hardware, it's just not the way the technology works. Counting it in powers of 2 would be a bizzare thing to do.
Anything involving time, like clock cycles or bandwidth (e.g. kBps) should be measured in powers of 10 since the 'second' is an arbitrary unit in the land of technology.
The only deviation from the metric system should be the necessary one. These people are complaining because they have been getting the bennefit of the difference between 2^20 and 10^6 in their 'M' prefix on RAM and they don't understand why they shouldn't get the same deal on their HDD's.
Finding more efficient ways to educate people only serves to concentrate wealth into the hands of the people who can out-compete the local educators. That's not about education, or quality of life anymore, it's about foreign companies making big bucks.
If that's the objective of U.S. foreign aid, it shoudn't be called "foreign aid" it should be called investing.
The objective should be on self-sufficiency and raising unnecessarily low quality of life. However, I don't think regional variations in quality of life can be prevented. Just softened.
What if you made a full internal combustion car with a lightweight aluminum chasis, a variable speed transmission, low resistance tires and sleek aerodynamics?
The hybrids are pretty much the same except they suffer heavy batteries, gain regenerative braking and have smaller lighter IC engines.
If the girl broke into a record store and stole thousands of CD's, it would be a criminal case.
If it were a criminal case, then the RIAA would be guilty of publishing the name of a young offender. That's a crime too. I think the maximum sentence is two years in prison.
There's this idea here that if a kid's name is published, it only serves to punish the kids who want to reform.
I imagine that the U.S. or New York State has similar laws. But this isn't the FBI or State police infiltrating a shoplifting ring, this is the RIAA executing a civil action.
If this were handled by the state police or the FBI, I imagine they would probably let the kid off with a warning, and the papers would publish about how some young offender was caught stealing, then post a nice story about how parents should watch their kids.
I would say that blaming the drive-thru for receiving a third degree burn from a coffee is closer to blaming Ford because somebody rear-ended you in the drive thru and your car erupted in a ball of fire.
Extreme? Hey, in both cases, you knew what you were getting yourself into. Competant drivers don't get rear-ended and competant coffee drinkers don't spill their coffee.
http://www.motherjones.com/mother_jones/SO77/dowie.html
Slashdot Mirror
Hey, if they keep doing that, they can confuse and drag out the court battle so long that browsers will be irrelevant and Elcomsoft will be bankrupt.
Sort of like what happened to Netscape.
There are multiple problems though, hiding addresses is a problem for spammers who build target lists.
Using the cgi where the client can control the recipient address for any mail is bad, even if can't be used for high-volume spam, it can still be used by stalkers, crackers, etc.
The method I posted above does not allow you to send email to anyone who's address did not appear in a page which has been viewed (relatively) recently. The hidden field is just for the random key. There isn't a lot you can do with that key --- a govenor will trip when you've sent (or attempted to send) too much mail too quickly... unless you rotate source IP addresses... a lot.... and you still can't see or control the recipient.
...but I've just been too busy to finish... and some Perl nut could probably finish it in 2 seconds:
An Apache plugin which will activate whenever a mail address is going to be rendered, and will do the following:
The mail database can clear itself every arbitrary period... something to make undeliverable messages meaningful.
The system could also be expanded to allow for total refusal of receipt from that server by the recipient clicking on a URL.
It's a bit hackish, and not a thorough plan, there are abuse details I've omitted, but I think it is a relatively simple system. The important part is to make the databases transparent and the system as trival to configure as just saying "load module"... The other important part is that nobody ever sees the email address, and being a module, it even applies to dynamic pages.
Quite right, but it would surprise me if they could write a plot where the 'real world' was not such a bad place, and still keep some kind of dramatic tension.
Yeah, like it was all some big mistake caused by some computer taking Douglas Adams too seriously.
The power idea is too stupid... the machines would have less trouble harvesting plants.
IMHO the "Zion destroyed 6 times before" line, Neo gaining power over the squid-things, and that matrix-virus travelling to a human host in Zion all reek of a plot based on Zion being part of the matrix used to deal with weird people like Neo.
It beats the Rambo plot of defeating the machines through brute force, or the Richie-Rich plot of introducing a new plot device right at the climax of the plot to save the world.
Of course it can't end entirely with Neo realizing that he loves big brother... or the matrix, so ultimately some American values theme has to come out, whether or not it results in squashing the matrix.
What would be kind of cool is if at the end of the story, Neo needs to decide whether to free people into a real world of inevitable famine and suffering or to leave them in the Matrix, freeing only his Zion friends and others around the world who don't fit in.
Of course I'm being hopeful. My pessimistic prediction is that they'll use a combination of the the Richie Rich plot, with a little bit of G.I. Joe thrown in for good measure. I mean, despite a valiant showing in a hopeless special-effects laden battle, Zion will be overwhelmned, Lots of nameless people will die, and a few strategically named people will die, humanity will be on the cusp of anihilation, all down to one terrible terrrible moment, all the doubt will lay heavily upon Morpheus' head, the audience will be left hanging... "maybe Morpheus was wrong", and that's when Neo will pull out his machine-blasting Richie-Rich superpower crap.
45 minutes of special effects later, Morpheus will be dead, Trinity will be pregnant, Neo will have a minor injury, Neo and Trinity will kiss and everyone will leave the theatre and say that the movie sucked.
But that's just my pessimistic prediction.
Users think they want a hammer, but if they had a hammer, they would have to know how to hold a hammer, how to swing a hammer, how to hold a nail, how to hit the nail... etc. The hammer is a beautiful example of a non-intuitive user friendly design. Most of us know how to use it simply because we've all learned how to use it.
Just imagine what would happen if you gave your end-users a hammer, a handful of nails and two peices of wood.
Other strong examples of non-intuitive user friendly design v.s. intuative non-friendly designs would be the qwerty keyboard v.s. alphabetaic keyboards, the mouse v.s. the touchscreen, vi and emacs v.s. notepad, the command line v.s. WIMP environments, etc.
It's the other way around. Correlation, causation and all that rot. "The loss of customers, stock value, and reputation" can be due to a percieved lack of ethics, but that doesn't mean that the corporation has motives which are ethical.
IMHO, the parent post is quite correct.
There are limits as to what a corporation can do unethically, but the only reason for that is convienient coincidence, the arbitrary ethics of the corporate leadership, the legal limits of unethical behavior, and as you say, the limit of the lack of ethics which customers and shareholders will bear.
Banks tend to be reluctant to reverse or block bogus transactions. It's good to have a middleman involved to keep your money safe.
Did I just say that?
You should throw out the frequency measurement for the same reason that the experiment threw out the speed of light measurement.
This thing isn't intended as an "experiment" anyways, it's just a cool thing to do with your microwave to demonstrate the nodes and troughs, and to understand how frequency, wavelength and the speed of light relate to one another.
Of course it's silly.
Now that we have this entirely new medium, I'm going to patent the following novel inventions:
If halftoning and similar techniques are impossible, you could alternate colours with appropriate duty cycles for the shade?
Hey, this is the kind of obvious solution which warrants an unnecssary patent!
Patents suck because they last too long in the software industry and they're being applied to things which really shouldn't be patentable.
Copyright sucks because it offers too little protection for the developer and there's a risk that many of the things which are copyrighted are fundamental mathematical algorithms in standard languages.
Bill Gates' solution sucks (Copyright + EULA) because it has little accountability for failed EULA's, offers unlimited powers to the developer and generally gives nothing back to society.
The FSF's solution, (Copyright + GPL) doesn't do anything to protect any model where a software industry profits from sale of a software 'product'
Throwing all of this aside, what is really unique about one particular solution for a problem vs. another? What makes the software industry valuable? What protections should they have? What horrors are we trying to prevent? What kinds of innovation would halt if we had no protections? What kind of innovation is being halted now because of the current system?
I didn't imply it, I outright stated it and backed it up with a link to a U.S. government website which contains the full text of the laws involved.
Reasoning that copyright covers usage of software because to do otherwise in a world of patents would be legally awkward, is to miss a goodly portion of the whole point of the argument against software patents. Further, it completely ignores the reason for bizzare things like click-wrap EULA's. Yes, it is absurd. It's absolutely nuts. You can lock yourself in a box with no access to the outside world, write a program to create an mp3 and legally, you could be torn out of your box and dragged into court for violating Frauenhofer's patents.
That aside, if you read your own quote:
"According to GPL section 7, patent licences that 'permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly' are acceptable to the GPL."
Say it with me: redistribution redistribution redistribution redistribution redistribution redistribution redistribution.
There are two parts of the GPL which cover usage. The first is Section 8, which says that you may opt to restrict distribution from areas where software patents might make usage of your software incur royalties.
The other part of the GPL which covers usage is the preamble, which promises that the text below restricts software in a way similar to what you describe, but nothing I can find in the GPL actually does this. I've writtent the FSF for clarification, but I can't see how the GPL could do this and at the same time be protected from a malicious company claiming a patent on some key Linux software and using the GPL to freeze all distribution. After all, that is the point of such a clause isn't it? To prevent distribution if a royalty demanding patent is inserted?
If you quote the whole paragraph in Section 0, it actually says precisely the opposite of what you're implying:
In the context, the GPL is talking about the GPL, not about placing restrictions on running a GPL'd program. It is saying quite clearly that the GPL does not place restrictions on running the program.
As for me pointing out laws "linking the execution of programs to patents and not to copyright", that's preposterous. I can't point you to a law which doesn't exist -- the "execution of programs" is not governed by copyright in any country.
In a nutshell, this is what copyright in the U.S. governs:
http://www.copyright.gov/title17/92chap1.html#106
I see no reason as to why somebody can't write a GPL'd program and insist that people pay to license their patents for its usage.
That has nothing to do with copyright. That's a private agreement made on the condition of receiving the product. It's all well and true that SCO has no such agreement with Linux users, big deal, that's not what I was talking about.
I think my point was quite clear. Groklaw emphasized that by releasing Linux under the GPL, SCO has no claim to the code for which they can claim a tithe.
Groklaw:
This is incorrect. The "Linux Intellectual Property" licenses are not incompatible with the GPL because the GPL doesn't touch usage, the GPL covers distribution. SCO is trying to license usage.
Point to the spot in the GPL where it restricts distributing code under the GPL, while charging money for usage? e.g. GPL'd code which implements patented algorithms?
Without the sort of private agreement you describe, as I said before, as far as I know, the only hold SCO can have over Linux users is in software patents.
Except that the GPL covers distribution, while SCO is licensing usage.
I don't think they've read SCO's license FAQ carefully enough, nor paid enough attention to paragraph 7 of the GPL.
If this weren't the case, then any patent-encumbured GPL'd software would fail under the GPL for distribution.
My question is how SCO can license usage for a product, even if there is copyrighted material in it. Patents cover usage. SCO has made no patent claims.
OLE has always been my greatest nightmare in Word and other Windows apps. Files bloat, software crashes, version upgrades do unpredictable things, ugh, to top it all off, for anything which is meant to leave your computer in binary form, OLE is almost useless.
I know how to use it, I know how to make it work, but I just haven't ever needed to use it... with the notable exception of putting spreadsheet cells into a document or presentation, or scripting OLE objects to perform certian actions... even then, I'm so worried about something crashing and corrupting the document, I wouldn't dream of using it for anything important.
Isn't this graphical search close to what would be in the domain of Corbis?
And isn't Corbis the oft-critisized Bill Gates' project to buy photographs and sell licenses for them?
I have to credit Bill Gates with two things, first, good or bad, he's a shrewd business man, second, unless he's crazy, he's doing his job because he enjoys it.
Any technology which is restricted by a power of 2 is best measured in powers of 2.
Any technology which is not restricted by a power of 2, is best measured in powers of 10.
RAM and Flash memory is addressed in powers of 2, it would be wasteful and technologically bizzare to count it in powers of 10.
Hard drives don't fill the storage capacity of their addressing hardware, it's just not the way the technology works. Counting it in powers of 2 would be a bizzare thing to do.
Anything involving time, like clock cycles or bandwidth (e.g. kBps) should be measured in powers of 10 since the 'second' is an arbitrary unit in the land of technology.
The only deviation from the metric system should be the necessary one. These people are complaining because they have been getting the bennefit of the difference between 2^20 and 10^6 in their 'M' prefix on RAM and they don't understand why they shouldn't get the same deal on their HDD's.
I can pick up a spindle of 50 disks for $20CDN. I don't know how they sell them so cheap.
For reading, writing and arithmetic.
Finding more efficient ways to educate people only serves to concentrate wealth into the hands of the people who can out-compete the local educators. That's not about education, or quality of life anymore, it's about foreign companies making big bucks.
If that's the objective of U.S. foreign aid, it shoudn't be called "foreign aid" it should be called investing.
The objective should be on self-sufficiency and raising unnecessarily low quality of life. However, I don't think regional variations in quality of life can be prevented. Just softened.
What if you made a full internal combustion car with a lightweight aluminum chasis, a variable speed transmission, low resistance tires and sleek aerodynamics?
The hybrids are pretty much the same except they suffer heavy batteries, gain regenerative braking and have smaller lighter IC engines.
(Ontario precisely)
If the girl broke into a record store and stole thousands of CD's, it would be a criminal case.
If it were a criminal case, then the RIAA would be guilty of publishing the name of a young offender. That's a crime too. I think the maximum sentence is two years in prison.
There's this idea here that if a kid's name is published, it only serves to punish the kids who want to reform.
I imagine that the U.S. or New York State has similar laws. But this isn't the FBI or State police infiltrating a shoplifting ring, this is the RIAA executing a civil action.
If this were handled by the state police or the FBI, I imagine they would probably let the kid off with a warning, and the papers would publish about how some young offender was caught stealing, then post a nice story about how parents should watch their kids.
Why is it o.k. to do this in the civil courts?
http://www.jython.org/
One insane analogy deserves another:
I would say that blaming the drive-thru for receiving a third degree burn from a coffee is closer to blaming Ford because somebody rear-ended you in the drive thru and your car erupted in a ball of fire.
Extreme? Hey, in both cases, you knew what you were getting yourself into. Competant drivers don't get rear-ended and competant coffee drinkers don't spill their coffee. http://www.motherjones.com/mother_jones/SO77/dowie .html