Which is more 'mindless', 'isolating', 'lonely' and 'arrogant': travelling to Washington to read paper books in an environment where anything you want to share has to be meticulously copied in one form or another, or cutting and pasting quotes with links so that everyone reading it can see the full context for themselves. Furthermore, the Internet in all of its forms encourages interaction between the reader and the writer: comments, corrections, additions. I've learned more from the replies to my comments on Slashdot in any given month than from any single book I've read.
I don't read books in electronic form very much. The hardware isn't as comfortable and convenient as traditional paper books... yet. I own paper copies of The Hacker Crackdown, The New Hacker's Dictionary and The Complete Works of William Shakespeare. When I want to quote them in comments on Slashdot or in e-mail to friends, I don't want to type the quotes every time. And I want to be able to refer them to something closer than the nearest library or bookstore to read a copy for themselves. If my friends are as much like me as I think they are, they do far too much of their reading at hours when libraries aren't open.
As has been pointed out here many times, there is a lot of sharing between Linux, FreeBSD and HURD. There are some differences in goals and licenses, but there's a lot of common ground. And there are some efforts underway to create a common interface for drivers. (other URLs anyone?) If that pans out, a Linux driver will very likely be usable as a FreeBSD driver and a HURD driver. That gives me a warm fuzzy feeling.
In case the/. HTML-eater is still awake, the link above to the article on drivers is:
Re:Read the motorola press release
on
WinDSL Coming?
·
· Score: 1
In fact the press release spoke glowingly of Linux specifically and open source in general. It spoke of companies using Linux enjoying lower hardware costs and avoiding software obsolescence, and users enjoying easier upgrades.
One of the big things Open Source advocates have been asking companies to do is BSD or GPL applications that they plan to drop support for. If they no longer plan to profit off them, why not allow the users to keep, maintain, and improve them. The real answer is the upgrade treadmill, but no need to discuss that.
I fully agree. In fact, I would like to suggest that companies with a history of doing this may enjoy better sales of their products overall, and less slow down in sales late in the product's life. The reason is that customers can expect to be able to get ports and bug fixes even if there are no new features. This is a kind of hybrid pricing/licensing scheme: new products sold as products, older ones sold as support contracts on open source code by the company most familiar with the code.
What I wonder, is when companies release their source code like this, how much do users pick it up, redistribute, etc.? I mean, Darwin and Mozilla looked dead for a while, but then they picked open some help (I know Mozilla did, did Darwin?). Mozilla and Darwin were HUGE projects, it makes sense that users took a while to jump in there. On the otherhand, will a project like this be picked up and improved by proponents of open source, or will we just say, "finally, they get it" take their software and run?
One way to encourage the products to live on would be to continue to offer support contracts for them after they are open sourced and to act as project coordinators. It could even be done with an official statement by the company that they aren't going to develop new features, but that they will roll them in if submitted and continue to do ports and any bug fixes needed by the people paying for the support contracts. As a good faith gesture, they can do the port to Linux as a kick off to open sourcing the project.
I just thought I would point out the review here a few weeks back of Database Nation: The Death of Privacy at the End of the 21st Century. by Simson Garfinkel. He gives some attention to the possible consequences of the increasing coalescing of information about us. I'm about two thirds of the way through it and there are no general ideas that aren't familiar ground for long time readers of the Risks Forum and the Privacy Forum, although there are some frighten examples that were new to me. However, if you need a book to explain to Mom why you are concerned about privacy issues, this is a good one.
Of course, there is the bonus that with predictive dialers, the telemarker doesn't hear your initial "hello". If you ever call my house, and you don't respond to my initial "hello" (or respond with a "hello?" that sounds like "is anyone there?"), expect to be hung up on.
Yes, predictive dialers help me distinguish between solicited and unsolicited calls. People returning my calls don't use them. Telemarketters seeking to maximize the number of contacts with people off a large list do. If I don't hear a person at the other end, I usually hang up.
Of course, telemarketters serve a very useful social function. It seems that every society defines a hierarchy of more or less respected professions. With the exception of malpracticing quacks and ambulance chasers, doctors and lawyers are afforded considerable respect. So are many other professions. And in these politically correct days, we try not to look down on those who take unpleasant jobs to pay the bills. But when someone calls to be rude....
And a note to all telemarketters: Put me on your Do Not Call List. Calling me is just a way to hear me say that with my own voice.
I like the question above, but I would refine it a bit:
The GNU project and the FSF seem to be succeeding. All of the tools that were intended to be freely available have been built. That isn't to say that the work is now, or ever will be complete. I merely mean that free software has achieved parity or better in the realm of development tools. I have also noticed in the past couple of years that certain projects have been handed off. egcs replaced gcc/g++ as the central compiler development effort. And you have handed off many of the day to day tasks with Emacs development. Are you freeing up time for advocacy, new programming projects or both? What do you have in mind for the future?
Alvin Toffler wrote about this idea several decades ago in Future Shock. He talked about the accelerating pace of our lives. He pointed out that past generations had seen a rate of change that could be ignored, or in the first half of this century slowly adapted to. That is obviously no longer the case. He spoke of obsolescent knowledge. It seems clear to me (although I haven't read it yet) that Faster explores those ideas with the perspective that several decades of living with constant change has brought.
Our rate of change carries with it twin dangers, the Scylla and Charibdis of our age. Merely to keep up, one must be an unrepentant neophile. And yet, we cannot blindly accept all that is new as a boon. I don't believe that we can keep new ideas from being distributed, nor would I approve of doing it. We need to constantly consider the consequences they may bring and prepare ourselves for them.
For anyone who actually knows the details of some of the systems mentioned: how much of a security compromise is this. For medical equipment, I suppose this might allow you to break into it and compromise it, causing it to operate in fatal ways. That would make tampering with over-the-counter medication look like the work of amateurs (wait, it was). However, since none of the systems mentioned are known for being on the Net, any tampering would require physical access. It would likely be an inside job.
Re:A warning that should have been heeded
on
RMS On eBooks
·
· Score: 2
I presented the link to his article as a counterpoint to RMS for people who find RMS's stridency to be too much. The fact that Garfinkel does not appear to be participating in the Amazon boycott certainly does indicate that his opinions differ from Stallman's on at least some subjects. I thought that the common ground that the two of them share might offer some insight.
Oh, and he also mentioned that he hasn't made much money off the link at this point. The way I see that link is Garfinkel's attempt to provide an easy way for anyone who has found his web site to order any of his books.
Another source for freely distributable books
on
RMS On eBooks
·
· Score: 1
A warning that should have been heeded
on
RMS On eBooks
·
· Score: 5
Simson Garfinkel, author of Database Nation among numerous other books, write an article for the Boston Globe nearly two years ago. His warning in that article was remarkable considering what happened with DeCCS. For those of you inclined to view RMS as an alarmist, read Garfinkel's article and consider the fact that he got it right.
I consider stories about the stock market to be CNNfn or CNBCs thing. We only bring up the market when it deals with specific companies that we're interested in, not when the story is "The Market".
If I want general, up-to-date news on the stock market, I have cable. I'll turn on CNBC. And allegedly some of the newsstands around here carry the Wall Street Journal. Slashdot is neither of these. That isn't central to it. To try to carry a story every day with the previous day's closing on all the stocks of interest to Slashdotters would detract from the character of Slashdot. However, running stories about the IPO behavior of Open Source companies and to track them from time to time is appropriate. It is a part of what those companies are. It can effect how they act.
For the stock market addicted, I suggest that Slashdot could offer a Slashbox that provides the latest quotes for the stocks of companies that Slashdot often runs stories about. Obviously any Open Source company would make the list. But Intel, Microsoft, Sun, Apple, and IBM get mentioned often enough to make the list. One of the greatest strengths of Slashdot is its configurability. Let the people who want the quotes get them.
As for the people who never want financial news, go customize your Slashdot homepage to exclude stories about "The Almighty Buck". And a note to Slashdot. It is a challenge to find it. I could understand alphabetizing it under "Almighty", "Buck" or even "The", but it seems to be alphabetized under "Mighty" perhaps to put it right under "Microsoft".
Dystopian fiction from Stallman
on
RMS On eBooks
·
· Score: 5
RMS also wrote a dystopian story story Right To Read which was supposed to depict a worst-case scenario. It is a quick read. Take a look and consider the question it indirectly poses: How do we retain the freedoms we love?
The number that I understand to be generally acceptable is that an employee costs twice what s/he is paid. So an employee that grosses $50k/yr costs the company $100k/yr when all the overhead, taxes and crap is figured in.
For back of the envelope calculations that is an excellent rule-of-thumb.
Look at it this way: Do you believe I underestimated these other costs by a factor of 1000? That's about what it takes to move $480,000 to the realm of "hundreds of millions of dollars". I agree there are other costs, but to say that an admin's paycheck makes up only 0.5% of the actual cost seems to be stretching things...
I'm sorry that I wasn't clear. I agree with your point. At worst, you underestimated by an order of magnitude. The only reason that I think you might be off is that I think the number of servers hit was underestimated because of under-reporting. I was trying to bring up some related issues.
Exact pricing will be finalized nearer to our late 2001 service launch. Our equipment will be comparably priced to digital satellite TV equipment. iSKY will feature an affordable flat monthly fee for unlimited high-speed access to the Internet. Developed from the customer up, iSKY is being built with affordability a key consideration in every technological and business decision.
There are a couple of external factors involved in the pricing of digital satellite service. First, it has to compete with cable in areas where cable is available. I've compared the pricing. It looks like satellite services substitute a few more (non-local) channels in place of the local service they can't provide at very similar price points for their packages in our area at least.
Second, they have to build sufficient volume to keep the price at that level. The equipment that they have to maintain is different from cable. The cost per user scales quite differently. With cable, each new neighborhood needs its own lines. With satellite, the infrastructure is centralized or in space. It gets paid for in big chunks.
They'll be competing with cable modems, DSL, and for rural users, ordinary dial-up access. And they are comping in after those services have gotten a head start. They are going to have to sell an attractive package, or they won't generate the customer base they need to keep going.
Number of affected servers: 3000 (This is almost certainly way low, in that I'm assuming 25% of approximately 12,000 servers on the Internet. I have no real idea how many servers there are.)
Number of hours spent by a net admin to following the problem, creating or downloading a patch, and verifying both the problem and the solution: 4?
Average hourly wage of a net admin: $40?
Putting it together: 3000x4x40 = $480,000. Unless my estimates are off by a combined three orders of magnitude, it's probably not going to be "hundreds of millions of dollars". Sorry, ESR.
I have two points to make about this. Farther along you did point out that installing patches is in the job description. True enough. Installing patches to correct a backdoor, as this was alleged to be, should not be. But, your calculations leave out a number of other factors. The cost of an employee's time doesn't stop at his paycheck. There was server downtime involved. There are also other non-salary costs in keeping employees: benefits, the employer's contribution to Social Security, office space, etc.
A former coworker of mine, and still a friend, pointed out something to me a couple of years ago about time spent by engineers, or anyone producing a product. It's value should be measured by what they could produce if not interrupted by whatever you are evaluating the cost of. Certainly, for a network administrator, this is part of the job. And what about the little start-up, where the three hackers with the brilliant product idea who are slinging code 18 hours a day also put together the company web site because they are paying themselves in equity in the company rather than in dollars, which they don't have. They can't hired a net admin right now, but the cost to them of those 4 hours may be huge.
NPR's All Things Considered show ran a story yesterday concerning a lawsuit by some retailers over Sony selling their artist's CD's directly via the Web. The interesting part for this discussion was the mention that Sony and BMG have plans to begin offering an online service to pay for downloads of music. The audio feeds for the story are here: 14.4, or 28.8 and up.
Unfortunately, this lawsuit will only further convince amazon that everyone is using patents and lawsuits and they must too in order to be competitive. As a matter of fact, our attempts to convince amazon to change it's position on this issue may be seriously set back. This is a sad, sad day...
This is really only going to be more fuel for the patent cross-licensing bonfire. One of the reasons that companies build patent portfolios is that they have assets to protect. Unfortunately, the way the patent laws are used these days, it is their economic assets that they are protecting, rather than their IP. When sued, they have a nice array of IP to cross-license. In case the point here has escaped anyone in the past discussions of patent abuses, the way patents are being used creates two playing fields: one for those with patents to trade and another for those without. Is anyone interested in donating patents to a free software organization to be used defensively?
So I'm locally pessimistic but globally optimistic! Who said AI is 10 years away? It's here now, in limited forms, yielding a lot of economic value, as your mouse clickstream is datamined so the ads which pop up are for things you might actually buy. But the SF ideal of a humanoid robot like Commander Data is centuries away. [emphasis added]
What scares me about this is that programmers learn early on the real meaning of GIGO: Garbage In, Garbage Out. Regrettably, in the real world, the garbage that goes in (sometimes the data, sometimes the assumptions on which the code is based) often remains unexamined. And the output is treated as gospel. In particular, with data mining, there is a crying need to educate the users of the output about the potential for errors in the input, bugs in the code and just incorrect assumptions (e.g., I bought one book in a series, I must want the rest). Data mining is a fine way to narrow down our guesses, increasing the odds that we are right. But it is not and never will be an exact science.
Savage attacks by insiders exiting are the worst thing in science, such as Bar Hillel's attack on Machine Translation in the 60's. Forty Years later, MT is "cool" again, in this month's issue of Wired.
Yes, this month's Wired gives a good summary of the history of MT. It gives the firm impression that MT will remain several years off for quite some time. And yet, we will continue to see useful, special purpose applications of what is learned in trying to achieve it. The article emphasizes that MT has had its greatest success within constrained domains. It even mentions that Babelfish, for all its warts, is particularly good at translating recipes.
The patch may consist of a tiny microchip that, after first determining exactly what your body needs, transfers vitamins to your body transdermally. Goodbye Penguin Mints, hello Penguin Patch!
The last time I checked, there was no USRDA for caffeine.
I found the exact quotes on the possible consequences of a Napster win to be more interesting than the summary of them:
"If Napster wins this, then presumably everybody that is propagating MP3 files and movie files will be protected," said attorney Carl Oppedahl of Oppedahl & Larson in Frisco, Colo. "And every time the music industry faces a technological change or an unfavorable ruling, they run to Congress to plug the latest hole in the dike."
"The defense is a novel one, but if Napster wins this, I predict the law will be rewritten in eight minutes," said Neil Rosini, a lawyer at New York law firm Franklin, Weinrib, Rudell & Vassallo, who represents online music firm MyPlay. "The DMCA was never intended for companies like Napster."
These quotes illustrate one of the biggest problems with trying to write specific laws to cover specific, narrow behaviors. The wording of the laws always misses important details. Give me a simple law, a good jury and a judge to answer their questions over endless legislative tweeking any day.
Slashdot Mirror
Which is more 'mindless', 'isolating', 'lonely' and 'arrogant': travelling to Washington to read paper books in an environment where anything you want to share has to be meticulously copied in one form or another, or cutting and pasting quotes with links so that everyone reading it can see the full context for themselves. Furthermore, the Internet in all of its forms encourages interaction between the reader and the writer: comments, corrections, additions. I've learned more from the replies to my comments on Slashdot in any given month than from any single book I've read.
... yet. I own paper copies of The Hacker Crackdown, The New Hacker's Dictionary and The Complete Works of William Shakespeare. When I want to quote them in comments on Slashdot or in e-mail to friends, I don't want to type the quotes every time. And I want to be able to refer them to something closer than the nearest library or bookstore to read a copy for themselves. If my friends are as much like me as I think they are, they do far too much of their reading at hours when libraries aren't open.
I don't read books in electronic form very much. The hardware isn't as comfortable and convenient as traditional paper books
As has been pointed out here many times, there is a lot of sharing between Linux, FreeBSD and HURD. There are some differences in goals and licenses, but there's a lot of common ground. And there are some efforts underway to create a common interface for drivers. (other URLs anyone?) If that pans out, a Linux driver will very likely be usable as a FreeBSD driver and a HURD driver. That gives me a warm fuzzy feeling.
/. HTML-eater is still awake, the link above to the article on drivers is:
9 210&mode=thread
In case the
http://slashdot.org/article.pl?sid=00/03/31/145
In fact the press release spoke glowingly of Linux specifically and open source in general. It spoke of companies using Linux enjoying lower hardware costs and avoiding software obsolescence, and users enjoying easier upgrades.
I fully agree. In fact, I would like to suggest that companies with a history of doing this may enjoy better sales of their products overall, and less slow down in sales late in the product's life. The reason is that customers can expect to be able to get ports and bug fixes even if there are no new features. This is a kind of hybrid pricing/licensing scheme: new products sold as products, older ones sold as support contracts on open source code by the company most familiar with the code.
One way to encourage the products to live on would be to continue to offer support contracts for them after they are open sourced and to act as project coordinators. It could even be done with an official statement by the company that they aren't going to develop new features, but that they will roll them in if submitted and continue to do ports and any bug fixes needed by the people paying for the support contracts. As a good faith gesture, they can do the port to Linux as a kick off to open sourcing the project.
I just thought I would point out the review here a few weeks back of Database Nation: The Death of Privacy at the End of the 21st Century. by Simson Garfinkel. He gives some attention to the possible consequences of the increasing coalescing of information about us. I'm about two thirds of the way through it and there are no general ideas that aren't familiar ground for long time readers of the Risks Forum and the Privacy Forum, although there are some frighten examples that were new to me. However, if you need a book to explain to Mom why you are concerned about privacy issues, this is a good one.
Yes, predictive dialers help me distinguish between solicited and unsolicited calls. People returning my calls don't use them. Telemarketters seeking to maximize the number of contacts with people off a large list do. If I don't hear a person at the other end, I usually hang up.
Of course, telemarketters serve a very useful social function. It seems that every society defines a hierarchy of more or less respected professions. With the exception of malpracticing quacks and ambulance chasers, doctors and lawyers are afforded considerable respect. So are many other professions. And in these politically correct days, we try not to look down on those who take unpleasant jobs to pay the bills. But when someone calls to be rude....
And a note to all telemarketters: Put me on your Do Not Call List. Calling me is just a way to hear me say that with my own voice.
I like the question above, but I would refine it a bit:
The GNU project and the FSF seem to be succeeding. All of the tools that were intended to be freely available have been built. That isn't to say that the work is now, or ever will be complete. I merely mean that free software has achieved parity or better in the realm of development tools. I have also noticed in the past couple of years that certain projects have been handed off. egcs replaced gcc/g++ as the central compiler development effort. And you have handed off many of the day to day tasks with Emacs development. Are you freeing up time for advocacy, new programming projects or both? What do you have in mind for the future?
Alvin Toffler wrote about this idea several decades ago in Future Shock. He talked about the accelerating pace of our lives. He pointed out that past generations had seen a rate of change that could be ignored, or in the first half of this century slowly adapted to. That is obviously no longer the case. He spoke of obsolescent knowledge. It seems clear to me (although I haven't read it yet) that Faster explores those ideas with the perspective that several decades of living with constant change has brought.
Our rate of change carries with it twin dangers, the Scylla and Charibdis of our age. Merely to keep up, one must be an unrepentant neophile. And yet, we cannot blindly accept all that is new as a boon. I don't believe that we can keep new ideas from being distributed, nor would I approve of doing it. We need to constantly consider the consequences they may bring and prepare ourselves for them.
For anyone who actually knows the details of some of the systems mentioned: how much of a security compromise is this. For medical equipment, I suppose this might allow you to break into it and compromise it, causing it to operate in fatal ways. That would make tampering with over-the-counter medication look like the work of amateurs (wait, it was). However, since none of the systems mentioned are known for being on the Net, any tampering would require physical access. It would likely be an inside job.
I presented the link to his article as a counterpoint to RMS for people who find RMS's stridency to be too much. The fact that Garfinkel does not appear to be participating in the Amazon boycott certainly does indicate that his opinions differ from Stallman's on at least some subjects. I thought that the common ground that the two of them share might offer some insight.
Oh, and he also mentioned that he hasn't made much money off the link at this point. The way I see that link is Garfinkel's attempt to provide an easy way for anyone who has found his web site to order any of his books.
The Open Content web site has a link to a page of open content works available on the net at http://wiley.byu.edu/opencontent/
Simson Garfinkel, author of Database Nation among numerous other books, write an article for the Boston Globe nearly two years ago. His warning in that article was remarkable considering what happened with DeCCS. For those of you inclined to view RMS as an alarmist, read Garfinkel's article and consider the fact that he got it right.
If I want general, up-to-date news on the stock market, I have cable. I'll turn on CNBC. And allegedly some of the newsstands around here carry the Wall Street Journal. Slashdot is neither of these. That isn't central to it. To try to carry a story every day with the previous day's closing on all the stocks of interest to Slashdotters would detract from the character of Slashdot. However, running stories about the IPO behavior of Open Source companies and to track them from time to time is appropriate. It is a part of what those companies are. It can effect how they act.
For the stock market addicted, I suggest that Slashdot could offer a Slashbox that provides the latest quotes for the stocks of companies that Slashdot often runs stories about. Obviously any Open Source company would make the list. But Intel, Microsoft, Sun, Apple, and IBM get mentioned often enough to make the list. One of the greatest strengths of Slashdot is its configurability. Let the people who want the quotes get them.
As for the people who never want financial news, go customize your Slashdot homepage to exclude stories about "The Almighty Buck". And a note to Slashdot. It is a challenge to find it. I could understand alphabetizing it under "Almighty", "Buck" or even "The", but it seems to be alphabetized under "Mighty" perhaps to put it right under "Microsoft".
RMS also wrote a dystopian story story Right To Read which was supposed to depict a worst-case scenario. It is a quick read. Take a look and consider the question it indirectly poses: How do we retain the freedoms we love?
For back of the envelope calculations that is an excellent rule-of-thumb.
I'm sorry that I wasn't clear. I agree with your point. At worst, you underestimated by an order of magnitude. The only reason that I think you might be off is that I think the number of servers hit was underestimated because of under-reporting. I was trying to bring up some related issues.
There are a couple of external factors involved in the pricing of digital satellite service. First, it has to compete with cable in areas where cable is available. I've compared the pricing. It looks like satellite services substitute a few more (non-local) channels in place of the local service they can't provide at very similar price points for their packages in our area at least.
Second, they have to build sufficient volume to keep the price at that level. The equipment that they have to maintain is different from cable. The cost per user scales quite differently. With cable, each new neighborhood needs its own lines. With satellite, the infrastructure is centralized or in space. It gets paid for in big chunks.
They'll be competing with cable modems, DSL, and for rural users, ordinary dial-up access. And they are comping in after those services have gotten a head start. They are going to have to sell an attractive package, or they won't generate the customer base they need to keep going.
I have two points to make about this. Farther along you did point out that installing patches is in the job description. True enough. Installing patches to correct a backdoor, as this was alleged to be, should not be. But, your calculations leave out a number of other factors. The cost of an employee's time doesn't stop at his paycheck. There was server downtime involved. There are also other non-salary costs in keeping employees: benefits, the employer's contribution to Social Security, office space, etc.
A former coworker of mine, and still a friend, pointed out something to me a couple of years ago about time spent by engineers, or anyone producing a product. It's value should be measured by what they could produce if not interrupted by whatever you are evaluating the cost of. Certainly, for a network administrator, this is part of the job. And what about the little start-up, where the three hackers with the brilliant product idea who are slinging code 18 hours a day also put together the company web site because they are paying themselves in equity in the company rather than in dollars, which they don't have. They can't hired a net admin right now, but the cost to them of those 4 hours may be huge.
NPR's All Things Considered show ran a story yesterday concerning a lawsuit by some retailers over Sony selling their artist's CD's directly via the Web. The interesting part for this discussion was the mention that Sony and BMG have plans to begin offering an online service to pay for downloads of music. The audio feeds for the story are here: 14.4, or 28.8 and up.
This is really only going to be more fuel for the patent cross-licensing bonfire. One of the reasons that companies build patent portfolios is that they have assets to protect. Unfortunately, the way the patent laws are used these days, it is their economic assets that they are protecting, rather than their IP. When sued, they have a nice array of IP to cross-license. In case the point here has escaped anyone in the past discussions of patent abuses, the way patents are being used creates two playing fields: one for those with patents to trade and another for those without. Is anyone interested in donating patents to a free software organization to be used defensively?
What scares me about this is that programmers learn early on the real meaning of GIGO: Garbage In, Garbage Out. Regrettably, in the real world, the garbage that goes in (sometimes the data, sometimes the assumptions on which the code is based) often remains unexamined. And the output is treated as gospel. In particular, with data mining, there is a crying need to educate the users of the output about the potential for errors in the input, bugs in the code and just incorrect assumptions (e.g., I bought one book in a series, I must want the rest). Data mining is a fine way to narrow down our guesses, increasing the odds that we are right. But it is not and never will be an exact science.
Yes, this month's Wired gives a good summary of the history of MT. It gives the firm impression that MT will remain several years off for quite some time. And yet, we will continue to see useful, special purpose applications of what is learned in trying to achieve it. The article emphasizes that MT has had its greatest success within constrained domains. It even mentions that Babelfish, for all its warts, is particularly good at translating recipes.
The last time I checked, there was no USRDA for caffeine.
These quotes illustrate one of the biggest problems with trying to write specific laws to cover specific, narrow behaviors. The wording of the laws always misses important details. Give me a simple law, a good jury and a judge to answer their questions over endless legislative tweeking any day.