Any score which is relative to the body of test takers is normalized every sitting by definition (ie, if you have the average student writing in a sitting getting 500 by definition, then the average student writing in a session will always get 500).
Reweighting or changing the statistical model, OTOH - that's something else entirely. If I were in the college board's shoes, I'd see it as something regrettable (because a lay audience just won't get it) but totally necessary to protect the integrity of one's rating scale.
I'd think that the difference is that any company can pick up a competitor's product.
Still, the innovator or developer of any product has three advantages when it comes to selling. They are the most familiar with the existing code base: this is very important when it comes to porting a given piece of software. RMS, for example, charges usurious rates to port GCC. Any complex piece of software is going to take a considerable amount of time and thus expense to familiarize oneself with.
They hold the copyright, and can offer alternative licensing terms. This may be important to some companies, who don't want an "open" piece of software or who may want to redistribute their modifications without including source and who will pay for source access.
Finally, they gain name recognition, credibility and thus consumer preference through the product's use. Even though Jim Joe Bob can put togeter a linux distribution based on Debian, they certainly aren't Debian and don't have the credibility of Debian. Red Hat is an even better example of this.
I think that they renormalize every year. That's the point of the test - to produce a ranking of test takers where the top scorers receive an 800, the bottom scorers receive a 200, and the median scorers receive a 500. This calculation is done by means of standard deviation from the mean; an 800 does not mean a perfect score, but it does mean a score in the 99th percentile of all test takers.
Why they'd change the scores would be because the median student would be because the distribution itself has changed shape. Hypothetically, we have a normal distribution (bell curve) which represents the scores of all test takers, and the apex of the curve is at 500.
If you had the median student scoring below 500, you'd have a heavy-tailed distribution - the bell curve is warped to one side (in this case, the tail would be heavy to the right) and the model has to be adjusted for that.
It's basic statistics, and I'm surprised that it's so widely misunderstood.
Do elite colleges matter? Are the worth the money? Consider:
In 2000 we will most likely have a choice between Harvard (Gore), Yale (Bush), or Princeton (Bradley).
Of our nine current Supreme Court justices, four attended the same undergraduate college (Stanford). Of the five others, we find graduates of Harvard, Cornell, Chicago, Georgetown, and Holy Cross. Five out of the nine attended a single law school (Harvard); the other law schools are Stanford (2), Yale, Columbia, and Northwestern. (Ginsberg attended Harvard Law for two years, then transferred to Columbia). Not one attended a public university or college, or an obscure private school.
Of the top four cabinet positions: State: Wellesley (Albright) Defense: Bowdoin (Cohen) Treasury: MIT (Summers, replacing Harvard (Rubin)) Attorney-General: Cornell (Reno)
Similar patterns, although not as extreme, can be found in business, the arts, etc. Keep in mind that only about 1-2% of all college graduates attended elite private colleges. Because larger numbers of Americans are attending (mostly public) universities and colleges, the percentage attending elite institutions has probably declined over time, which makes the disproportionate success of elite graduates even more striking.
(from the LBO-talk mailing list, posted by Kelley Walker)
Any college anywhere has the "right" to choose whom to accept. That's the purpose of an admissions process. However, most universities, and especially private ones, realize that while undergraduates pay the bills, faculty and graduate students make the reputation, and that's where the effort goes. Perhaps you had a good experience as an undergrad at CMU but it's certainly not generalizable.
I think that eCollegebid is an excellent way of running a clearings process, although perhaps not a primary admissions process.
Universities in the UK run their clearings process this way: when the second round of admissions is done and all students have committed or declined, they post the program vacancies and invite applications to the remaining spots rather than take students who don't meet their standards or run programs half full.
You don't bid on tuition in the UK, but for private schools in the US that should definitely be an option; they have the option of subsidizing you or not, just like they would have the option of offering you a scholarship or a bursary if you participated in the primary applications process. If they're already running a program, the incremental cost for additional students is quite small.
If collegebid became widely adopted, I think that it would definitely fill a niche in the US' private educational program. If you don't like the quality offers you get in the primary process, decline them and attempt to bid up. If you can't afford tuition where you did get in, try to bid down for schools of lesser quality. This freedom isn't presently available, but in a private system perhaps it should be.
It might be wonderful but still I doubt that any more than a tiny fraction of the internet community would use encryption. It's changed from the days where traffic was mainly academic; now the "average user" is a suburbanite with no grasp and no willingness to inform him or herself about issues of privacy, security, cryptography or the like.
Of course, that may not be necessary; it might just be necessary to encrypt a small fraction of network traffic to result in changes at Echelon. Of course, Echelon is reportedly a global operation, so it's necessarily a slow process. Also, I doubt that Echelon is the only communications intercept operation out there. It's something that any government or even any multinational corporation would be interested in, no matter where they're located.
However, I don't think that's as relevant as other considerations. One should be encouraging prudent and secure communications for one's own good, not just to thumb one's nose at the US government. Make your PGP or GPG key available; use encryption routinely, and expect your correspondents to reply in kind. In other words, take responsibility for your own security.
I wouldn't go that far. It's certainly not a bad thing, and definitely not pathologically manipulative, to want your nearest and dearest to be running a decent computer.
Nor are many octogenarians going to be using a computer at all. You aren't denying someone tech support (more to the point, you're setting up the computer to require little if any tech support) from their peers if their peers aren't going to be offering support.
Likewise, I don't think that the "community of users" argument follows. It takes an effort to seek out and belong to any community, and most people who aren't geeks really have no desire to belong to a community of computer users. The extent that they *will* belong to a community of users is through things like email, will be with people whom they have pre-existing social relationships with, and it'll be primarily social, not primarily technical.
--
Re:Don't disagree with what it didn't say
on
Gartner Slams Linux
·
· Score: 1
I recall that the (inter|intra)net-connected server market was NT's big marketing goal only a couple of years ago. But I might recall incorrectly.
Still. If you need NT to provide a decent single user workstation it's a great failure of design; it should be massive overkill given its technical roots (VAX, OS/2). Rather, if all NT can provide is a decent single user workstation, it's massive overkill. Which is fine so long as you have the resources, and most corporations can and do fork up the cash on an ongoing basis.
This specific report is referring to a single sector of the computing market: business productivity platforms, or desktops running MS Office. There's no port of Office and no analogous system which provides 100% compatibility (and/or superior functionality), so the verdict will be negative, will have to be negative. It's phrased in such a way that there is no alternative.
Is that dishonest of the Gartner group? Well, yes, it is; it implies that a valid comparison is possible. The preconceptions intrinsic to the question are (or seem to be) so strong that starting from a level playing field is impossible. Nor would it be possible to fairly compare other unices/unix-like OSs on the same question.
At the same time: it's nothing that a linux user doesn't already know, and it's nothing that any current linux user should worry about. (Nor does it say anything that you couldn't find out in ten minutes of asking questions and searching and thinking... but YMMV, once again).
You have heard about displaying an xterm from the server box on your own desktop. Or perhaps you haven't.
Whether you have or you haven't, it's one of the biggest and best real-world uses of X - you can run X applications on a distributed model and display them locally.
Symptoms, no; they're different disorders (although they are differential diagnoses of each other).
What has been suggested quite recently in clinical research is that ADHD/ADD as a child or juvenile increases the risk of developing bipolar disorder later in life (bear in mind that the average time of diagnosis for bipolar disorder is in one's late 20's).
One could make arguments as to whether there is a shared cause or that the former causes the latter through some biochemical or even through a social mechanism. Nobody knows what causes either ADD/ADHD or bipolar disorder; it's even possible that there are multiple possible causes producing the same symptomology.
I used to favour the explanation that there was a genetic component to all cases of bipolar disorder; however, right now I would have to demur. Perhaps someone on/. does research in this field; I'm merely an educated layperson.
The example of quarantine (which I suspect you don't know much about) is a fine example. People with TB, even today in some cases, are quarantined until they get sufficiently better so that releasing them is a good risk.
Of course, it wasn't the existence of quarantine laws that stopped widespread tb; it was the existence of effective antibiotics. People who were badly enough off to go to a tb sanatorium were by and large already bedridden.
HIV's spread isn't going to be confined by quarantine; we know that people can be asymptomatic for years, decades even. Those whom we have the most to fear from are people who carry HIV but haven't yet developed AIDS, who haven't been tested (most people aren't, unless they're blood donors), and who thus haven't adopted safer sex practises.
They wouldn't be caught by a medical dragnet through online records. They wouldn't be caught by quarantine laws. They would be caught by better sex education, which would both keep them from getting it in the first place and from transmitting it in the second.
Having online and fully accessible medical records strikes me as so much expensive but trendy claptrap. Compare the huge effort of putting records for 300 million people securely online to the relatively tiny cost of adequate and universal public health. Of course, adequate and universal public health won't make headlines and won't make millionaires - so it's neglected.
I studied statistics at university, and that kind of large-scale correlative study that you described was touched upon.
However, this kind of thing - online medical records - does not lead directly to large-scale longitudinal studies. If anything, the restrictions on use of data will render such studies impossible. Likewise, reporting standards are likely to be variable at best. The data, if you can get it, will be only mildly usable. Big correlations and interrelations will be visible; subtler ones will not.
IMHO, of course; I'd like to have that much data at my disposal, too. But it's much more likely to happen through big (and, yes, expensive, but good data is always expensive) longitudinal studies, not through electronic record keeping.
Firstly, someone who is said to engineer a system increases the information represented within. This is one of the creationist pseudo-arguments - it doesn't work there, but it works here.
Secondly, the sanctity of the fertilized egg (assuming you recognize such a sanctity) is undisturbed. The engineering merely must be done prior to fertilization. I've never seen anyone claim that the unfertilized egg or sperm is a legitimate life.
That said, you do need to confront items of Catholic doctrine, I believe, which deal with the integrity of normal conception. However, they're different from those which deal with the termination of pregnancy.
I recall reading an interesting study which looked at the positive factors for economic success.
Being white was positively correlated to economic success. No surprise there.
Being male was positively correlated to economic success. No surprise there, either.
Going to a good college was positively correlated to economic success. Again, no surprise.
However, intelligence/academic success was only mildly correlated to economic success. What the study found was that the highest earning group was that the stupidest white males who went to top colleges were in fact the top earners.
Admittedly, we're looking at a complex system; very smart people in colleges, especially top colleges, are tempted by academe, which pays poorly. Still, I don't think that a race of genetically engineered super-men, who dominate positions at the head of industry and government, is in the offing.
That said, the expectation is probably much more powerful than the reality. If we expect super-men to be the best leaders, no doubt they'll be picked early as leaders, given the necessary background for leadership, and then moved up to positions of leadership. No doubt a disproportionate amount of the best leaders will come from this group.
It will hardly amount to proof of their superiority, however. It is merely proof of the natural credulity of the human animal.
The main reason is to give a hierarchy for authoritative lookups. You could certainly place all domains in a unified namespace (bear in mind that the namespace would shrink proportionally), but you'd have to have a single authoritative register.
Um, no thanks.
The way it is now, NSI has.com,.org,.net,.edu, etc etc, but they don't have to - they could just as easily spin off the edu's, for example, to the NSF, and that would work seamlessly. You could even have one company holding the authoritative server for one tld and one tld only - no problem there. When you *don't* have tld's, on the other hand, you have to figure out some other method to apportion and split authority, or you have to stick them together.
Remember that the existence of tld's may not be (and isn't) an advantage when it comes to database searching, but it's certainly an advantage when it comes to determining the proper progress of nameserver authority. And that's the sine qua non of DNS - the existence of a proper progression of authorities for any given name.
I bought a 128Mb stick of SDRAM for $100 Canadian (~= 69$US) at the absolute bottom market; I'm glad I did. Admittedly, that was a used purchase, but the multiplier in the great white north was about.75$US per meg for new RAM.
I doubt that Silicon Valley is on a totally different banking system than the rest of America (where people who don't know better are legion). If SV is, feel free to disregard my comment.
The main problem for direct billing systems - well, the main problems - are twofold.
First, we have the problem of liability. Banks, especially banks, have to know who's going to pay if something goes wrong. They like to divide up and preferably cap liability before such a system goes into operation. Bear in mind that many large organizations, such as banks, self-insure (that is, rather than pay a certain amount of money to an insurer, they are big enough to merely set money aside and act as their own insurer), and that increases the stakes.
Second, we have the problem of incompetence. Most software projects are failures to some extent - nearly a third get trashed because they don't satisfy spec, or the direction of the corp changes, or they're just plain not used. This doesn't mean that online banking services isn't a possibility, but you don't want to fail while doing this kind of thing. In other words, it's likely that online banking will remain in beta for half of forever.
To the second point you can add the observation that the cheque-clearing and banking laws in the US drastically and desperately need rationalization - there are few provisions for banks with no physical presence to operate in an arbitrary state. But YMMV.
The japanese case is probably because Japan has fallen behind North America and Europe when it comes to networking and the like. It's more a cultural thing than anything else; Japan does very well when it comes to other aspects of computing.
The world has quite a bit of variance when it comes to access to computing facilities - really, when it comes to telecom facilities at all. I recall seeing a map which showed that most of the world has only limited access to the internet - most of africa, for example, is UUCP-only. No doubt that's changed in the intervening years, but the problem remains that almost nobody is going to make money off wide availability of internet services to that part of the world, so you're unlikely to see it happen.
Well, you can look at it from two perspectives: from that of a hardware vendor and that of a software vendor.
The hardware vendor should definitely go open source. Most likely they have patents on their hardware; while the source may give someone an edge in reverse engineering their hardware, it's not really that much of an edge, nor will it be a *competitive* edge. That is to say, no one is going to compete with Company X if all they do is copy what Company X does, merely two years later. Technology moves too fast for that.
For software companies, it's a different paradigm. Consider that most big companies, such as IBM, make a substantial chunk of their software dollar by providing systems integrations and customization on said software. Companies like IBM have more work than they can bother to do - that's why they develop "business partner" networks. There's substantial money to be made by going open source with your product and making money by supporting it. Is it sufficient to give away the software in its entirety? It is for some, probably not for others.
Likewise, very few companies are going to take your custom job, with open source, and distribute the source to it, even if they have the right to. Their competitive edge is tied into having your software - not to mention that it's worth quite a bit of money.
Consider the case of WordPerfect, which became the dominant word processing application by foregoing the copy protection which other such software was using. WP tacitly encouraged software piracy in order to build their install base; open source just extends that principle. The greater market penetration you have, the greater chance of an organization (because we all know, corporations are where the big money are) standardizing on you and putting money in your pocket for support.
It is possible that they could go to someone else for support or customization. Which is a fair risk. But I doubt that that would be the first choice; very few other companies are likely to know the application as well as you do.
Would Red Hat would make money if it wasn't open source? I don't think so; they'd have to deal with the expense of developing and maintaining a fairly huge body of software in-house. Operating systems development is not for the weak: it takes money, it takes bodies, and it takes time. Going open source, while it doesn't make for a guarantee that bodies and time will come your way gratis, at least gives you a decent chance.
In Canada, as well, people can do business under their own name without restrictions (they have to register a "fictitious name" if they wish to do business under another name).
It's not the same as a trademark, of course, but if you can show that you've been doing business under said name, even the jellyfish at the NIC would be hard pressed to take it away from you.
I think that you're taking natural selection in the wrong way. Poor eyesight or what have you are compensated for in the modern world. There is no reason to select against poor eyesight, unless you're striving to build the ubermensch. (snarf)
Singer is, as you might guess, a strong utilitarian. Everything, in his view, must be measured against its consequences. His support for animal rights is reasoned in that the benefit of using animals for food, clothing, and such is overwhelmed by the suffering that such use causes to animals.
Likewise, the suffering that a life as disabled might cause the individual must be measured against the benefit (or potential benefit) of their lives, the pain that they suffer must be balanced against the pleasure that they will experience. Singer is a renegade in the academic community for saying so - but I don't see how you can adopt utilitarian ethics without saying so.
Let's look at the example above: poor eyesight. I have very poor eyesight, and I wouldn't pretend to say that it's no inconvenience at all - it is a horrible inconvenience. Enough so that life is not worth living? Of course not. The calculation, however, is not one that you can take for granted.
I went to a day-long workshop at CASCON'97 which dealt with the use of communities in ecommerce. I'd think that VRML, intelligently used (and with properly wide net connections), would provide an excellent platform for ecommerce situations.
Not that this is entirely good news for everyone (especially the segment of/.ers that looks down their nose at ecommerce), but a VR world has the capacity to be quite engaging, to build community and allow for both synchronous and asynchronous interaction.
No doubt people who want to sell us stuff (whether it's advertising pageviews - yes, those are bought and sold - or conventional goods and services) will jump on the bandwagon, since the longer they can keep you there means the longer that they have to pitch to you - and that means greater sales.
Slashdot Mirror
Any score which is relative to the body of test takers is normalized every sitting by definition (ie, if you have the average student writing in a sitting getting 500 by definition, then the average student writing in a session will always get 500).
Reweighting or changing the statistical model, OTOH - that's something else entirely. If I were in the college board's shoes, I'd see it as something regrettable (because a lay audience just won't get it) but totally necessary to protect the integrity of one's rating scale.
--
I'd think that the difference is that any company can pick up a competitor's product.
Still, the innovator or developer of any product has three advantages when it comes to selling.
They are the most familiar with the existing code base: this is very important when it comes to porting a given piece of software. RMS, for example, charges usurious rates to port GCC. Any complex piece of software is going to take a considerable amount of time and thus expense to familiarize oneself with.
They hold the copyright, and can offer alternative licensing terms. This may be important to some companies, who don't want an "open" piece of software or who may want to redistribute their modifications without including source and who will pay for source access.
Finally, they gain name recognition, credibility and thus consumer preference through the product's use. Even though Jim Joe Bob can put togeter a linux distribution based on Debian, they certainly aren't Debian and don't have the credibility of Debian. Red Hat is an even better example of this.
--
I think that they renormalize every year. That's the point of the test - to produce a ranking of test takers where the top scorers receive an 800, the bottom scorers receive a 200, and the median scorers receive a 500. This calculation is done by means of standard deviation from the mean; an 800 does not mean a perfect score, but it does mean a score in the 99th percentile of all test takers.
Why they'd change the scores would be because the median student would be because the distribution itself has changed shape. Hypothetically, we have a normal distribution (bell curve) which represents the scores of all test takers, and the apex of the curve is at 500.
If you had the median student scoring below 500, you'd have a heavy-tailed distribution - the bell curve is warped to one side (in this case, the tail would be heavy to the right) and the model has to be adjusted for that.
It's basic statistics, and I'm surprised that it's so widely misunderstood.
--
Do elite colleges matter? Are the worth the money? Consider:
In 2000 we will most likely have a choice between Harvard (Gore), Yale (Bush), or Princeton (Bradley).
Of our nine current Supreme Court justices, four attended the same undergraduate college (Stanford). Of the five others, we find graduates of Harvard, Cornell, Chicago, Georgetown, and Holy Cross. Five out of the nine attended a single law school (Harvard); the other law schools are Stanford (2), Yale, Columbia, and Northwestern. (Ginsberg attended Harvard Law for two years, then transferred to Columbia). Not one attended a public university or college, or an obscure private school.
Of the top four cabinet positions:
State: Wellesley (Albright)
Defense: Bowdoin (Cohen)
Treasury: MIT (Summers, replacing Harvard (Rubin))
Attorney-General: Cornell (Reno)
Similar patterns, although not as extreme, can be found in business, the arts, etc. Keep in mind that only about 1-2% of all college graduates attended elite private colleges. Because larger numbers of Americans are attending (mostly public) universities and colleges, the percentage attending elite institutions has probably declined over time, which makes the disproportionate success of elite graduates even more striking.
(from the LBO-talk mailing list, posted by Kelley Walker)
--
Any college anywhere has the "right" to choose whom to accept. That's the purpose of an admissions process. However, most universities, and especially private ones, realize that while undergraduates pay the bills, faculty and graduate students make the reputation, and that's where the effort goes. Perhaps you had a good experience as an undergrad at CMU but it's certainly not generalizable.
I think that eCollegebid is an excellent way of running a clearings process, although perhaps not a primary admissions process.
Universities in the UK run their clearings process this way: when the second round of admissions is done and all students have committed or declined, they post the program vacancies and invite applications to the remaining spots rather than take students who don't meet their standards or run programs half full.
You don't bid on tuition in the UK, but for private schools in the US that should definitely be an option; they have the option of subsidizing you or not, just like they would have the option of offering you a scholarship or a bursary if you participated in the primary applications process. If they're already running a program, the incremental cost for additional students is quite small.
If collegebid became widely adopted, I think that it would definitely fill a niche in the US' private educational program. If you don't like the quality offers you get in the primary process, decline them and attempt to bid up. If you can't afford tuition where you did get in, try to bid down for schools of lesser quality. This freedom isn't presently available, but in a private system perhaps it should be.
--
It might be wonderful but still I doubt that any more than a tiny fraction of the internet community would use encryption. It's changed from the days where traffic was mainly academic; now the "average user" is a suburbanite with no grasp and no willingness to inform him or herself about issues of privacy, security, cryptography or the like.
Of course, that may not be necessary; it might just be necessary to encrypt a small fraction of network traffic to result in changes at Echelon. Of course, Echelon is reportedly a global operation, so it's necessarily a slow process. Also, I doubt that Echelon is the only communications intercept operation out there. It's something that any government or even any multinational corporation would be interested in, no matter where they're located.
However, I don't think that's as relevant as other considerations. One should be encouraging prudent and secure communications for one's own good, not just to thumb one's nose at the US government. Make your PGP or GPG key available; use encryption routinely, and expect your correspondents to reply in kind. In other words, take responsibility for your own security.
--
I wouldn't go that far. It's certainly not a bad thing, and definitely not pathologically manipulative, to want your nearest and dearest to be running a decent computer.
Nor are many octogenarians going to be using a computer at all. You aren't denying someone tech support (more to the point, you're setting up the computer to require little if any tech support) from their peers if their peers aren't going to be offering support.
Likewise, I don't think that the "community of users" argument follows. It takes an effort to seek out and belong to any community, and most people who aren't geeks really have no desire to belong to a community of computer users. The extent that they *will* belong to a community of users is through things like email, will be with people whom they have pre-existing social relationships with, and it'll be primarily social, not primarily technical.
--
I recall that the (inter|intra)net-connected server market was NT's big marketing goal only a couple of years ago. But I might recall incorrectly.
... but YMMV, once again).
Still. If you need NT to provide a decent single user workstation it's a great failure of design; it should be massive overkill given its technical roots (VAX, OS/2). Rather, if all NT can provide is a decent single user workstation, it's massive overkill. Which is fine so long as you have the resources, and most corporations can and do fork up the cash on an ongoing basis.
This specific report is referring to a single sector of the computing market: business productivity platforms, or desktops running MS Office. There's no port of Office and no analogous system which provides 100% compatibility (and/or superior functionality), so the verdict will be negative, will have to be negative. It's phrased in such a way that there is no alternative.
Is that dishonest of the Gartner group? Well, yes, it is; it implies that a valid comparison is possible. The preconceptions intrinsic to the question are (or seem to be) so strong that starting from a level playing field is impossible. Nor would it be possible to fairly compare other unices/unix-like OSs on the same question.
At the same time: it's nothing that a linux user doesn't already know, and it's nothing that any current linux user should worry about. (Nor does it say anything that you couldn't find out in ten minutes of asking questions and searching and thinking
--
You have heard about displaying an xterm from the server box on your own desktop. Or perhaps you haven't.
Whether you have or you haven't, it's one of the biggest and best real-world uses of X - you can run X applications on a distributed model and display them locally.
--
Symptoms, no; they're different disorders (although they are differential diagnoses of each other).
/. does research in this field; I'm merely an educated layperson.
What has been suggested quite recently in clinical research is that ADHD/ADD as a child or juvenile increases the risk of developing bipolar disorder later in life (bear in mind that the average time of diagnosis for bipolar disorder is in one's late 20's).
One could make arguments as to whether there is a shared cause or that the former causes the latter through some biochemical or even through a social mechanism. Nobody knows what causes either ADD/ADHD or bipolar disorder; it's even possible that there are multiple possible causes producing the same symptomology.
I used to favour the explanation that there was a genetic component to all cases of bipolar disorder; however, right now I would have to demur. Perhaps someone on
--
Don't be silly. Everything is a matter of risk.
The example of quarantine (which I suspect you don't know much about) is a fine example. People with TB, even today in some cases, are quarantined until they get sufficiently better so that releasing them is a good risk.
Of course, it wasn't the existence of quarantine laws that stopped widespread tb; it was the existence of effective antibiotics. People who were badly enough off to go to a tb sanatorium were by and large already bedridden.
HIV's spread isn't going to be confined by quarantine; we know that people can be asymptomatic for years, decades even. Those whom we have the most to fear from are people who carry HIV but haven't yet developed AIDS, who haven't been tested (most people aren't, unless they're blood donors), and who thus haven't adopted safer sex practises.
They wouldn't be caught by a medical dragnet through online records. They wouldn't be caught by quarantine laws. They would be caught by better sex education, which would both keep them from getting it in the first place and from transmitting it in the second.
Having online and fully accessible medical records strikes me as so much expensive but trendy claptrap. Compare the huge effort of putting records for 300 million people securely online to the relatively tiny cost of adequate and universal public health. Of course, adequate and universal public health won't make headlines and won't make millionaires - so it's neglected.
--
I studied statistics at university, and that kind of large-scale correlative study that you described was touched upon.
However, this kind of thing - online medical records - does not lead directly to large-scale longitudinal studies. If anything, the restrictions on use of data will render such studies impossible. Likewise, reporting standards are likely to be variable at best. The data, if you can get it, will be only mildly usable. Big correlations and interrelations will be visible; subtler ones will not.
IMHO, of course; I'd like to have that much data at my disposal, too. But it's much more likely to happen through big (and, yes, expensive, but good data is always expensive) longitudinal studies, not through electronic record keeping.
--
On the contrary.
Firstly, someone who is said to engineer a system increases the information represented within. This is one of the creationist pseudo-arguments - it doesn't work there, but it works here.
Secondly, the sanctity of the fertilized egg (assuming you recognize such a sanctity) is undisturbed. The engineering merely must be done prior to fertilization. I've never seen anyone claim that the unfertilized egg or sperm is a legitimate life.
That said, you do need to confront items of Catholic doctrine, I believe, which deal with the integrity of normal conception. However, they're different from those which deal with the termination of pregnancy.
--
I recall reading an interesting study which looked at the positive factors for economic success.
Being white was positively correlated to economic success. No surprise there.
Being male was positively correlated to economic success. No surprise there, either.
Going to a good college was positively correlated to economic success. Again, no surprise.
However, intelligence/academic success was only mildly correlated to economic success. What the study found was that the highest earning group was that the stupidest white males who went to top colleges were in fact the top earners.
Admittedly, we're looking at a complex system; very smart people in colleges, especially top colleges, are tempted by academe, which pays poorly. Still, I don't think that a race of genetically engineered super-men, who dominate positions at the head of industry and government, is in the offing.
That said, the expectation is probably much more powerful than the reality. If we expect super-men to be the best leaders, no doubt they'll be picked early as leaders, given the necessary background for leadership, and then moved up to positions of leadership. No doubt a disproportionate amount of the best leaders will come from this group.
It will hardly amount to proof of their superiority, however. It is merely proof of the natural credulity of the human animal.
--
The main reason is to give a hierarchy for authoritative lookups. You could certainly place all domains in a unified namespace (bear in mind that the namespace would shrink proportionally), but you'd have to have a single authoritative register.
.com, .org, .net, .edu, etc etc, but they don't have to - they could just as easily spin off the edu's, for example, to the NSF, and that would work seamlessly. You could even have one company holding the authoritative server for one tld and one tld only - no problem there. When you *don't* have tld's, on the other hand, you have to figure out some other method to apportion and split authority, or you have to stick them together.
Um, no thanks.
The way it is now, NSI has
Remember that the existence of tld's may not be (and isn't) an advantage when it comes to database searching, but it's certainly an advantage when it comes to determining the proper progress of nameserver authority. And that's the sine qua non of DNS - the existence of a proper progression of authorities for any given name.
--
You can get bankbook updaters in Canada, too. :) They're not uncommon, at least in major centers (I live and work in Toronto).
--
I bought a 128Mb stick of SDRAM for $100 Canadian (~= 69$US) at the absolute bottom market; I'm glad I did. Admittedly, that was a used purchase, but the multiplier in the great white north was about .75$US per meg for new RAM.
--
A classy response from what seems like a rather classy guy. Plaudits to Jane's.
--
I doubt that Silicon Valley is on a totally different banking system than the rest of America (where people who don't know better are legion). If SV is, feel free to disregard my comment.
The main problem for direct billing systems - well, the main problems - are twofold.
First, we have the problem of liability. Banks, especially banks, have to know who's going to pay if something goes wrong. They like to divide up and preferably cap liability before such a system goes into operation. Bear in mind that many large organizations, such as banks, self-insure (that is, rather than pay a certain amount of money to an insurer, they are big enough to merely set money aside and act as their own insurer), and that increases the stakes.
Second, we have the problem of incompetence. Most software projects are failures to some extent - nearly a third get trashed because they don't satisfy spec, or the direction of the corp changes, or they're just plain not used. This doesn't mean that online banking services isn't a possibility, but you don't want to fail while doing this kind of thing. In other words, it's likely that online banking will remain in beta for half of forever.
To the second point you can add the observation that the cheque-clearing and banking laws in the US drastically and desperately need rationalization - there are few provisions for banks with no physical presence to operate in an arbitrary state. But YMMV.
--
The japanese case is probably because Japan has fallen behind North America and Europe when it comes to networking and the like. It's more a cultural thing than anything else; Japan does very well when it comes to other aspects of computing.
The world has quite a bit of variance when it comes to access to computing facilities - really, when it comes to telecom facilities at all. I recall seeing a map which showed that most of the world has only limited access to the internet - most of africa, for example, is UUCP-only. No doubt that's changed in the intervening years, but the problem remains that almost nobody is going to make money off wide availability of internet services to that part of the world, so you're unlikely to see it happen.
--
I don't think that he was a founder (IIRC, the guy who designed the UltraSPARC for Sun is), but he's certainly a funder :).
(Rather like the story of John Harvard and Harvard University, I guess, except they didn't name the shop after him.)
--
Well, you can look at it from two perspectives: from that of a hardware vendor and that of a software vendor.
The hardware vendor should definitely go open source. Most likely they have patents on their hardware; while the source may give someone an edge in reverse engineering their hardware, it's not really that much of an edge, nor will it be a *competitive* edge. That is to say, no one is going to compete with Company X if all they do is copy what Company X does, merely two years later. Technology moves too fast for that.
For software companies, it's a different paradigm. Consider that most big companies, such as IBM, make a substantial chunk of their software dollar by providing systems integrations and customization on said software. Companies like IBM have more work than they can bother to do - that's why they develop "business partner" networks. There's substantial money to be made by going open source with your product and making money by supporting it. Is it sufficient to give away the software in its entirety? It is for some, probably not for others.
Likewise, very few companies are going to take your custom job, with open source, and distribute the source to it, even if they have the right to. Their competitive edge is tied into having your software - not to mention that it's worth quite a bit of money.
Consider the case of WordPerfect, which became the dominant word processing application by foregoing the copy protection which other such software was using. WP tacitly encouraged software piracy in order to build their install base; open source just extends that principle. The greater market penetration you have, the greater chance of an organization (because we all know, corporations are where the big money are) standardizing on you and putting money in your pocket for support.
It is possible that they could go to someone else for support or customization. Which is a fair risk. But I doubt that that would be the first choice; very few other companies are likely to know the application as well as you do.
Would Red Hat would make money if it wasn't open source? I don't think so; they'd have to deal with the expense of developing and maintaining a fairly huge body of software in-house. Operating systems development is not for the weak: it takes money, it takes bodies, and it takes time. Going open source, while it doesn't make for a guarantee that bodies and time will come your way gratis, at least gives you a decent chance.
--
In Canada, as well, people can do business under their own name without restrictions (they have to register a "fictitious name" if they wish to do business under another name).
It's not the same as a trademark, of course, but if you can show that you've been doing business under said name, even the jellyfish at the NIC would be hard pressed to take it away from you.
--
I think that you're taking natural selection in the wrong way. Poor eyesight or what have you are compensated for in the modern world. There is no reason to select against poor eyesight, unless you're striving to build the ubermensch. (snarf)
Singer is, as you might guess, a strong utilitarian. Everything, in his view, must be measured against its consequences. His support for animal rights is reasoned in that the benefit of using animals for food, clothing, and such is overwhelmed by the suffering that such use causes to animals.
Likewise, the suffering that a life as disabled might cause the individual must be measured against the benefit (or potential benefit) of their lives, the pain that they suffer must be balanced against the pleasure that they will experience. Singer is a renegade in the academic community for saying so - but I don't see how you can adopt utilitarian ethics without saying so.
Let's look at the example above: poor eyesight. I have very poor eyesight, and I wouldn't pretend to say that it's no inconvenience at all - it is a horrible inconvenience. Enough so that life is not worth living? Of course not. The calculation, however, is not one that you can take for granted.
--
I went to a day-long workshop at CASCON'97 which dealt with the use of communities in ecommerce. I'd think that VRML, intelligently used (and with properly wide net connections), would provide an excellent platform for ecommerce situations.
/.ers that looks down their nose at ecommerce), but a VR world has the capacity to be quite engaging, to build community and allow for both synchronous and asynchronous interaction.
Not that this is entirely good news for everyone (especially the segment of
No doubt people who want to sell us stuff (whether it's advertising pageviews - yes, those are bought and sold - or conventional goods and services) will jump on the bandwagon, since the longer they can keep you there means the longer that they have to pitch to you - and that means greater sales.
--