Although some of his arguments are not convincing or even insulting ("Did Israeli astronaut Ilan Ramon really have to be there to push a couple of buttons...")
Perhaps this is a little bit insulting, but he makes some good points about Space Shuttle Science. Particularly with regards to the types of experiment that can be performed in the short time-frame available to Shuttle astronauts, and also regarding the need for human involvement in many of these experiments.
Can anyone provide a convincing rebuttal to Mr. Easterbrook's contentions?
"This is an unusually large amount of plutonium to be unaccounted for, which makes me uncomfortable, although I think it's highly unlikely that it was stolen," said Tatsujiro Suzuki, senior research scientist at the Central Research Institute of Electric Power Industry.
and...
The ministry admitted it was baffled by the remaining 86kg of absent plutonium, but insisted it was "not missing". Initial projections for plutonium output might have been too high and the missing amount might never have been produced, the ministry said.
So essentially, the experts express their discomfort with the figures but say that they're "pretty sure" the Plutonium couldn't have been stolen (based only on the plant managers' evaluation of plant security, which might not be the most objective measure). And you express that with the blanket assertion that the Plutonium "cannot have left the factory by these means."
There is a large difference between banks and physical applications.
Banks work with exact numbers, when you measure you don't. You calculate with unexact values.
1. Clearly if they failed to notice the discrepancies until they reached 206kg, so there must not have been a lot of day-to-day measurement. One assumes that they would have adjusted their predictions prior to this point if they were consistently seeing less output than expected.
2. In the absence of such day-to-day measurement I stand by my point. If someone had the opportunity to embezzle several kg of plutonium, there would be a reasonable chance that the plant managers would never find out (if they did find out, they could sound the alarm to the authorities, investigate their workers, etc.) So there's essentially one less line of security between the plant and potential purchasers than I'm comfortable with.
The plant security may be more than adequate. But if anyone can find a way to penetrate that security and hijack some Plutonium, this plant has advertised a huge vulnerability.
This is the kind of thing that makes me want to scream at the breeder reactor advocates. We can't seem to keep track of the relatively small amount of Plutonium that we generate. Imagine if we move to a Plutonium-economy and have to ship the stuff from place to place.
These discrepancies do always occur since no measurement is perfect. An error margin of 1% can be attributed to the measurement alone. The factory had a discrepancy of 3%.
If banks advertised regular "it's no big deal" accounting errors of +-3% in large bills and bearer bonds, I imagine you'd see a lot more embezzlement than you do today.
Xupiter has been around for a while. And it's NOT hard to get rid off: http://www.xupiter.com/uninstall That's it. Way to overreact guys.
I don't know about this week's version of the uninstaller, but previous versions were nice enough to leave behind big chunks of the program. Still running. Sort of the way a tick will leave its head behind if you yank it out with tweezers.
This is a pretty common and ugly tactic among spyware developers.
This little beastie's been around for a while now; when I first got infected with it, most of the removal sites I found on Google were in Asian character sets, so I suppose it was pretty big over there first.
The problem with these damn things is that you never quite know how they got onto your machine. I'm always very careful about what I install, and which dialog boxes I say "OK" to, but there's always the possibility that I accidentally let something slip through. I suppose that's why people aren't 100% sure if it can install itself right from IE without confirmation.
I'm getting increasingly sick of using IE, but I'm constantly running across sites that Mozilla just can't handle properly (or swiftly). And yes, I've cranked up the security level, though god knows why there exists any level of "security" that would allow unconfirmed installs.
The constitution doesn't give them much wiggle room since it assigns Congress the right to regulate copyrights to their hearts desire.
The Constitution does no such thing. Read the damned thing before you comment. The Copyright Clause requires copyright terms to have "limited times". A term that can be repeatedly extended (retroactively) at the whim of Congress is not limited.
The issue here is: does the Supreme Court have the right to enforce this sort of Constitutional limitation on Congress, or does that responsibility fall to Congress itself? The traditional answer is "Congress can do whatever it wants and the Court must restrain itself". In 1995, Justices Rehnquist, Scalia, Thomas, O'Connor and Kennedy threw this understanding on its head and declared that they have the power to enforce the restrictions on enumerated powers (the right to grant copyrights is one of these.)
The plaintiffs in Eldred asked the Court to apply this plain-stated logic with respect to the copyright clause, and the Court's response was... well, nothing. They didn't even bother to explain why they would do it in some circumstances and not in this one.
Eldred & Co had a strong Constitutional argument for limiting Congressional power if the Court was willing to obey its own precedent. This Court just didn't feel the need to do so, or even to explain why this case was different.
Project Gutenberg doesn't mind markup lanuages. They just like to have a copy of the plain ASCII as well.
There are a bunch of folks on the Gutenberg mailing lists who are ostensibly trying to settle on a markup language and a setup for quickly typesetting books out of it.
I don't know what kind of progress they've made. Last time I tuned in, they were getting nowhere because the idea of going through PG's back-catalog and adding markup was too daunting (though it's not gonna get any easier...)
The recording labels are not collaborating on a music download service. It is music retailers that are doing this.
You're right. My mistake.
I recall an article about the formation of one of the download services (pressplay?) that was a collaboration between RIAA labels. The article mentioned that their lawyers were acutely aware of antitrust laws to the point of not even allowing executives of the different companies in the same room when discussing the service.
IMHO, to illegally collaborate, the labels don't need to work together on the music download service. They only need to collaborate in their efforts to push outside competitors out of the business. One that's done, they can divide up the space by themselves without any further communication.
If someone can prove that they deliberately collaborated on the first step, their later efforts to avoid anti-trust violations don't necessarily matter.
So the point of this post is: why not ask publishers for the material? If it's already public domain, it's not like they'll lose profits, and maybe Project Gutenberg could let them put a little
Yup. Except that the vast majority of publishers won't give out their digital masters, even if the work in question is public domain. The formatting and page layout cost them money, and they (rightly or wrongly) feel that such a release would undercut their sales.
And even if you could get hold of the digital representation, it'd very likely be copyrighted as a "derivative work" (due to the layout info, page numbers, and even spelling corrections).
Didn't we just have a set of articles on Distributed Proofreaders? Those guys are harnessing technology to churn out books at a mad rate. Seems to me that Wired's reporter is maybe just a tad uninformed.
In any case, the real obstacle to a useful electronic library isn't labor. It's copyright.
A different situation
on
Kazaa Fights Back
·
· Score: 2, Interesting
That antitrust claim probably won't work. For one thing, Napster tried it, and it didn't work for them.
Yes, but I think that the new issue here is the Recording Industry's attempts to promote its own download services in competition to things like Kazaa. It's one thing for the major labels to protect their copyrights, it's another for them to use their copyrights as a lever to put companies out of business so they'll have no competition in a business area that they've already entered.
The hotel case didn't pack the same argument. Even Napster didn't have the same case, because at that time the music industry hadn't clearly demonstrated its desire to control the music download business. The situation is different now; for instance, six major labels just announced that they're collaborating on a music download service, which creates a much more compelling case that the labels are leveraging their copyrights to advance their own distribution businesses. This is further evidenced by the fact that many of these label-controlled download businesses have failed to pay royalties and yet the music industry has neglected to use its legal powers against them.
I doubt that the conservative court system is going to buy it, but I do think that Kazaa has a valid argument in this case. It's an argument that's going to become increasingly valid until it's obvious to even the most skeptical among us.
If you violate my copyright, then I want you punished
You seem to be under the impression that Federal Law lacks provisions for punishing copyright infringement. Wouldn't it be better if we used some of those laws instead of twisting draconian electronic "theft" laws to serve the purpose?
If you, as an artist, really want to preserve control over your works, then you'd better not support anyone who wants to go down this road. Most Americans understand the necessity of copyright, but have very little tolerance for it where it impacts their lives. Nailing "regular people" with excessive punishments for the crime of file-sharing is a great way to push wipe away that last vestige of respect and make Americans really hate copyright holders.
Enough of that and you can pretty much forget about controlling your work.
Look around at the applications that people run in corporations -- 30 year old COBOL apps, 10 year old UNIX apps, 15 year old XBASE apps that have been ported to the latest Visual FoxPro, 7 year old Visual BASIC apps that may or may not have been rebuilt with a recent version, DOS applications from the 80s, things that use BTrieve. All of this stuff is very real and in production.
And everything you say makes sense, if you're talking about servers. But we're not talking about servers, we're talking about desktops.
To carry your argument through to its logical conclusion, you'd see millions of employees interfacing with those machines using 20-year old clients: teletypes, maybe early CRT machines. Yuck.
But you don't see that. Go into the average large business today, even one that makes use of some ancient customer app, and you'll see a bunch of relatively modern Windows boxes interfaced through telnet or some web interface. And that's the key: does a free client OS have a decent telnet or web interface? Then why in god's name should I spend $xx for all of those clients?
DAT is inherently a heck of a lot more complex than analog cassette
Sure, but if you read my post you'll see I was comparing them to CD players, not analog tapes. CD players may still be slightly less complex than DAT decks, but not outrageously so. And CD players retail for a fraction of what DAT decks go for.
The point remains the same: DAT decks could have sold for reasonable prices (way less than the prohibitive $500 barrier) if they'd really taken off.
You get twenty-eight years to milk something for all the millions it's worth, AND you get to crush utterly and punitively anyone who dares steal even a penny's worth from you? Sounds like a good deal to me.
The record companies are confident that they can get all that and a bag of chips (er, 95+ year copyright terms, that is.) So why settle for what's behind door number three when you can get doors number one and two at the same time?
I'm no Linux zealot, I'm just trying to explain why the "Linux expects the consumer to come to the product" argument is crap. However:
if you already have Windows or a Mac, and nothing compels you to abandon it, why would you spend money and time converting to an OS that merely tries to mimic what you already have?
You pose your argument as though Linux is asking users to throw out their perfectly good machine and set up a brand new one. That's not it at all. Computers and OSes are an inherently short-lived product. All the Open Source movement needs to do is convince you to make your next one Linux, rather than paying a bunch of money for something with essentially the same features.
I just helped a friend of mine buy a laptop for grad school. With XP and the latest version of Office, she wound up paying several hundred dollars-- more than 25% of the price of the machine itself! Linux and OpenOffice aren't good enough yet that I would really recommend them to my friend, but in five years they probably will be. Why in god's name should she pay that much money for something that she could get for $50 on a CD with a support contract?
I'd say this argument is particularly compelling if you run a business. When Linux gets "good enough" to replace Windows on the desktop (and I don't think it's quite there yet), it'll be awfully hard to justify spending thousands to millions on commercial licensing fees when you could get the same thing for free (or cheap, with a support contract).
I'm thinking particulary of all of those industries where each of ten-thousand employees sits in front of Windows box doing basic word-processing, data-entry or web-browsing. If Linux becomes as easy-to-use as Windows, sending million dollar checks to Microsoft will be increasingly difficult to justify for this purpose (not to mention the hassle from MS licensing). And employees tend to get used to the OS they use at work, which often encourages them to use it at home.
Anyone who wants to see Linux become a mainstream consumer product should try to answer these questions: If Windows/Mac reliably meets your needs, and you aren't interested in open versus proprietary software issues, and you buy shrinkwrapped software rather than do installs off the net, why would you want Linux?
Again, you could flip this the other way. If Linux reliably meets your needs and is cheap or free, why would you spend a lot more money on a commercial OS?
I imagine the reasons are a) inertia, and b) compatibility with existing apps. Inertia's a really weak premise to base superiority on, and comptibility might well be fleeting. As the above article points out, Beta once had 100% of the market, but something cheaper and more convenient came out and knocked it down.
It's much worse than just taking money away from the farmer so he can "buy more things." Farmers go into debt at the beginning of each year, and basically hope to make enough money out of their crop to pay that debt off and pay their bills for the rest of the year.
Well, hopefully there are enough circle-watchers in the world that the farmer can sell tickets and T-shirts to recoup some of his losses.
"Beware the bearers of FALSE gifts & their BROKEN PROMISES. Much PAIN but still time. BELIEvE. There is GOOD out there. We OPpose DECEPTION. Conduit CLOSING. Acknowledge."
My favorite bit:
Richard Brain was unable initially to identify the word "BELIEvE" (initially seen as 'EELIJ?E'), saying he thought it was a corrupted word. The problem was that he read the first 'B' as an 'E'. The binary code for E is 01000101. The binary code for B is 01000010. In the field spiral (see marked photo) there is an encoding error (marked with a red X in the photo above right)! This letter has nine individual bits and appears to read 010000101.
I've never seen a DAT deck that was under $500 list
DAT decks never really got produced in large enough numbers to compete with CD players or other consumer goods. That certainly isn't the only reason they were more expensive, but it's a big one. Same thing for the Minidisc and other technologies.
This just makes me laugh and laugh and laugh. I mean, to the larger portion of the educated population, this is so obvious as to be not even worth mentioning.
Actually, most of the people here are making a valid point, which is that this columnist totally misses the point. He's trying to take geeks to task for focusing on technical superiority... and then he's going on to focus on what? Technical superiority, just in a different aspect (longer tapes, which IIRC, were also eventually available for Beta.)
Quite clearly the issue here was cost, which was largely driven by Sony's "no-license" business model. If you're going to write a snotty article deriding geeks for looking at too narrow of a picture, it really doesn't help to be guilty of the same.
In other words, they expect the customer to change rather than the product.
Try the following. Grab a computer and install a version of RedHat linux from 1999. Now install the latest version. You'll notice a phenomenal difference between the two products.
The more recent version will have a simple, pretty graphical installer that recognizes just about any hardware and self-configures. It'll have a nice desktop interface that's clearly modeled after Windows/Macintosh. It'll have an office suite designed to be comfortable for someone who's used to MS Office. Almost all of the day-to-day configuration issues (think editing text files) from the 1999 version will have been moved into simple-to-use control panels accessable from the desktop.
Sure, the current version isn't perfect, and it may not be enough to convince most users to switch. But to claim that Linux "expects the customer to change rather than the product" is to set up a strawman that has little to do with reality.
Reverse lookups go by sending a PTR request containing an IP address to a DNS server
Absolutely right. I was looking at an older, unused mechanism for doing reverse lookups that includes an "IQUERY" status in the header. The actual mechanism is as you say-- it's actually a QTYPE=PTR message and the addition of in-addr.arpa to the IP address.
Leads to the same conclusion, though. Sending an IP address without the QTYPE=PTR and in-addr.arpa add-on is pointless and wastes server bandwidth.
Ah, so the problem is that some software that should be recognizing an IP as a clue that a reverse lookup is required is instead sending the request as a "standard" query?
Or, equally likely, that some software that should know the difference between an IP address and a domain name isn't kicking in.
For instance, if I enter http://87.23.54.34 (made that up) into my browser, the browser/network stack should be smart enough not to issue a DNS query. Even if my computer is dumb, the local DNS server should recognize that there's no need to pass that query on to the root server.
If those two things aren't happening in a certain percentage of machines/ISPs, then it's not surprising that the root servers are being deluged with these pointless requests. Probably doesn't take a lot of CPU on the root server, but it does eat up bandwidth.
Is it just me, or is this a description of a reverse lookup? How does that qualify as unnecessary?
I believe that reverse lookups are identified by an "inverse" status flag in the request header. One can only assume that the authors were not counting this sort of valid query, and were only focusing on the "standard" queries that contained IP addresses. Those certainly would, I think, be rather pointless.
About 12 percent of the queries received by the root server on Oct. 4, were for nonexistent top-level domains, such as ".elvis", ".corp", and ".localhost".
And that's a problem? My understanding was dealing with this sort of thing was exactly the purpose of the root DNS servers. If every ISP's DNS server was pre-configured to recognize valid and invalid top-level domains, you could just set them up to go straight to the specific DNS servers handling those domains (.com,.net,.org, etc.) There would be no need for a root-level system.
The argument for allowing this kind of cracked query through to the root server is that it makes it easy to add new domains (.elvis,.corp, what have you) without forcing everyone to reconfigure their DNS boxes for each new top-level domain.
Slashdot Mirror
Perhaps this is a little bit insulting, but he makes some good points about Space Shuttle Science. Particularly with regards to the types of experiment that can be performed in the short time-frame available to Shuttle astronauts, and also regarding the need for human involvement in many of these experiments.
Can anyone provide a convincing rebuttal to Mr. Easterbrook's contentions?
There is a large difference between banks and physical applications. Banks work with exact numbers, when you measure you don't. You calculate with unexact values.
1. Clearly if they failed to notice the discrepancies until they reached 206kg, so there must not have been a lot of day-to-day measurement. One assumes that they would have adjusted their predictions prior to this point if they were consistently seeing less output than expected.
2. In the absence of such day-to-day measurement I stand by my point. If someone had the opportunity to embezzle several kg of plutonium, there would be a reasonable chance that the plant managers would never find out (if they did find out, they could sound the alarm to the authorities, investigate their workers, etc.) So there's essentially one less line of security between the plant and potential purchasers than I'm comfortable with.
The plant security may be more than adequate. But if anyone can find a way to penetrate that security and hijack some Plutonium, this plant has advertised a huge vulnerability.
This is the kind of thing that makes me want to scream at the breeder reactor advocates. We can't seem to keep track of the relatively small amount of Plutonium that we generate. Imagine if we move to a Plutonium-economy and have to ship the stuff from place to place.
If banks advertised regular "it's no big deal" accounting errors of +-3% in large bills and bearer bonds, I imagine you'd see a lot more embezzlement than you do today.
Just a thought.
I don't know about this week's version of the uninstaller, but previous versions were nice enough to leave behind big chunks of the program. Still running. Sort of the way a tick will leave its head behind if you yank it out with tweezers.
This is a pretty common and ugly tactic among spyware developers.
The problem with these damn things is that you never quite know how they got onto your machine. I'm always very careful about what I install, and which dialog boxes I say "OK" to, but there's always the possibility that I accidentally let something slip through. I suppose that's why people aren't 100% sure if it can install itself right from IE without confirmation.
I'm getting increasingly sick of using IE, but I'm constantly running across sites that Mozilla just can't handle properly (or swiftly). And yes, I've cranked up the security level, though god knows why there exists any level of "security" that would allow unconfirmed installs.
The Constitution does no such thing. Read the damned thing before you comment. The Copyright Clause requires copyright terms to have "limited times". A term that can be repeatedly extended (retroactively) at the whim of Congress is not limited.
The issue here is: does the Supreme Court have the right to enforce this sort of Constitutional limitation on Congress, or does that responsibility fall to Congress itself? The traditional answer is "Congress can do whatever it wants and the Court must restrain itself". In 1995, Justices Rehnquist, Scalia, Thomas, O'Connor and Kennedy threw this understanding on its head and declared that they have the power to enforce the restrictions on enumerated powers (the right to grant copyrights is one of these.)
The plaintiffs in Eldred asked the Court to apply this plain-stated logic with respect to the copyright clause, and the Court's response was... well, nothing. They didn't even bother to explain why they would do it in some circumstances and not in this one.
Eldred & Co had a strong Constitutional argument for limiting Congressional power if the Court was willing to obey its own precedent. This Court just didn't feel the need to do so, or even to explain why this case was different.
There are a bunch of folks on the Gutenberg mailing lists who are ostensibly trying to settle on a markup language and a setup for quickly typesetting books out of it.
I don't know what kind of progress they've made. Last time I tuned in, they were getting nowhere because the idea of going through PG's back-catalog and adding markup was too daunting (though it's not gonna get any easier...)
You're right. My mistake.
I recall an article about the formation of one of the download services (pressplay?) that was a collaboration between RIAA labels. The article mentioned that their lawyers were acutely aware of antitrust laws to the point of not even allowing executives of the different companies in the same room when discussing the service.
IMHO, to illegally collaborate, the labels don't need to work together on the music download service. They only need to collaborate in their efforts to push outside competitors out of the business. One that's done, they can divide up the space by themselves without any further communication.
If someone can prove that they deliberately collaborated on the first step, their later efforts to avoid anti-trust violations don't necessarily matter.
Yup. Except that the vast majority of publishers won't give out their digital masters, even if the work in question is public domain. The formatting and page layout cost them money, and they (rightly or wrongly) feel that such a release would undercut their sales.
And even if you could get hold of the digital representation, it'd very likely be copyrighted as a "derivative work" (due to the layout info, page numbers, and even spelling corrections).
In any case, the real obstacle to a useful electronic library isn't labor. It's copyright.
Yes, but I think that the new issue here is the Recording Industry's attempts to promote its own download services in competition to things like Kazaa. It's one thing for the major labels to protect their copyrights, it's another for them to use their copyrights as a lever to put companies out of business so they'll have no competition in a business area that they've already entered.
The hotel case didn't pack the same argument. Even Napster didn't have the same case, because at that time the music industry hadn't clearly demonstrated its desire to control the music download business. The situation is different now; for instance, six major labels just announced that they're collaborating on a music download service, which creates a much more compelling case that the labels are leveraging their copyrights to advance their own distribution businesses. This is further evidenced by the fact that many of these label-controlled download businesses have failed to pay royalties and yet the music industry has neglected to use its legal powers against them.
I doubt that the conservative court system is going to buy it, but I do think that Kazaa has a valid argument in this case. It's an argument that's going to become increasingly valid until it's obvious to even the most skeptical among us.
You seem to be under the impression that Federal Law lacks provisions for punishing copyright infringement. Wouldn't it be better if we used some of those laws instead of twisting draconian electronic "theft" laws to serve the purpose?
If you, as an artist, really want to preserve control over your works, then you'd better not support anyone who wants to go down this road. Most Americans understand the necessity of copyright, but have very little tolerance for it where it impacts their lives. Nailing "regular people" with excessive punishments for the crime of file-sharing is a great way to push wipe away that last vestige of respect and make Americans really hate copyright holders.
Enough of that and you can pretty much forget about controlling your work.
And everything you say makes sense, if you're talking about servers. But we're not talking about servers, we're talking about desktops.
To carry your argument through to its logical conclusion, you'd see millions of employees interfacing with those machines using 20-year old clients: teletypes, maybe early CRT machines. Yuck.
But you don't see that. Go into the average large business today, even one that makes use of some ancient customer app, and you'll see a bunch of relatively modern Windows boxes interfaced through telnet or some web interface. And that's the key: does a free client OS have a decent telnet or web interface? Then why in god's name should I spend $xx for all of those clients?
Sure, but if you read my post you'll see I was comparing them to CD players, not analog tapes. CD players may still be slightly less complex than DAT decks, but not outrageously so. And CD players retail for a fraction of what DAT decks go for.
The point remains the same: DAT decks could have sold for reasonable prices (way less than the prohibitive $500 barrier) if they'd really taken off.
The record companies are confident that they can get all that and a bag of chips (er, 95+ year copyright terms, that is.) So why settle for what's behind door number three when you can get doors number one and two at the same time?
if you already have Windows or a Mac, and nothing compels you to abandon it, why would you spend money and time converting to an OS that merely tries to mimic what you already have?
You pose your argument as though Linux is asking users to throw out their perfectly good machine and set up a brand new one. That's not it at all. Computers and OSes are an inherently short-lived product. All the Open Source movement needs to do is convince you to make your next one Linux, rather than paying a bunch of money for something with essentially the same features.
I just helped a friend of mine buy a laptop for grad school. With XP and the latest version of Office, she wound up paying several hundred dollars-- more than 25% of the price of the machine itself! Linux and OpenOffice aren't good enough yet that I would really recommend them to my friend, but in five years they probably will be. Why in god's name should she pay that much money for something that she could get for $50 on a CD with a support contract?
I'd say this argument is particularly compelling if you run a business. When Linux gets "good enough" to replace Windows on the desktop (and I don't think it's quite there yet), it'll be awfully hard to justify spending thousands to millions on commercial licensing fees when you could get the same thing for free (or cheap, with a support contract).
I'm thinking particulary of all of those industries where each of ten-thousand employees sits in front of Windows box doing basic word-processing, data-entry or web-browsing. If Linux becomes as easy-to-use as Windows, sending million dollar checks to Microsoft will be increasingly difficult to justify for this purpose (not to mention the hassle from MS licensing). And employees tend to get used to the OS they use at work, which often encourages them to use it at home.
Anyone who wants to see Linux become a mainstream consumer product should try to answer these questions: If Windows/Mac reliably meets your needs, and you aren't interested in open versus proprietary software issues, and you buy shrinkwrapped software rather than do installs off the net, why would you want Linux?
Again, you could flip this the other way. If Linux reliably meets your needs and is cheap or free, why would you spend a lot more money on a commercial OS?
I imagine the reasons are a) inertia, and b) compatibility with existing apps. Inertia's a really weak premise to base superiority on, and comptibility might well be fleeting. As the above article points out, Beta once had 100% of the market, but something cheaper and more convenient came out and knocked it down.
Well, hopefully there are enough circle-watchers in the world that the farmer can sell tickets and T-shirts to recoup some of his losses.
My favorite bit:
DAT decks never really got produced in large enough numbers to compete with CD players or other consumer goods. That certainly isn't the only reason they were more expensive, but it's a big one. Same thing for the Minidisc and other technologies.
Actually, most of the people here are making a valid point, which is that this columnist totally misses the point. He's trying to take geeks to task for focusing on technical superiority... and then he's going on to focus on what? Technical superiority, just in a different aspect (longer tapes, which IIRC, were also eventually available for Beta.)
Quite clearly the issue here was cost, which was largely driven by Sony's "no-license" business model. If you're going to write a snotty article deriding geeks for looking at too narrow of a picture, it really doesn't help to be guilty of the same.
Try the following. Grab a computer and install a version of RedHat linux from 1999. Now install the latest version. You'll notice a phenomenal difference between the two products.
The more recent version will have a simple, pretty graphical installer that recognizes just about any hardware and self-configures. It'll have a nice desktop interface that's clearly modeled after Windows/Macintosh. It'll have an office suite designed to be comfortable for someone who's used to MS Office. Almost all of the day-to-day configuration issues (think editing text files) from the 1999 version will have been moved into simple-to-use control panels accessable from the desktop.
Sure, the current version isn't perfect, and it may not be enough to convince most users to switch. But to claim that Linux "expects the customer to change rather than the product" is to set up a strawman that has little to do with reality.
Absolutely right. I was looking at an older, unused mechanism for doing reverse lookups that includes an "IQUERY" status in the header. The actual mechanism is as you say-- it's actually a QTYPE=PTR message and the addition of in-addr.arpa to the IP address.
Leads to the same conclusion, though. Sending an IP address without the QTYPE=PTR and in-addr.arpa add-on is pointless and wastes server bandwidth.
Or, equally likely, that some software that should know the difference between an IP address and a domain name isn't kicking in.
For instance, if I enter http://87.23.54.34 (made that up) into my browser, the browser/network stack should be smart enough not to issue a DNS query. Even if my computer is dumb, the local DNS server should recognize that there's no need to pass that query on to the root server.
If those two things aren't happening in a certain percentage of machines/ISPs, then it's not surprising that the root servers are being deluged with these pointless requests. Probably doesn't take a lot of CPU on the root server, but it does eat up bandwidth.
I believe that reverse lookups are identified by an "inverse" status flag in the request header. One can only assume that the authors were not counting this sort of valid query, and were only focusing on the "standard" queries that contained IP addresses. Those certainly would, I think, be rather pointless.
And that's a problem? My understanding was dealing with this sort of thing was exactly the purpose of the root DNS servers. If every ISP's DNS server was pre-configured to recognize valid and invalid top-level domains, you could just set them up to go straight to the specific DNS servers handling those domains (.com, .net, .org, etc.) There would be no need for a root-level system.
The argument for allowing this kind of cracked query through to the root server is that it makes it easy to add new domains (.elvis, .corp, what have you) without forcing everyone to reconfigure their DNS boxes for each new top-level domain.