"Why do I need to carry around this box? Why do I want to carry around data? That's what the Internet is for."
In theory, but not in practice yet. My home machine is on dialup, and data on it is not available when I am not there. Servers at work are for work, not personal use, and anyway I wouldn't put personal stuff on a machine where some BOFH has the root password unless I absolutely had to.
And this would still only work if whatever machine I was trying to use also had a broadband connection. "Hey want to see the photos I scanned at nice high resolution? Let's connect to the ftp server and come back in three days when it's done."
People recognize this problem. It is why they carry laptops around with them everywhere. I see people all the time using a laptop while sitting next to a desktop or an X server. This is the easiest way to carry their data around.
"In Starship Troopers, the voting franchise was restricted to Veterans (Either Military veterans, or Veterans of Federal Service) because they had demonstrated a willingness, and an ability, to put the interests of society above that of their own well being."
Except they haven't, and they don't. It is one of RH's simple-minded fantasies. This is not to say that many in the armed forces don't have a great interest in the good of society, but as a whole they are not uniquely gifted in this respect. Read some history. Read the news. Meet some people. You'll figure out that RH was full of it.
I found it interesting that in Grumbles from Beyond the Grave RH backtracked on this in personal letters at the time. He claimed that really all he meant was that we out to try a different system, because he thought that government based on the US Constitution was not working too well. He also suggested we try allowing only women to vote for a while. His infatuation with soldiers was a little too obvious for me to think he meant this.
"A substantial chunk of artistry is knowing what isn't worth publishing. Now, we've got Douglas Adams and Heinlein releasing stuff from beyond the grave that they might not deem publishable, given the option."
Pizza is a wonderful food because even bad pizza is still pretty good food. (Same for cheesecake.) I'd bet bad Douglas Adams is a lot better then other authors' good stuff.
This also reminds me that RH had a long list of pseudonyms that he used, based on how good he thought a story was. He had some names that had some respect he wanted preserved and some that he would slap on any piece of crap that someone would publish. (Another reason for this was so to disguise that a few authors were writing almost all the stories in the magazine.)
And the man getting shot at the table was in The Cat who walks through Walls. RH put such visible, painful effort into creating a male lead that was not a clone of Lazarus Long.
"What editor? If you read the top, it was published with only minor spelling corrections. This is similar to the tack that was taken with 2 other works after his death. They were re-published the way HE wrote them, not the way they were first published."
This could make it very interesting reading. It is great to read the two versions of Stranger in a Strange Land together. (For those who missed it, there is a version of SiaSL published that is much closer RH's version before some serious editing.) A lot of the writing was tightened up signficantly by the editor. This made it more even, but got rid of some descriptive language that was almost poetic, but completely tangential and distracting. OTOH, the editor's squeamishness about sex eviscerated at least one major scene and ruined the impact of it.
"the group marriage thing was in that book, but somehow it seemed to flow with the plot a lot better. I accepted it and liked how it shaped the characters... it had *meaning*, but did not define or control the main plot line. It added to, rather than distracting from, the overall story."
Good way of putting it. RH has these minor themes like group marriage, the gold standard, some mathematically-based predictional group psychology crap, soldiers are better people than everyone else, religion as opiate-to-the-masses, and so on that are OK once or twice. But they are as you said shoehorned into almost every book, with nothing new added, and all nuance eschewed.
It makes me think of a series of books by a modern SF author that spends a lot of time exploring identity, how you define yourself with respect to your history, family, spouse, responsibilities, flaws, mistakes, achievements, and so on. But each book adds a new twist or new perspective. (Any guesses as to who I am talking about?) There is depth and empathy here that RH never approached. He had more of an engineering approach; a smart person can find the right solution to any problem, and then there's nothing more to be said about it.
"I don't know how to use TeX, it isn't installed on my machine or anyone else's I know, and I don't make the purchasing or IT decisions for my company.
This blew my mind. Someone can't use TeX because they don't make the purchasing decisions at their company. An innocent trapped in a paradigm who doesn't even know it.
"Phew, with all those emotive phrases, I'm finding that I have strong opinions on this, despite not having access to the data behind the complaint...
When I read Herbert's column today, I immediately thought of onion-esque headlines like "New Study Shows Half of Student Test Scores Below Mean". Or claims like "alarming and puzzling new studies show that some cancer rates are above average at IBM labs and an equal number are below average, while most are pretty close to average."
People also don't understand that extremely unlikely things will happen if you have a large enough sample size. There are probably several companies in the US where the cancer rate is three standard deviations above average just due to chance. This is why elevated rates by themselves prove nothing. If you select some company to study because they have been seriously violating OSHA standards and then find that illness rates are high, you have good reason to suspect causality. But if you select a company to study because illness rates are higher than average, you then need to find some other evidence that they were doing something unusually dangerous to show it is not random chance.
"That link went to an Op-Ed, which is, er, an EDITORIAL."
No, it is an opinion piece. Editorials are written by, now stay with me here, the editors!
However, it is an opinion piece written by a NYTimes columnist, not an unsolicited submission. This association increases his credibility and therefore gives the NYTimes a greater responsibility to see that he does not fabricate or mislead.
"At the very least sending fraudulent invoices..."
Emphasis mine. Because if/when they drop those invoices in a box owned by the United States Postal Service they have possibly committed mail fraud. I predict that SCO will never mail one single invoice. They might FedEx them, they might send McBride knocking on your front door to hand it to you, but you will never receive one through the USPS.
"SCO is awesome because they guys are just so UTTERLY ridiculous. as idiotic as their thinking is, you have to respect the confidence they have to send out invoices."
I might agree if SCO was a person, but it is not. If SCO were a person, you could admire (in some twisted way) the brazen disregard for sane self interest. But it is execs like McBride who are committing these bizarre acts, and they risk nothing (as long as they can keep the SEC off their backs). SCO will die, and they'll be rich. People dumb enough to have bought SCO stock at $16.3 per share (10AM this morning) will be the ones losing out. And the taxpayers who have to fund the legal system. There is nothing worthy of admiration in SCO's actions.
In practice, I make sure everything is stored in plain ascii and run grep alot. What file did I put that phone number in? grep -i smith * */* | less
Keeping email organized is a lot harder than it should be. There is no good way to deal with things like a seminar announcement that I need to keep for two weeks but is junk after that, or stuff that I need to remember to read or reply to but don't want to read right now (or stuff I keep because I should read it but don't want to actually read ever).
It is also hard to remember that, when someone emails me some document, the place to store it is not in an email folder, but a directory dedicated to that project or subject. Like if someone sends a reference for a paper I am writing, it should go in ~/papers/journalname/papername/references or something, not just stay as an attachment in my inbox.
And once in a while, you have to waste a day or two reorganizing your crap and deleting old email. This is especially hard when I have copies of documents or programs on different computers, because I have to figure out which ones are the most recent and are the authoritative copy. CVS and rsync help here; CVS makes it obvious which copy is the best one (the one in CVS), and rsync makes it easy to keep things identical on different machines so you don't have the problem to begin with.
What was the question? Oh yeah. Let google index your entire file tree and use it to find stuff.
"IIRC, the DMCA specifically permits circumvention of copy protection/DRM/anything else if it is done specifically for purposes of interoperability (not just to allow unauthorized access to information). That means that OpenOffice or any other competitor would be allowed to crack their encryption in order to allow their users to read.doc files. Right?"
This may be true in theory. But how does someone who writes the code prove that it was for interoperability and not to steal data? We have seen from the DeCSS cases that the law will sometimes assume that piracy and theft are the intentions even in the face of contrary evidence. A private hacker who has no financial incentive in producing code so OO and MSWord or whatever can share files may have difficulty in arguing that the code is for OO interoperability when he is not employed by OO.
"The article points out, and I agree, that it's unlikely DRM will be applied to documents by default, since implementing it requires configuring Windows Server 2003 and ensuring both the creator and reader of the document have access/accounts on the Rights server."
Your claim is that setting up the system is too much work now, so for the forseeable future it will not be used. One might think that MS will expend some effort to make it very easy to configure the server and clients, and thereby reduce resistance to using it. They may also tie it in to some other system that will be necessary, perhaps something to do with software site licenses or email or anything.
'A private company introduces software that basically introduces built-in encryption for word documents, spreadsheets, and email...And the availability of this product is somehow an example of "blatant abuse of the law"?'
You miss something. These features are awarded special protections under the law. The concern is that MS is trying to use a loophole to extend these special protections beyond what they were meant to be. This is an abuse of the law. The features themselves are not at issue, but the effect of the features in the context of the DMCA.
"Or you just get out your trusty camera and take a picture of it. If you want to get higher tech, capture the EM signal generated by the monitor. It's just like bypassing music DRM by recording from a line out. This sort of security will stop casual snoops, but somebody who wants the information will get it."
All this only means something if the point is to protect information. Office workers are not going to set up van whatever-it-is phreaking rigs so they can copy stuff onto their old computers to use in applications without DRM. The purpose is to get J. Random User to follow the upgrade path on orders from One Microsoft Way. Probably another purpose is the appearance of security.
It is like checking photo IDs in airports. It doesn't stop terrorists, who have photo IDs. It stops people selling airline tickets they can't use to people who can, and gives an appearance of serious security procedures.
"Some of the evidence Sontag showed us is straightforward:...He said this code was not added to Linux by IBM but by someone else, and that it's a violation of SCO's copyright."
They admitted this after ESR, Bruce Perens, and others showed that SCO's example of SMP code illegally put in the kernel by IBM was in fact code for Itanium support put in by SGI. It was only there for a few releases, and is nearly identical to public domain Unix code. In other words, they are admitting that they lied.
I have ceased to be shocked by journalists who take press releases and treat them as unquestionable truth, even when contrary evidence is easily available. I have not ceased to be appalled by it.
I am regularly asked what the average Internet user can do to ensure his security. My first answer is usually "Nothing; you're screwed." -- Bruce Schneier
The sentiment behind this pithy observation is that computers are designed to be versatile and powerful, and this concept is fundamentally antithetical to security.
Unless you are going to make everyone run identical computers and software, you can't lock down the computers. The only choice is to lock down the network. Improve the ability to detect, diagnose, and isolate problem machines. Assume that the worms will come and they will infect machines, and plan on how to catch them, disable them, and reduce their impact as fast as possible. This requires investment ahead of time in software, hardware, and training. Make this a primary consideration in how you design your networks.
Additionally, make it as easy as possible for students to get security updates. Mirror them locally and make lots of cd's, if is legal. Produce your own clearly worded intructions/FAQ on where to get the software and information. People are lazy; you can make it hard for them to find stuff and then feel superior when they don't upgrade, or you can make it as easy as possible to do the upgrade and save yourself pain.
Can probably find people in the dorms with sufficient knowledge to help J. Random User with updates who will work for pizza.
"Case in point -- back in 2000, even though I had about four years Linux experience by then, I managed to bring down Internet access for an entire dorm (about 900 students) for a week."
So, for an entire week no one was able to take PC with the equivalent of tcpdump on it, hook it up to the network, and look at the packets. Or look at what was making it out of the local network and see that only one machine was getting through, and suppose that machine might be part of the problem. This doesn't sound like competent system administration.
This sounds like people whose idea of testing the network connection is starting IE and going to www.espn.com.
Re:McBride and capitalism
on
SCO Roundup
·
· Score: 1
"There is nothing "communist" about Linux."
I think many people fail to understand the context of the GPL and RMS's ideas. They are not against capitalism at all. But the key is to realize the difference in point of view. The GPL rejects the idea of software as a piece of property that can be owned (more or less). Instead, the market should be oriented toward payment for the service of writing software. If IBM wants Linux to run on its hardware, it should be ready to pay for the services of people to write the code to do this. The philosophy is that people should be paid for writing code, not for "owning" code.
This seems strange in today's world partly because of the shift to shrink-wrapped software. It was assumed that people would pay to get software to do what they wanted. People now mostly accept what is offered to them, and adjust their expectations instead. This makes it difficult to pay someone for writing code as separate from paying someone for "owning" code.
The GPL is not anti-capitalist if you refuse to let companies like MS define your terms for you.
Re:Dear Comrade McBride....
on
SCO Roundup
·
· Score: 1
"What I want to know is, what is so wrong with communism. Can you give me a clear, logical answer?"
It just doesn't work. At a simplistic level, communism is an extension of how families often work. You will feed and house a four year old or a sick parent, even though they do not materially contribute. But people don't feel about everyone in the world they way they feel about their close family, so extending the economy of a family to a whole country or planet doesn't work. This may be related to family members building up trust over time, the "economy" being local so each member can observe it, and genetic hard-wiring.
It is easy to find examples of human actions that prove that communism is unstable, because there are things that we effectively own in common. Think about the people who weave through tight traffic and make everyone else's trip longer. Look at how many people litter. Spam! There are enough people who abuse any common resource to make communism impossible.
"Mark Heise...believes SCO is entitled to pursue users based on its claims. 'End users are improperly using this copyrighted material, and under copyright law SCO is entitled to damages and injunctive relief,' he said."
It is beautiful to see how they shoot themselves in the foot (head?) whenever they say something the least bit concrete.
Heise says it is copyright that is at issue. There is no reason they can't detail what code is in Linux that they own the copyright to. They only have that excuse for trade secrets. Oops. And copyright law was violated by whoever did the copying, i.e. the distributor of the code, not the user. Oops. And it is one thing for SCO to make outrageous and unfounded legal claims, but a lawyer who publicly makes outrageous and unfounded legal claims could be (should be) sanctioned or disbarred. Oops.
"I don't want to spit in the soup, I think Moz rocks the boat, but apart from the oh-so-welcome stability issues, it's more or less functionally equivalent to Netscape Communicator 4.7 to me"
This may be true for you, but I think that you are an exception, and I use moz on a P200 with 64M of RAM. Two examples: First is rendering tables. Netscape 4 had awful problems with some large tables, such that scrolling such a page induced a slow, screen-flashing nightmare. Moz is significantly faster at rendering some things than Netscape. Another is CSS. If I want to make even trivial use of standard CSS in writing a web page, I can't do it primarily testing with Netscape 4. (And image blocking on selected sites is incredibly nice. Amazing how many sites like Slate serve their ads from a different server.) There are many real improvements over NS4, not just fluff.
Slashdot Mirror
In theory, but not in practice yet. My home machine is on dialup, and data on it is not available when I am not there. Servers at work are for work, not personal use, and anyway I wouldn't put personal stuff on a machine where some BOFH has the root password unless I absolutely had to.
And this would still only work if whatever machine I was trying to use also had a broadband connection. "Hey want to see the photos I scanned at nice high resolution? Let's connect to the ftp server and come back in three days when it's done."
People recognize this problem. It is why they carry laptops around with them everywhere. I see people all the time using a laptop while sitting next to a desktop or an X server. This is the easiest way to carry their data around.
Except they haven't, and they don't. It is one of RH's simple-minded fantasies. This is not to say that many in the armed forces don't have a great interest in the good of society, but as a whole they are not uniquely gifted in this respect. Read some history. Read the news. Meet some people. You'll figure out that RH was full of it.
I found it interesting that in Grumbles from Beyond the Grave RH backtracked on this in personal letters at the time. He claimed that really all he meant was that we out to try a different system, because he thought that government based on the US Constitution was not working too well. He also suggested we try allowing only women to vote for a while. His infatuation with soldiers was a little too obvious for me to think he meant this.
Pizza is a wonderful food because even bad pizza is still pretty good food. (Same for cheesecake.) I'd bet bad Douglas Adams is a lot better then other authors' good stuff.
This also reminds me that RH had a long list of pseudonyms that he used, based on how good he thought a story was. He had some names that had some respect he wanted preserved and some that he would slap on any piece of crap that someone would publish. (Another reason for this was so to disguise that a few authors were writing almost all the stories in the magazine.)
And the man getting shot at the table was in The Cat who walks through Walls. RH put such visible, painful effort into creating a male lead that was not a clone of Lazarus Long.
This could make it very interesting reading. It is great to read the two versions of Stranger in a Strange Land together. (For those who missed it, there is a version of SiaSL published that is much closer RH's version before some serious editing.) A lot of the writing was tightened up signficantly by the editor. This made it more even, but got rid of some descriptive language that was almost poetic, but completely tangential and distracting. OTOH, the editor's squeamishness about sex eviscerated at least one major scene and ruined the impact of it.
I hope you mean he made reading about sex boring to you.
Good way of putting it. RH has these minor themes like group marriage, the gold standard, some mathematically-based predictional group psychology crap, soldiers are better people than everyone else, religion as opiate-to-the-masses, and so on that are OK once or twice. But they are as you said shoehorned into almost every book, with nothing new added, and all nuance eschewed.
It makes me think of a series of books by a modern SF author that spends a lot of time exploring identity, how you define yourself with respect to your history, family, spouse, responsibilities, flaws, mistakes, achievements, and so on. But each book adds a new twist or new perspective. (Any guesses as to who I am talking about?) There is depth and empathy here that RH never approached. He had more of an engineering approach; a smart person can find the right solution to any problem, and then there's nothing more to be said about it.
This blew my mind. Someone can't use TeX because they don't make the purchasing decisions at their company. An innocent trapped in a paradigm who doesn't even know it.
When I read Herbert's column today, I immediately thought of onion-esque headlines like "New Study Shows Half of Student Test Scores Below Mean". Or claims like "alarming and puzzling new studies show that some cancer rates are above average at IBM labs and an equal number are below average, while most are pretty close to average."
People also don't understand that extremely unlikely things will happen if you have a large enough sample size. There are probably several companies in the US where the cancer rate is three standard deviations above average just due to chance. This is why elevated rates by themselves prove nothing. If you select some company to study because they have been seriously violating OSHA standards and then find that illness rates are high, you have good reason to suspect causality. But if you select a company to study because illness rates are higher than average, you then need to find some other evidence that they were doing something unusually dangerous to show it is not random chance.
No, it is an opinion piece. Editorials are written by, now stay with me here, the editors!
However, it is an opinion piece written by a NYTimes columnist, not an unsolicited submission. This association increases his credibility and therefore gives the NYTimes a greater responsibility to see that he does not fabricate or mislead.
Emphasis mine. Because if/when they drop those invoices in a box owned by the United States Postal Service they have possibly committed mail fraud. I predict that SCO will never mail one single invoice. They might FedEx them, they might send McBride knocking on your front door to hand it to you, but you will never receive one through the USPS.
I might agree if SCO was a person, but it is not. If SCO were a person, you could admire (in some twisted way) the brazen disregard for sane self interest. But it is execs like McBride who are committing these bizarre acts, and they risk nothing (as long as they can keep the SEC off their backs). SCO will die, and they'll be rich. People dumb enough to have bought SCO stock at $16.3 per share (10AM this morning) will be the ones losing out. And the taxpayers who have to fund the legal system. There is nothing worthy of admiration in SCO's actions.
Keeping email organized is a lot harder than it should be. There is no good way to deal with things like a seminar announcement that I need to keep for two weeks but is junk after that, or stuff that I need to remember to read or reply to but don't want to read right now (or stuff I keep because I should read it but don't want to actually read ever).
It is also hard to remember that, when someone emails me some document, the place to store it is not in an email folder, but a directory dedicated to that project or subject. Like if someone sends a reference for a paper I am writing, it should go in ~/papers/journalname/papername/references or something, not just stay as an attachment in my inbox.
And once in a while, you have to waste a day or two reorganizing your crap and deleting old email. This is especially hard when I have copies of documents or programs on different computers, because I have to figure out which ones are the most recent and are the authoritative copy. CVS and rsync help here; CVS makes it obvious which copy is the best one (the one in CVS), and rsync makes it easy to keep things identical on different machines so you don't have the problem to begin with.
What was the question? Oh yeah. Let google index your entire file tree and use it to find stuff.
This may be true in theory. But how does someone who writes the code prove that it was for interoperability and not to steal data? We have seen from the DeCSS cases that the law will sometimes assume that piracy and theft are the intentions even in the face of contrary evidence. A private hacker who has no financial incentive in producing code so OO and MSWord or whatever can share files may have difficulty in arguing that the code is for OO interoperability when he is not employed by OO.
Your claim is that setting up the system is too much work now, so for the forseeable future it will not be used. One might think that MS will expend some effort to make it very easy to configure the server and clients, and thereby reduce resistance to using it. They may also tie it in to some other system that will be necessary, perhaps something to do with software site licenses or email or anything.
You miss something. These features are awarded special protections under the law. The concern is that MS is trying to use a loophole to extend these special protections beyond what they were meant to be. This is an abuse of the law. The features themselves are not at issue, but the effect of the features in the context of the DMCA.
And stop calling me Shirley.
If you're drinking all that Jolt, you might want to make that two three-inch holes.
All this only means something if the point is to protect information. Office workers are not going to set up van whatever-it-is phreaking rigs so they can copy stuff onto their old computers to use in applications without DRM. The purpose is to get J. Random User to follow the upgrade path on orders from One Microsoft Way. Probably another purpose is the appearance of security.
It is like checking photo IDs in airports. It doesn't stop terrorists, who have photo IDs. It stops people selling airline tickets they can't use to people who can, and gives an appearance of serious security procedures.
They admitted this after ESR, Bruce Perens, and others showed that SCO's example of SMP code illegally put in the kernel by IBM was in fact code for Itanium support put in by SGI. It was only there for a few releases, and is nearly identical to public domain Unix code. In other words, they are admitting that they lied.
I have ceased to be shocked by journalists who take press releases and treat them as unquestionable truth, even when contrary evidence is easily available. I have not ceased to be appalled by it.
The sentiment behind this pithy observation is that computers are designed to be versatile and powerful, and this concept is fundamentally antithetical to security.
Unless you are going to make everyone run identical computers and software, you can't lock down the computers. The only choice is to lock down the network. Improve the ability to detect, diagnose, and isolate problem machines. Assume that the worms will come and they will infect machines, and plan on how to catch them, disable them, and reduce their impact as fast as possible. This requires investment ahead of time in software, hardware, and training. Make this a primary consideration in how you design your networks.
Additionally, make it as easy as possible for students to get security updates. Mirror them locally and make lots of cd's, if is legal. Produce your own clearly worded intructions/FAQ on where to get the software and information. People are lazy; you can make it hard for them to find stuff and then feel superior when they don't upgrade, or you can make it as easy as possible to do the upgrade and save yourself pain.
Can probably find people in the dorms with sufficient knowledge to help J. Random User with updates who will work for pizza.
So, for an entire week no one was able to take PC with the equivalent of tcpdump on it, hook it up to the network, and look at the packets. Or look at what was making it out of the local network and see that only one machine was getting through, and suppose that machine might be part of the problem. This doesn't sound like competent system administration.
This sounds like people whose idea of testing the network connection is starting IE and going to www.espn.com.
I think many people fail to understand the context of the GPL and RMS's ideas. They are not against capitalism at all. But the key is to realize the difference in point of view. The GPL rejects the idea of software as a piece of property that can be owned (more or less). Instead, the market should be oriented toward payment for the service of writing software. If IBM wants Linux to run on its hardware, it should be ready to pay for the services of people to write the code to do this. The philosophy is that people should be paid for writing code, not for "owning" code.
This seems strange in today's world partly because of the shift to shrink-wrapped software. It was assumed that people would pay to get software to do what they wanted. People now mostly accept what is offered to them, and adjust their expectations instead. This makes it difficult to pay someone for writing code as separate from paying someone for "owning" code.
The GPL is not anti-capitalist if you refuse to let companies like MS define your terms for you.
It just doesn't work. At a simplistic level, communism is an extension of how families often work. You will feed and house a four year old or a sick parent, even though they do not materially contribute. But people don't feel about everyone in the world they way they feel about their close family, so extending the economy of a family to a whole country or planet doesn't work. This may be related to family members building up trust over time, the "economy" being local so each member can observe it, and genetic hard-wiring.
It is easy to find examples of human actions that prove that communism is unstable, because there are things that we effectively own in common. Think about the people who weave through tight traffic and make everyone else's trip longer. Look at how many people litter. Spam! There are enough people who abuse any common resource to make communism impossible.
It is beautiful to see how they shoot themselves in the foot (head?) whenever they say something the least bit concrete.
Heise says it is copyright that is at issue. There is no reason they can't detail what code is in Linux that they own the copyright to. They only have that excuse for trade secrets. Oops. And copyright law was violated by whoever did the copying, i.e. the distributor of the code, not the user. Oops. And it is one thing for SCO to make outrageous and unfounded legal claims, but a lawyer who publicly makes outrageous and unfounded legal claims could be (should be) sanctioned or disbarred. Oops.
This may be true for you, but I think that you are an exception, and I use moz on a P200 with 64M of RAM. Two examples: First is rendering tables. Netscape 4 had awful problems with some large tables, such that scrolling such a page induced a slow, screen-flashing nightmare. Moz is significantly faster at rendering some things than Netscape. Another is CSS. If I want to make even trivial use of standard CSS in writing a web page, I can't do it primarily testing with Netscape 4. (And image blocking on selected sites is incredibly nice. Amazing how many sites like Slate serve their ads from a different server.) There are many real improvements over NS4, not just fluff.