Recently, I served on a committee for our college that did some strategic planning. You know... the whole "strengths, weaknesses, opportunities, threats" deal.
Anyway, one of the ideas that someone brought up was the notion of trying to influence our ranking in the U.S. News annual report. So we looked into how the rankings are done.
As I recall, it turned out that the main factor in the rankings of universities as a whole was the peer assessment (other deans of universities and colleges). To this end, all of the institutions who put a priority on being near the top of the list make sure to send out promo material to everyone that U.S. News queries... ideally a few weeks before U.S. News sends out the queries, so that the promo material is still fresh in the mind of the voters.
For either the overall rankings or the rankings of the individual programs (like engineering, business, etc), there were some other very interesting quantitative measures that came into play. One of them was something like the percentage of classes with fewer than, say, 21 students (which increase a school's score) and another was the percentage of classes with more than about 35 students (which lower a school's score).
One insteresting suggestion someone on the committee made was, if we had any classes with a maximum class size of 21 or 22, lower it to 20. Only one or two students have to wait until next quarter for the class, and the college gets a discreet jump in its score. Same goes for lowering classes with a max of 35 or 36 to 34. Every little bit helps.
Anyway, the long and short of the story is that... there are a lot of clever people who make it their business to juice the scores that their school gets. If a school isn't very high on the list, it doesn't necessarily mean that it's a bad school. It might just mean that they haven't found out how the ranking game is really played. (Kinda like an athlete who doesn't realize that everyone else is using steroids yet).
I can't believe this made Slashdot! I coulda been famous.... I coulda BEEN somebody!:)
Actually, I've done this several times. The toughest thing about the whole process is actually finding the identical "donor" drive. The first time was probably 3 or 4 years ago when a faculty member fried her hard drive. You could actually *see* which chip blew out. There was a little scorched hole and soot around it and everything.
So, I managed to find another drive on eBay and ordered it. I got it in the mail, swapped the cards, and boom, it worked.
A year and a half later, she shows up at my door again... same hard drive... SAME blown chip. The only problem is that, these days, it's not easy to find this particular Quantum Fireball 10GB. (Actually, it has just struck me how apropos the "Fireball" moniker is for this drive).
Anyway, I eventually found a dude selling the same drive, same model number and even the same controller board revision markings. Bid... won... swapped and then immediately copied everything onto one of our servers, burned her personal files onto CD, and told her to copy her hd to some different model of drive and to toss the Fireball.
Seriously though... this kinda thing gets you featured on Slashdot? Sheesh. I supposed next you're going to feature some "printer hacker" who refills his own ink cartridges.
How is *that* not news in itself? Either they've got a robot that can ride a bike, or they've got a pilotless Segway. Either way, that's pretty impressive.:)
Of course, there's always the other possibility that (casters|outriggers|nylon sliding feet|articulated legs) don't count as "wheels" and shouldn't be mentioned.
CNN has a story on 'exclusive' Peer to Peer networks, that require 'knowing the right people and having a wealth of content on your hard disk to get into the clique
Over the last several months, I've begun to conclude that something like this is the only way that file swapping can really endure. Basically, my idea was that each person's file swapping client would only make/accept connections to/from people that you trust: friends, family, etc.
The twist would be that the system would allow relaying of searches and of actual files. In other words, if I request a file that is on my friend's friend's computer, then the file has to come through the computer of our mutual friend. The whole idea is to keep things as encapsulated as possible... kinda like how terrorist cells work.
Now, I know that this increases network traffic... adds a lot of opportunities for a "weaker link" in the chain (imagine if one of the people in the relay chain is using a 56k modem)... decreases the "connectedness" of the whole sharing network, etc. However, I think this is the only real way to keep the RIAA from just being able to download a song and, *pow*, have the IP of someone to sue.
Also, some of these problems mentioned might be assuaged by the fact that people might feel more comfortable leaving their stuff shared. I, for one, have gobs and gobs of stuff that I could share, but I don't... because I have way too much to lose. However, if I knew that the only people who could connect to me would be people that I know... I'd have tons of stuff up and shared... 24/7.
The strange thing is that it seems to me that this was Aimster's plan, but they got shut down for some reason. But I don't know why.
Sure, you let a friend use your handgun for target practice and he shoots the cashier. Why wouldn't you be guilty.
In some situations, the law sort of allows for a "we're going to have SOMEBODY'S head in the basket when this is all over..." approach.
As an example, take cars. Remember when those auotmatic speeder-detector-license-plate-photographing trailers were introduced? They'd snap a picture of your car going 55 in a 25mph zone and the cops would mail you the photo along with instructions on where to send the money? Remember when people initially said "Hey... you can see who'd driving, so you can't prove it was me.". As far as I know, that ended up not mattering much.
I guess the courts figured that, if it wasn't you who was speeding, then you lent your car to someone that you should have suspected would speed.... which is as bad as speeding yourself. However, this is just bullsh*t legal justification for the real underlying position of "Somebody was speeding and we're going to get our freakin' money... and you're the dude we have the most justification for.". But it works... that's the important part.
Same goes for having alcohol and minors at a party. If the cops bust in and find 16 year-olds drinking from the keg, and you own the house, then, unless someone else comes forward and confesses, your ass is in trouble. You might not have bought the beer... you might not have invited the kids... you might not have even let them in, but that doesn't matter. You're as far as they can follow the clues and you're the one they're going to roast. They must figure that you know who's really responsible for it (and aren't talkin') and you'll probably punish the real guilty party in your own way later. So... it's justice by proxy, in a way.
Now, in the case of downloading stuff on someone else's computer, I can see them making the same argument. There's really no accurate way for them to find out who was using your computer at any given time. They can't exactly dust the keyboard for fingerprints. So, what they do is follow the trail of clues to you and figure that you know who was using your computer. So, they sue you for $XXX,000 and leave it to you to squeeze that out of the person who's really guilty.
Holly smokes! Am I the only one who thinks that the dude on the bike on the rocket pack page looks like a teen-aged Bob Lazar, the guy who claims to have worked at Area 51?!?!? Propane my eye... I'll bet that thing runs on antimatter for sure!
On a side note, check out the lame-ass shifter and the headlight mounted on the bike. All I can say is that I sure hope the rocket worked, or that dude must have gotten beat up a lot.
From the early 1950's until 1991, copyright registrations rise exponentially. In fact, a simple quadratic fit shows an Rsquare of over.99
Ummm... a quadratic (ie, 2nd-order polynomial... usually thought of by the masses as y=x^2) is not exponential (ie, e^x). Exponential growth is so much greater that there is no polynomial (not 2nd-order... not 100th-order...) that can keep pace with it.
Looks like you need to spend a little more time in the math section of the library.
Actually, the book contained a very interesting point that relates directly to this.
The issue was that the German hacker was able to take a bunch of individual bits of information (which, by themselves, weren't classified) and bring them together to deduce some information which is classified.
Cliff Stoll mentioned that this type of research prompted some of the intelligence officials to propose a new classification ("sensitive") to represent information that wasn't classified but which could be used to deduce classified info.
As far as I can tell, in the U.S., if you throw something away, the police don't need a search warrant to obtain it. They can go through your trash all they want.
It goes likewise for eavesdropping on conversations. There's something called "reasonable expectation of privacy". If I'm talking to someone on a crowded street, then the cops can record my conversation and use it as evidence because I wasn't taking steps that someone would reasonably take if they were trying to keep something secret (as in, secret from everybody and not just from the police).
This DNA thing strikes me as something similar. Just like if you throw a murder weapon with your fingerprints on it into your trash, I think you're "discarding" your DNA if you were to, say, have children. The parallel to "reasonable expectation of privacy" is that, if you were someone who really didn't want their DNA (or half of it) out there running around loose, then you would elect to not have kids.
So, if you have kids, then I think that the cops should, clearly, have a right to use the DNA of those kids (provided they're over the age of consent and provided that they volunteer their DNA) to catch you. What's a little more murky is whether the cops should be allowed to catch you based on DNA from any familial relative... including ones you have no control over the production of (like, your parents, cousins, etc.). Again, I think it would hinge upon whether you had kids. Going back to "reasonable expectation of privacy". If you have kids, then you aren't exercising what little control you have over the dissemination of your DNA... so it must not matter to you that much... so everything's fair game at that point.
Legal issues aside, I must say that this is one really cool thing about DNA. Everybody knows the obvious advantages that DNA has over fingerprints: you get usable evidence from smaller samples, from a wider variety of fluids, smudges, etc. What's less obvious is the idea of "proximity". Two people with similar fingerprints are not necessarily closely related, and two people who are closely related don't necessarily have similar prints. With DNA, that's not the case. With DNA, you're able to tell when you've got some DNA from someone closely related to some "target DNA". Even cooler, you can probably tell how many generations away they are.
...ink for home printers is now seven times more expensive than vintage champagne
This makes sense for our current economic climate. Consider that champagne is needed for when your company goes public and ink is what's needed when it goes bust (ie. resume').
"...we had about eighteen different things we had in mind for the browser"
Oooh, yeah... like VRML?:)
I remember seeing an interview with Mark Pesche, the dude who was regarded as the author of the VRML spec, and he was going on and on about how cumbersome it is to keep track of URL's when we could be navigating in a 3D space for our documents....
Could you just see that? "Come visit Jiffy Lube on the web! Start at the Origin, go down Street 1 until you come to the big purplish billboard, bear left and continue through the pasture... go under the spaceship and then head 4 spaces east and you can't miss us!". And this is more intuitive than "www.jiffylube.com" because... why?
I'm sure that, of those 18 improvements to the browser, many or all of them promised to "completely change the way we think about browsing". However, like VRML, it's not necessarily a change for the better.
and if everyone comes to that conclusion then it's bye-bye P2P network
I haven't been following the P2P judgements closely, so I'm a little curious as to how they calculate the monetary "damages" figure.
First off, wouldn't they have to demonstrate that the file I'm sharing titled "Doobie Brothers - China Grove.mp3" is really what it says it is? And if they have to verify each song to be able to sue me for it, then wouldn't they have to download a whole lot of my songs? If that's the case, you'd think it'd be possible to make a P2P client that was resistant to repeated download requests from single (or small ranges of) IP's... in the same way that some network firewalls block port scanning.
Next, don't they have to show that I actually supplied mp3's to people and cost the RIAA revenue? How many downloads would they actually be able to prove? And how much lost revenue could there really be? I mean, when iTunes is selling songs for $1, not all of that is going to the record companies, right? So you figure that the lost revenue will be something less than $1 per song. At that rate, it would take quite a while to rack up even a measley $100 in potential damage to the record companies using a household broadband connection.
It makes me wonder if there'd be a need for a P2P client that was RIAA-judgement-aware... that would let you limit your "reparations" exposure by limiting the number of songs that could be downloaded per time period. In other words, if you used to spend, say, $50 per month on CD's, then you could tell the P2P client to, each month, only allow downloads until the amount the RIAA could sue you for went up by $50. This way, you could get access to a bunch of songs, and you're only maybe out that $50 per month.
Another thing I'm wondering about is the partial-download problem. With most P2P networks, you can download a song by getting a portion of it from many different sharers. If the RIAA downloads a song and it comes from 5 different people, are they expecting to (I know they're hoping to, but are they expecting to) be able to sue each of those five for the full "lost revenue" amount?
As time goes on, I keep discovering that more and more commonly-used cliche's trace back to famous pieces of literature. (ie, "Neither a borrower, nor a lender be", "Good fences make good neighbors", "Out, out.. damned spot!", etc.). It surprises me how ignorant most people (including me) are about where these came from.
Now, looking back on my English experiences, I think it would have been pretty cool if each student were given a phrase and they had to use the net to find out what literature it originally came from and have to read enough of the surrounding text to be able to describe the context of the scene where the phrase occured (like Lady MacBeth trying to wash the blood off, etc).
Back when the web first started, there were a lot of web-page creators scrambling for ways to make their page viewable, but not able to be saved, printed... whatever. The end conclusion was always the same: "If it can be viewed, it can be printed".
The same goes for "rip-proof" CD's. At some point, it has to be listenable to a human. When that happens, the song is vulnerable to being copied.
The obvious way to do this is just to route your "Line Out" into your "Line In" on your PC and then just have a sound recorder going while your CD plays. Of course, this carries the problem of converting from digital, to analog, and then to digital again.
What's only a little less obvious and a little less difficult (so much so that I can hardly believe I haven't seen it available yet) would be to have a pseudo sound output device. Assuming that the CD would be playable (but not rippable) on a normal PC CD-ROM drive, you could tell your CD player app to use this pseudo sound device as the output. To the app, it would look like a regular sound card (kinda like how Adobe Acrobat appears to be a printer), but it would actually just write the digital data to a file (again, like Acrobat does).
The nice thing here is that, the CD could even be restricted to only being played on a DRM-enabled player. At some point, that player has to send the audio off to what it thinks are speakers. If you have a pseudo device that intercepts the audio, then there you go.
Of course, I wouldn't be surprised if Paladium had components to prevent this... but that's a different story. The point here is that, if you had a pseudo sound card, you could still rip AND keep it all digital. Granted, the rip would happen at 1x... but that's why I have a second PC in my office with lots of games on it.:)
...a keen teenager who code programs computers quickly than an older person...
First off, let me confess that, at 34, I don't program nearly as fast as I did back when I was, say, 20.
However, let me also state that I think that this is because I only bother to actually enter the code that doesn't have bugs.
I'm quite serious here. Although I would be cranking out more code when I was younger, I'd end up debugging, deleting, and rewriting it for just as long as it took me to write the first draft. Nowadays, I'm armed with so much "been down that road before" experience that I can see where the coding perils are before I even start.
Look at it this way... if you wanted to make an expedition to the south pole, would you want the strapping teenager who goes "Cool, I'll grab my backpack and then we can go!", or the veteran who says "Okay... give me a few weeks to get our equipment and supplies and to plan our route..."?
Coding is kinda the same way. There are, at first glance, a million ways of getting to the goal, but many of them aren't going to work.... and many of those won't reveal themselves as such until it's too late. Just like an expedition, much of the work in coding is in the planning before you even start.
But officials in other parks, faced with an onslaught of geocachers, are scrambling to develop restrictions
I agree.... the last thing we want is more people in public parks and natural spaces.:)
Now, on a more serious note, when I explain the geocaching thing to friends or co-workers, if they still look at me funny after I explain the whole GPS and prize-exchange thing, then I just use my standard: "Okay... just think of it as an excuse to go hiking.". Ultimately, I think that that's the end result: marginally more people getting out of their homes and walking around in nice, big public areas and getting some fresh air.
Now, although the rangers and park officials claim that they'd like more people to appreciate the outdoors, I think they'd prefer that they appreciate them from afar. Like just about every other line of work, I'm sure they often mumble amongst themselves: "This job would be great if it weren't for the customers".
Bottom line: If you want more people at the park, then you've got more people at the park. Deal with the increased foot traffic... that's what you're paid to do... to manage the park. However, if you *don't* want more people, then either close the damn park entirely or require permits or something. But don't go preaching about how people should get back to nature if you're not prepared for them to do it during your shift.
...a lot of the ideas are modeled off of major car free cities in Europe (like Venice)
Venice shouldn't be used as a model for anything except what not to do.
Besides, making a car-free city by making cars impossible ("Hey, let's flood the streets!") is not the goal. The goal is to make it possible for humans to retain their current level of productivity without needing cars. Also, within that goal, I think there's the implicit constraint that the cars don't get replaced with something just as bad (like boats in Venice, for example).
I once asked a (more "hip" and "now") co-worker why they call them "blogs" and he responded that the probably call them that so as to not confuse them with the logfiles from your webserver.
Personally, I think it's so that the little morons don't forget how to spell it, but it got me thinking about all of the possible contractions of "weblog" and why they weren't used...
"weblog" - could be confused with actual web logfiles... which, by contrast, contain useful information.
"eblog" - could be confused with driftwood washed up by the tide
"log" - could be confused with something that "rolls down stairs, alone or in pairs..."
"og" - is reserved for a blogger's contraction of "ogg".
"g" - it probably already some stupid IM acronym for something (like "gee" or something).
So... you see... "blog" was the only choice the morons had.
I wish SCO would just go ahead and announce what parts of the Linux kernel have their code so that we can get started on replacing it with something that works.
What are they waiting for.... a bug-finder's fee?:P
1) Don't put DSL to each unit - pull CAT-5 and run Ethernet. Your residents will have a much easier time getting hardware than with DSL, and your costs will be less.
2) Pull the wire to a common router closet.
3) One port per unit
4) Lock down the ports that aren't being used.
5) Use DHCP to assign addresses.
6) Set up your own caching server. I would recommend using Squid.
7) Force all outbound port 25 (SMTP) through your mail server.
8) Run a virus scanner on your mail server. Scan all incoming AND outgoing mail.
9) Don't route the Microsoft file sharing ports or Apple Rendezvous ports between units.
10) Insist customers keep their machines virus free. Disconnect any who don't IMMEDIATELY.
11) Write into your rental contracts that you ARE NOT RESPONSIBLE for maintaining your customer's machines or security - if they are scared let them run their own firewall.
12) Offer your own space, accessible to your users, with virus scanner updates, MS patches, and so on. Encourage them to use that to save bandwidth.
13) Routinely sniff around for WAPs. Handle them as you see fit - disconnect, or verify they are set up sanely. Don't ignore them.
I agree with juat about all of this. Ethernet hardware is going to be much cheaper (for the end-users and probably for you, too). Compare the price of an ethernet card with a DSL modem. Now, multiply that by 160....
I'm not sure I'd use Squid for caching. Although it's a great cache, it requires that users take action on their end in order to use it... which means that it won't get used all that much and you're likely to get support calls from those people who *do* try to use it. Instead, if it's worth it to you, you might want to look at some of the routers that have transparent caching. Although I've never used one, I'm told that Cisco (and, if so, then the others like Ascend, etc. have them) has routers that magically detect port-80 traffic and do caching.... in a way that is: A) transparent and, B) non-optional.
I'm not sure I'd block outbound SMTP... but it depends upon which set of hassles you like the most. There seems to be a slow (or, rather, glacial...) drift on the net toward authenticated SMTP for laptop users who move around a lot and who don't want to have to keep changing their outbound SMTP settings. If you block outbound SMTP, then those mobile users who *are* lucky enough to deal with a single, authenticated SMTP server out there... they'll be out of luck. With only 160 stations... and with them being stations with relatively low turnover (and high degree of being able to track down the "owner" of each station), you might get a lot of mileage out of a strict spam policy.
Of course, the other alternative (to authenticated SMTP) that I've seen for laptop users is to support assignment of the SMTP server through DHCP. I'm not sure how many DHCP servers (or client OS's, DHCP clients, and mail clients) support this... but you may want to add something like this to your wishlist if you do decide to block outbound SMTP.
Also, more than considering blocking outbound SMTP, consider blocking *inbound* port-80 web traffic.... and write a prohibition against web hosting into the agreement.
A markup language for real-time monitoring of remote events... I can see it now...
<ex-wife id=1>
<screwing>
milkman
</screwing>
</ex-wife>
<ex-wife id=2>
<screwing>
that dickhead with the Trans-Am
</screwing>
<spending>
most of my money
</spending>
</ex-wife>
Slashdot Mirror
Recently, I served on a committee for our college that did some strategic planning. You know... the whole "strengths, weaknesses, opportunities, threats" deal.
Anyway, one of the ideas that someone brought up was the notion of trying to influence our ranking in the U.S. News annual report. So we looked into how the rankings are done.
As I recall, it turned out that the main factor in the rankings of universities as a whole was the peer assessment (other deans of universities and colleges). To this end, all of the institutions who put a priority on being near the top of the list make sure to send out promo material to everyone that U.S. News queries... ideally a few weeks before U.S. News sends out the queries, so that the promo material is still fresh in the mind of the voters.
For either the overall rankings or the rankings of the individual programs (like engineering, business, etc), there were some other very interesting quantitative measures that came into play. One of them was something like the percentage of classes with fewer than, say, 21 students (which increase a school's score) and another was the percentage of classes with more than about 35 students (which lower a school's score).
One insteresting suggestion someone on the committee made was, if we had any classes with a maximum class size of 21 or 22, lower it to 20. Only one or two students have to wait until next quarter for the class, and the college gets a discreet jump in its score. Same goes for lowering classes with a max of 35 or 36 to 34. Every little bit helps.
Anyway, the long and short of the story is that... there are a lot of clever people who make it their business to juice the scores that their school gets. If a school isn't very high on the list, it doesn't necessarily mean that it's a bad school. It might just mean that they haven't found out how the ranking game is really played. (Kinda like an athlete who doesn't realize that everyone else is using steroids yet).
I can't believe this made Slashdot! I coulda been famous.... I coulda BEEN somebody! :)
Actually, I've done this several times. The toughest thing about the whole process is actually finding the identical "donor" drive. The first time was probably 3 or 4 years ago when a faculty member fried her hard drive. You could actually *see* which chip blew out. There was a little scorched hole and soot around it and everything.
So, I managed to find another drive on eBay and ordered it. I got it in the mail, swapped the cards, and boom, it worked.
A year and a half later, she shows up at my door again... same hard drive... SAME blown chip. The only problem is that, these days, it's not easy to find this particular Quantum Fireball 10GB. (Actually, it has just struck me how apropos the "Fireball" moniker is for this drive).
Anyway, I eventually found a dude selling the same drive, same model number and even the same controller board revision markings. Bid... won... swapped and then immediately copied everything onto one of our servers, burned her personal files onto CD, and told her to copy her hd to some different model of drive and to toss the Fireball.
Seriously though... this kinda thing gets you featured on Slashdot? Sheesh. I supposed next you're going to feature some "printer hacker" who refills his own ink cartridges.
Of course, there's always the other possibility that (casters|outriggers|nylon sliding feet|articulated legs) don't count as "wheels" and shouldn't be mentioned.
The twist would be that the system would allow relaying of searches and of actual files. In other words, if I request a file that is on my friend's friend's computer, then the file has to come through the computer of our mutual friend. The whole idea is to keep things as encapsulated as possible... kinda like how terrorist cells work.
Now, I know that this increases network traffic... adds a lot of opportunities for a "weaker link" in the chain (imagine if one of the people in the relay chain is using a 56k modem)... decreases the "connectedness" of the whole sharing network, etc. However, I think this is the only real way to keep the RIAA from just being able to download a song and, *pow*, have the IP of someone to sue.
Also, some of these problems mentioned might be assuaged by the fact that people might feel more comfortable leaving their stuff shared. I, for one, have gobs and gobs of stuff that I could share, but I don't... because I have way too much to lose. However, if I knew that the only people who could connect to me would be people that I know... I'd have tons of stuff up and shared... 24/7.
The strange thing is that it seems to me that this was Aimster's plan, but they got shut down for some reason. But I don't know why.
And for quick trips to the shore and scouting, it comes with a dinghy made out of a Fiat convertable. :)
As an example, take cars. Remember when those auotmatic speeder-detector-license-plate-photographing trailers were introduced? They'd snap a picture of your car going 55 in a 25mph zone and the cops would mail you the photo along with instructions on where to send the money? Remember when people initially said "Hey... you can see who'd driving, so you can't prove it was me.". As far as I know, that ended up not mattering much.
I guess the courts figured that, if it wasn't you who was speeding, then you lent your car to someone that you should have suspected would speed.... which is as bad as speeding yourself. However, this is just bullsh*t legal justification for the real underlying position of "Somebody was speeding and we're going to get our freakin' money... and you're the dude we have the most justification for.". But it works... that's the important part.
Same goes for having alcohol and minors at a party. If the cops bust in and find 16 year-olds drinking from the keg, and you own the house, then, unless someone else comes forward and confesses, your ass is in trouble. You might not have bought the beer... you might not have invited the kids... you might not have even let them in, but that doesn't matter. You're as far as they can follow the clues and you're the one they're going to roast. They must figure that you know who's really responsible for it (and aren't talkin') and you'll probably punish the real guilty party in your own way later. So... it's justice by proxy, in a way.
Now, in the case of downloading stuff on someone else's computer, I can see them making the same argument. There's really no accurate way for them to find out who was using your computer at any given time. They can't exactly dust the keyboard for fingerprints. So, what they do is follow the trail of clues to you and figure that you know who was using your computer. So, they sue you for $XXX,000 and leave it to you to squeeze that out of the person who's really guilty.
Holly smokes! Am I the only one who thinks that the dude on the bike on the rocket pack page looks like a teen-aged Bob Lazar, the guy who claims to have worked at Area 51?!?!? Propane my eye... I'll bet that thing runs on antimatter for sure!
On a side note, check out the lame-ass shifter and the headlight mounted on the bike. All I can say is that I sure hope the rocket worked, or that dude must have gotten beat up a lot.
Looks like you need to spend a little more time in the math section of the library.
Actually, the book contained a very interesting point that relates directly to this.
The issue was that the German hacker was able to take a bunch of individual bits of information (which, by themselves, weren't classified) and bring them together to deduce some information which is classified.
Cliff Stoll mentioned that this type of research prompted some of the intelligence officials to propose a new classification ("sensitive") to represent information that wasn't classified but which could be used to deduce classified info.
I guess that's proving a barrier to this student.
As far as I can tell, in the U.S., if you throw something away, the police don't need a search warrant to obtain it. They can go through your trash all they want.
It goes likewise for eavesdropping on conversations. There's something called "reasonable expectation of privacy". If I'm talking to someone on a crowded street, then the cops can record my conversation and use it as evidence because I wasn't taking steps that someone would reasonably take if they were trying to keep something secret (as in, secret from everybody and not just from the police).
This DNA thing strikes me as something similar. Just like if you throw a murder weapon with your fingerprints on it into your trash, I think you're "discarding" your DNA if you were to, say, have children. The parallel to "reasonable expectation of privacy" is that, if you were someone who really didn't want their DNA (or half of it) out there running around loose, then you would elect to not have kids.
So, if you have kids, then I think that the cops should, clearly, have a right to use the DNA of those kids (provided they're over the age of consent and provided that they volunteer their DNA) to catch you. What's a little more murky is whether the cops should be allowed to catch you based on DNA from any familial relative... including ones you have no control over the production of (like, your parents, cousins, etc.). Again, I think it would hinge upon whether you had kids. Going back to "reasonable expectation of privacy". If you have kids, then you aren't exercising what little control you have over the dissemination of your DNA... so it must not matter to you that much... so everything's fair game at that point.
Legal issues aside, I must say that this is one really cool thing about DNA. Everybody knows the obvious advantages that DNA has over fingerprints: you get usable evidence from smaller samples, from a wider variety of fluids, smudges, etc. What's less obvious is the idea of "proximity". Two people with similar fingerprints are not necessarily closely related, and two people who are closely related don't necessarily have similar prints. With DNA, that's not the case. With DNA, you're able to tell when you've got some DNA from someone closely related to some "target DNA". Even cooler, you can probably tell how many generations away they are.
I remember seeing an interview with Mark Pesche, the dude who was regarded as the author of the VRML spec, and he was going on and on about how cumbersome it is to keep track of URL's when we could be navigating in a 3D space for our documents....
Could you just see that? "Come visit Jiffy Lube on the web! Start at the Origin, go down Street 1 until you come to the big purplish billboard, bear left and continue through the pasture... go under the spaceship and then head 4 spaces east and you can't miss us!". And this is more intuitive than "www.jiffylube.com" because... why?
I'm sure that, of those 18 improvements to the browser, many or all of them promised to "completely change the way we think about browsing". However, like VRML, it's not necessarily a change for the better.
First off, wouldn't they have to demonstrate that the file I'm sharing titled "Doobie Brothers - China Grove.mp3" is really what it says it is? And if they have to verify each song to be able to sue me for it, then wouldn't they have to download a whole lot of my songs? If that's the case, you'd think it'd be possible to make a P2P client that was resistant to repeated download requests from single (or small ranges of) IP's... in the same way that some network firewalls block port scanning.
Next, don't they have to show that I actually supplied mp3's to people and cost the RIAA revenue? How many downloads would they actually be able to prove? And how much lost revenue could there really be? I mean, when iTunes is selling songs for $1, not all of that is going to the record companies, right? So you figure that the lost revenue will be something less than $1 per song. At that rate, it would take quite a while to rack up even a measley $100 in potential damage to the record companies using a household broadband connection.
It makes me wonder if there'd be a need for a P2P client that was RIAA-judgement-aware... that would let you limit your "reparations" exposure by limiting the number of songs that could be downloaded per time period. In other words, if you used to spend, say, $50 per month on CD's, then you could tell the P2P client to, each month, only allow downloads until the amount the RIAA could sue you for went up by $50. This way, you could get access to a bunch of songs, and you're only maybe out that $50 per month.
Another thing I'm wondering about is the partial-download problem. With most P2P networks, you can download a song by getting a portion of it from many different sharers. If the RIAA downloads a song and it comes from 5 different people, are they expecting to (I know they're hoping to, but are they expecting to) be able to sue each of those five for the full "lost revenue" amount?
As time goes on, I keep discovering that more and more commonly-used cliche's trace back to famous pieces of literature. (ie, "Neither a borrower, nor a lender be", "Good fences make good neighbors", "Out, out.. damned spot!", etc.). It surprises me how ignorant most people (including me) are about where these came from.
Now, looking back on my English experiences, I think it would have been pretty cool if each student were given a phrase and they had to use the net to find out what literature it originally came from and have to read enough of the surrounding text to be able to describe the context of the scene where the phrase occured (like Lady MacBeth trying to wash the blood off, etc).
The idea of a rip-proof CD amuses me.
:)
Back when the web first started, there were a lot of web-page creators scrambling for ways to make their page viewable, but not able to be saved, printed... whatever. The end conclusion was always the same: "If it can be viewed, it can be printed".
The same goes for "rip-proof" CD's. At some point, it has to be listenable to a human. When that happens, the song is vulnerable to being copied.
The obvious way to do this is just to route your "Line Out" into your "Line In" on your PC and then just have a sound recorder going while your CD plays. Of course, this carries the problem of converting from digital, to analog, and then to digital again.
What's only a little less obvious and a little less difficult (so much so that I can hardly believe I haven't seen it available yet) would be to have a pseudo sound output device. Assuming that the CD would be playable (but not rippable) on a normal PC CD-ROM drive, you could tell your CD player app to use this pseudo sound device as the output. To the app, it would look like a regular sound card (kinda like how Adobe Acrobat appears to be a printer), but it would actually just write the digital data to a file (again, like Acrobat does).
The nice thing here is that, the CD could even be restricted to only being played on a DRM-enabled player. At some point, that player has to send the audio off to what it thinks are speakers. If you have a pseudo device that intercepts the audio, then there you go.
Of course, I wouldn't be surprised if Paladium had components to prevent this... but that's a different story. The point here is that, if you had a pseudo sound card, you could still rip AND keep it all digital. Granted, the rip would happen at 1x... but that's why I have a second PC in my office with lots of games on it.
However, let me also state that I think that this is because I only bother to actually enter the code that doesn't have bugs.
I'm quite serious here. Although I would be cranking out more code when I was younger, I'd end up debugging, deleting, and rewriting it for just as long as it took me to write the first draft. Nowadays, I'm armed with so much "been down that road before" experience that I can see where the coding perils are before I even start.
Look at it this way... if you wanted to make an expedition to the south pole, would you want the strapping teenager who goes "Cool, I'll grab my backpack and then we can go!", or the veteran who says "Okay... give me a few weeks to get our equipment and supplies and to plan our route..."?
Coding is kinda the same way. There are, at first glance, a million ways of getting to the goal, but many of them aren't going to work.... and many of those won't reveal themselves as such until it's too late. Just like an expedition, much of the work in coding is in the planning before you even start.
Now, on a more serious note, when I explain the geocaching thing to friends or co-workers, if they still look at me funny after I explain the whole GPS and prize-exchange thing, then I just use my standard: "Okay... just think of it as an excuse to go hiking.". Ultimately, I think that that's the end result: marginally more people getting out of their homes and walking around in nice, big public areas and getting some fresh air.
Now, although the rangers and park officials claim that they'd like more people to appreciate the outdoors, I think they'd prefer that they appreciate them from afar. Like just about every other line of work, I'm sure they often mumble amongst themselves: "This job would be great if it weren't for the customers".
Bottom line: If you want more people at the park, then you've got more people at the park. Deal with the increased foot traffic... that's what you're paid to do... to manage the park. However, if you *don't* want more people, then either close the damn park entirely or require permits or something. But don't go preaching about how people should get back to nature if you're not prepared for them to do it during your shift.
Besides, making a car-free city by making cars impossible ("Hey, let's flood the streets!") is not the goal. The goal is to make it possible for humans to retain their current level of productivity without needing cars. Also, within that goal, I think there's the implicit constraint that the cars don't get replaced with something just as bad (like boats in Venice, for example).
Personally, I think it's so that the little morons don't forget how to spell it, but it got me thinking about all of the possible contractions of "weblog" and why they weren't used...
"weblog" - could be confused with actual web logfiles... which, by contrast, contain useful information.
"eblog" - could be confused with driftwood washed up by the tide
"log" - could be confused with something that "rolls down stairs, alone or in pairs..."
"og" - is reserved for a blogger's contraction of "ogg".
"g" - it probably already some stupid IM acronym for something (like "gee" or something).
So... you see... "blog" was the only choice the morons had.
I wish SCO would just go ahead and announce what parts of the Linux kernel have their code so that we can get started on replacing it with something that works.
:P
What are they waiting for.... a bug-finder's fee?
I'm not sure I'd use Squid for caching. Although it's a great cache, it requires that users take action on their end in order to use it... which means that it won't get used all that much and you're likely to get support calls from those people who *do* try to use it. Instead, if it's worth it to you, you might want to look at some of the routers that have transparent caching. Although I've never used one, I'm told that Cisco (and, if so, then the others like Ascend, etc. have them) has routers that magically detect port-80 traffic and do caching.... in a way that is: A) transparent and, B) non-optional.
I'm not sure I'd block outbound SMTP... but it depends upon which set of hassles you like the most. There seems to be a slow (or, rather, glacial...) drift on the net toward authenticated SMTP for laptop users who move around a lot and who don't want to have to keep changing their outbound SMTP settings. If you block outbound SMTP, then those mobile users who *are* lucky enough to deal with a single, authenticated SMTP server out there... they'll be out of luck. With only 160 stations... and with them being stations with relatively low turnover (and high degree of being able to track down the "owner" of each station), you might get a lot of mileage out of a strict spam policy.
Of course, the other alternative (to authenticated SMTP) that I've seen for laptop users is to support assignment of the SMTP server through DHCP. I'm not sure how many DHCP servers (or client OS's, DHCP clients, and mail clients) support this... but you may want to add something like this to your wishlist if you do decide to block outbound SMTP.
Also, more than considering blocking outbound SMTP, consider blocking *inbound* port-80 web traffic.... and write a prohibition against web hosting into the agreement.
A markup language for real-time monitoring of remote events... I can see it now...
<ex-wife id=1>
<screwing>
milkman
</screwing>
</ex-wife>
<ex-wife id=2>
<screwing>
that dickhead with the Trans-Am
</screwing>
<spending>
most of my money
</spending>
</ex-wife>