Building height is just as bad as any other type of zoning. If you make the buildings short, they will have to spread out. The result is a nightmare like Phoenix where you can live in the city and work in the city and commute 30 miles one way to work.
Zoning should be restricted to keeping heavy industry reasonably separated from business and residential, but NOT separate business from residential. The heavy industry should also be located in one spot, so you can run a train out there. If you spread the factories out too much, you have to run 50 trains in different directions, which gets expensive. Topology is important.
Blame the city planners. Nobody can build anything convenient to me because of zoning. To go anywhere I have to drive, or walk 40 minutes at minimum. My next house is not going to be in the burbs. The quality of life here is not high enough.
Wireless networks are what we need. Whatever problems exist are at least partially because we don't own the wires. But citizens can have more control of a section of the spectrum, and we can build little networks with that.
But who really said that the Internet has to be a bunch of commercial sites and spam-laden e-mail? If individual servers get too cumbersome to use, then they will be replaced with something else. When http and port 80 become nothing but a vast marketing wasteland, we can make something new. We've got our own publishing rights, so we don't have to eat what's shoveled to us.
Item #1 from the game designer's book of good ideas:
1) When the player dies in a game, it's a good idea to render that part a little less realistically than other parts of the game. We do want them to play the game more than once.
There must be someone out there who is on our side who has access to System V source code, right? That's what SCO is saying.
So, let's find this person, and get him to do a comparison of source code for us. If SCO won't find the infringing code, then we should try to find it ourselves.
Obviously the kernel developers need OUR help to sort this sorry mess out. Everyone, please make a google news account ASAP and put your two cents in. If all of us together put our minds to it, and posted our opinions on that thread, I'm sure they would appreciate our help in solving this problem quickly and efficiently. Thanks.
Suppose you constructed a message that said in part: 'You have an interesting definition of "trivial".' You ran this through your perfect classification function and it said "NOT SPAM".
I received the message and said to myself "well, look at that spam." Clearly the spam classifier mis-identified the spam as being ham, because I would consider it spam.
I'm not trying to duck your question. OK, well maybe I am. I spent a bit of time thinking about it, and it seems that diagonalization would be hard, if not impossible.
My original point is that spam is partly in the eye of the beholder, and I'll leave my argument at that, and abandon the statement about exact equivalence to the halting problem.
Re:you're forgeting the test flights
on
Shuttle Politics
·
· Score: 2, Insightful
1) There were no suborbital shuttle tests. 2) The test flight shuttle was Enterprise. 3) The shuttle is 1970's technology.
Simple. Spams or programs can be represented as sequences of numbers, making them indistinguishable at the machine representation level. Take all of those bits of those numbers in a row and read them as a single number that represents that spam.
At this point, Cantor's diagonalization is trivial.
In Dec. 1995 I got a 1 gig drive for $500. Someone else posted that they can get a 120 gig drive for 80 cents a gigabyte. That's 625 times better in 7.5 years. That means roughly at the end of the year 2010, $100 will buy you about 7.8 million gigabytes.
That doesn't seem quite right, and I suspect that the recent market forces driving prices down might be skewing the results.
Maybe we should be looking at the size of the largest hard drive available instead of the price.
Classifying spam is essentially the same problem as classifying programs into those that terminate that those that don't (the halting problem). This leads us to the following conclusions:
1) Filtering spam is not trivial. A program that filters spam X% better than another program will be X^2% more complicated or worse.
2) You can't write a program that will filter perfectly. At best, all you can do is develop a set of heuristics that you hope aren't too complicated. The less complicated the heuristic, the fewer resources it will require.
3) There's a limit to how simple your heuristics can be.
4) The system of spam is not just the message: it's the spammer, plus the message, plus the recipient. This is because a certain message considered spam by some will not be considered spam by others. That means that the heuristics that account for the person reading the spam will be better than those that don't. The source of a spam is also important: a message consisting of a spam report to a spam newsgroup is not a spam, though it may contain a complete spam message.
5) The best spam filters will eventually be AI's that understand human language. That means that the ultime spam filter will require enough processing power to model human cognitive abilities. In short, you're going to see an endless increase in the number of processor cycles consumed by spam filters, asymptotically approaching the requirements of a full-up human brain simulation.
On the other hand, this will sell a hell of a lot of computers.
Re:Hot Tips for Slashdotters?
on
Wall Street Meat
·
· Score: 2, Funny
Slashdot Mirror
How about engine parts? They might make an engine that would run for a half million miles with normal oil changes.
Building height is just as bad as any other type of zoning. If you make the buildings short, they will have to spread out. The result is a nightmare like Phoenix where you can live in the city and work in the city and commute 30 miles one way to work.
Zoning should be restricted to keeping heavy industry reasonably separated from business and residential, but NOT separate business from residential. The heavy industry should also be located in one spot, so you can run a train out there. If you spread the factories out too much, you have to run 50 trains in different directions, which gets expensive. Topology is important.
It's probably not a joke. Plenty of morons actually think (if you can call it that) like that. Rush Limbaugh is the most famous example.
Dallas! It's a great place to live if you happen to be a pickup truck. Just who are we building these cities for, anyways?
Blame the city planners. Nobody can build anything convenient to me because of zoning. To go anywhere I have to drive, or walk 40 minutes at minimum. My next house is not going to be in the burbs. The quality of life here is not high enough.
This one also has a commercial off-the-shelf computer in the heart of the design:
It uses a Palm Pilot
They might do that. So we encode other protocols for a replacement to work on port 80. Regular http would be rejected.
Or, at least I hope it will...
Wireless networks are what we need. Whatever problems exist are at least partially because we don't own the wires. But citizens can have more control of a section of the spectrum, and we can build little networks with that.
But who really said that the Internet has to be a bunch of commercial sites and spam-laden e-mail? If individual servers get too cumbersome to use, then they will be replaced with something else. When http and port 80 become nothing but a vast marketing wasteland, we can make something new. We've got our own publishing rights, so we don't have to eat what's shoveled to us.
I hope he's good on TV, and has a penchant for writing to the masses. A strange pronounciation of a common word would help too.
We need someone to fill the shoes of Carl Sagan
Earth has a 'moon' in a very strange horseshoe shaped orbit, and it's named Cruithne.
m l
http://burtleburtle.net/bob/physics/cruithne.ht
It's pronounced Croo-EEN-ya, which is Celtic I think.
Item #1 from the game designer's book of good ideas:
1) When the player dies in a game, it's a good idea to render that part a little less realistically than other parts of the game. We do want them to play the game more than once.
And I even e-mailed daddypants to notify him that it was a dupe.
There must be someone out there who is on our side who has access to System V source code, right? That's what SCO is saying.
So, let's find this person, and get him to do a comparison of source code for us. If SCO won't find the infringing code, then we should try to find it ourselves.
Obviously the kernel developers need OUR help to sort this sorry mess out. Everyone, please make a google news account ASAP and put your two cents in. If all of us together put our minds to it, and posted our opinions on that thread, I'm sure they would appreciate our help in solving this problem quickly and efficiently. Thanks.
That's pretty cool. A nail gun that doesn't need either an air compressor, an extension cord, or a battery.
You mean it comes with a little band that plays "Nearer my God to Thee" when the server crashes? Cool!
There are only three networks that carry cartoons on Saturday mornings?
Excuse me, but this is EXACTLY what it was like in the 1970's. We had ABC, NBC, and CBS. PBS didn't have any cartoons worth watching.
So why is this news again?
Whoops.
l ?m enu=
http://www.ananova.com/news/story/sm_283669.htm
Let's get real theoretical now: :-)
Suppose you constructed a message that said in part: 'You have an interesting definition of "trivial".' You ran this through your perfect classification function and it said "NOT SPAM".
I received the message and said to myself "well, look at that spam." Clearly the spam classifier mis-identified the spam as being ham, because I would consider it spam.
I'm not trying to duck your question. OK, well maybe I am. I spent a bit of time thinking about it, and it seems that diagonalization would be hard, if not impossible.
My original point is that spam is partly in the eye of the beholder, and I'll leave my argument at that, and abandon the statement about exact equivalence to the halting problem.
1) There were no suborbital shuttle tests.
2) The test flight shuttle was Enterprise.
3) The shuttle is 1970's technology.
Simple. Spams or programs can be represented as sequences of numbers, making them indistinguishable at the machine representation level. Take all of those bits of those numbers in a row and read them as a single number that represents that spam.
At this point, Cantor's diagonalization is trivial.
Ah crap, I think I mixed up dollars and cents, so I'm 100 times too large. Should be 78 terabytes for $100 in 7.5 years.
In Dec. 1995 I got a 1 gig drive for $500. Someone else posted that they can get a 120 gig drive for 80 cents a gigabyte. That's 625 times better in 7.5 years. That means roughly at the end of the year 2010, $100 will buy you about 7.8 million gigabytes.
That doesn't seem quite right, and I suspect that the recent market forces driving prices down might be skewing the results.
Maybe we should be looking at the size of the largest hard drive available instead of the price.
Classifying spam is essentially the same problem as classifying programs into those that terminate that those that don't (the halting problem). This leads us to the following conclusions:
1) Filtering spam is not trivial. A program that filters spam X% better than another program will be X^2% more complicated or worse.
2) You can't write a program that will filter perfectly. At best, all you can do is develop a set of heuristics that you hope aren't too complicated. The less complicated the heuristic, the fewer resources it will require.
3) There's a limit to how simple your heuristics can be.
4) The system of spam is not just the message: it's the spammer, plus the message, plus the recipient. This is because a certain message considered spam by some will not be considered spam by others. That means that the heuristics that account for the person reading the spam will be better than those that don't. The source of a spam is also important: a message consisting of a spam report to a spam newsgroup is not a spam, though it may contain a complete spam message.
5) The best spam filters will eventually be AI's that understand human language. That means that the ultime spam filter will require enough processing power to model human cognitive abilities. In short, you're going to see an endless increase in the number of processor cycles consumed by spam filters, asymptotically approaching the requirements of a full-up human brain simulation.
On the other hand, this will sell a hell of a lot of computers.
Buy low, sell high.