Actually, I was worried already about the way
gentoo might be heading. See
this
proposal which seems to make a lot of people enthusiastic. It talks about a system where "everyone" can contribute with packages/ebuilds and there is a voting system that determines whether this package is tagged new, or approved or whatever. There is a thread on the Debian mailing list about it starting here.
Isn't there a "largest" known prime number? Take double that + 2 and isn't it disproven?
Wow, you've solved goldbach! Just kidding. No, there are infinitly many primes due to a trick of Euclid: suppose there are finitely many: P1,...pn. Multiply and add one: p1...pn+1. This number is divisible by a prime number (as are all numbers) and this number cannot be p1 or p2 or... or pn. So, this is a new prime number. Contradiction.
Some people are now even worried about email messages containing ASCII! My girlfriend sent a message to someone in a dutch organization `cadans' in ASCII and got the following automated response (in dutch, which I translated to english for your convenience):
Dear Sender,
The message you sent, does not pass the security policy of our organisation.
The files we do let through are: word documents excel documents powerpoint documents
And if you want to waste money solving an undecidable problem, it's your money, not mine...:-)
Trying to solve an undecidable problem might not be such a good idea, but that doesn't mean we should immediately stop doing research. Take the Halting Problem, for instance. As you point out, there is no algorithm that tells you if a computer program halts, but that doesn't mean we should not try to find an algorithm that that solves the halting problem for everyday programs. This might also be a very hard problem, but there are a lot of programs where it is easy to see that it will stop.
If you think trying to solve the halting problem
`as good as you can' is still silly, consider another example. Finding the limit of some mathematical function in infinity is in general also undecidable. Yet, there are many math packages that try to solve the problem as good as they can. And there is a lot of research done.
Slashdot Mirror
Actually, I was worried already about the way gentoo might be heading. See this proposal which seems to make a lot of people enthusiastic. It talks about a system where "everyone" can contribute with packages/ebuilds and there is a voting system that determines whether this package is tagged new, or approved or whatever. There is a thread on the Debian mailing list about it starting here.
Richard
Wow, you've solved goldbach! Just kidding. No, there are infinitly many primes due to a trick of Euclid: suppose there are finitely many: P1,...pn. Multiply and add one: p1...pn+1. This number is divisible by a prime number (as are all numbers) and this number cannot be p1 or p2 or ... or pn. So, this is a new prime number. Contradiction.
Richard
I wish I was joking.
Richard
Trying to solve an undecidable problem might not be such a good idea, but that doesn't mean we should immediately stop doing research. Take the Halting Problem, for instance. As you point out, there is no algorithm that tells you if a computer program halts, but that doesn't mean we should not try to find an algorithm that that solves the halting problem for everyday programs. This might also be a very hard problem, but there are a lot of programs where it is easy to see that it will stop.
If you think trying to solve the halting problem `as good as you can' is still silly, consider another example. Finding the limit of some mathematical function in infinity is in general also undecidable. Yet, there are many math packages that try to solve the problem as good as they can. And there is a lot of research done.