As I have always understood it, evolution in and of itself implies beneficial changes that allow the species to survive. The individual changing you are referring to is mutation, not evolution. Mutation is a part of evolution, wherein the mutations that benefit a species tend to be passed on, but evolution is by definition always an overall improvement.
And if the "survivability" of everyone goes to zero in the long-time limit, well
maybe that is where some religious guys could take their inspiration from evolution?
Evolution doesn't have a standpoint. Would you claim that ohm's law is false on the grounds that whatever happens to the current is fine from the voltage's standpoint?
So you deny a standpoint to evolution but nonetheless talk about "grounds" from which
one can, well, what else?, have a standpoint to talk about Ohm's law?
Imagine a species has 100 million members, and lets say it is a large-sized species which experiences a generation turnover every 20 years or so. Lets say there is a low mutation rate of perhaps 1% of offspring having some mutation. Let us also say that 99% of mutations are harmful, or perhaps even fatal, and a mere 1% are beneficial. Now we do the math:
But before we do the math, don't we need to figure out how to determine what
the words "harmful" and "beneficial" mean? Naturally, we can just keep on with
doing math, but you also keep on with the same terminology throughout the
calculation, like when you say "Lets say each beneficial mutation is only
beneficial by a very tiny amount", which makes me think that you assume that
there are some kind of "amounts" of "beneficial" and "harmful", that these are
mathematical quantities described by numbers. So how do you determine these
numbers?
Real, hardcore, scientific guys know that the word 'evolve' does not imply objective betterment of the organism/population in question. The idea of evolution as a process of constant improvement is a common misconception among laymen. Organisms just evolve, ie. they become different from their ancestors. Whether this change makes them more or less fit (depends on the conditions) doesn't change the concept of evolving.
Does this mean that whatever happens to organisms/populations is an evolutionary process?
What is then "the concept of evolving" according to these "real, hardcore, scientific guys"
whose knowledge you know so well?
Your message reminds me of the Popper's objection to evolution: it is impossible to disprove it since whichever way organisms turn out is fine from the evolutionary standpoint. He concluded then that evolution is not a scientific theory according to his definition.
Third, merge sort is O(n log n) time, hard to beat for random data.
Provably impossible, in fact.
So what is the proof of impossibility, in fact?
to which you answer:
Please make at least a vague attempt to research before you post. Googling for "proof sort nlogn" turns up several papers giving the standard arguments for the lower bound.
which does not answer the question, does it?
(if the lower bound for merge-sort is O(N log N), then what is "the proof of impossibility, in fact",
in fact?)
(And yes, I'm only talking about comparison sorting here. Obviously you can use radix/bucket/whatever for specialist applications and these are lower than the bound, but they're not exactly equivalent algorithms to the merge sort referenced in the headline, are they?)
(every sorting algorithm is comparison based, by definition. Then "specialist application" means that there is some additional knowledge about objects that simplifies sorting, like properties of associated keys or, for a (non)trivial example, when one wants to sort an array that is known to be in descending order into ascending order, etc. Once one learns something about the dataset one is sorting, it might be preferable to get an efficient sorting algorithm from another standard sorting algorithm even if they are "equivalent".)
If we're dealing entirely with unique keys, this is of course >= log(N), by the pigeonhole principle and all that. If we may have duplicate keys, however, there may be more keys than container space, and radix *does* scale linearly to the keys.
What is then the sorting time of the Radix algorithm?
I mean, mathematically it cannot be just O(k*N) where k is the
number of keys, because k depends on N in general. Now, if all
the keys are different (which is for many applications the most
important case), what is the sorting time in terms of N?
As far as I know, Quicksort's typical sorting times are
much better than those of the Radix.
I'd definitely recommend reading the article.
The best part concerns Germany:
"This is an incredibly bad thing in terms of privacy, since people have grown up with the idea that you ought to be able to have an anonymous e-mail account," Mr. Fleischer said. "Moreover, it's totally unenforceable and would never work."
Ok, now that the thing is incredibly bad and would never work, the article gives
Mr. Fleischer's explanation on how to do it:
Mr. Fleischer said the law would have to require some kind of identity verification, "like you may have to register for an e-mail address with your national ID card."
Like imagine rows of grudging Germans waiting in a perfect line to register
their non-anonymous private e-mails. Ironic?
Mr. Fleischer said: "It's ironic, because Germany is one of the countries in Europe where people talk the most about privacy. In terms of consciousness of privacy in general, I would put Germany at the extreme end."
Ironic indeed if you ask yourself which end is this "extreme end" and what kind
of consciousness of privacy is in question.
Hint: I have heard (here in Germany) with my own ears a friend complaining to a friend:
"You behave very individualistically lately!" [translation mine].
Slashdot Mirror
What is then "the concept of evolving" according to these "real, hardcore, scientific guys" whose knowledge you know so well?
Your message reminds me of the Popper's objection to evolution: it is impossible to disprove it since whichever way organisms turn out is fine from the evolutionary standpoint. He concluded then that evolution is not a scientific theory according to his definition.
(if the lower bound for merge-sort is O(N log N), then what is "the proof of impossibility, in fact", in fact?) (every sorting algorithm is comparison based, by definition. Then "specialist application" means that there is some additional knowledge about objects that simplifies sorting, like properties of associated keys or, for a (non)trivial example, when one wants to sort an array that is known to be in descending order into ascending order, etc. Once one learns something about the dataset one is sorting, it might be preferable to get an efficient sorting algorithm from another standard sorting algorithm even if they are "equivalent".)