the point made still has some value. Whenever individiuals take it upon them to enforce what is right and wrong, there's a risk of abuse or simply them doing a poor job.
But I guess (hope) that all who use these blacklists, monitor how it's working and are prepared to act if errors are being made. If the guy who runs your mailserver thinks it's added value for his users to block from servers on the lists, then he's going to use them - otherwise not.
The speed of a drive consist of both raw transfer rate and seek time.
Let's look at transfer rate first:
If you double the density how will this affect transfer rate? Let's assume that the increase (it's doubled) in density is achieved by having sqrt(2) times more tracks and sqrt(2) times more bits in each track - a fair assumption IMO. The transfer rate of a new 5400 RPM disk compared to an old 7200 is then (5400*1.41)/7200 = 7636 / 7200 = 1.06. The 320 GB disk's transfer rate is 6% better - not very impressive.
The theoretical average seek time for the old 7200 RPM drive on the other hand is 1/(7200*2) = 6.9 ms compared to the new disk's 1/(5400*2) = 9.3 ms. That's 25% better - which I think is quite a lot.
In real life I think that you'll find an old 7200 RPM drive quite a bit snappier than a new 5400.
So now it's possible to fit a thousand ripped movies on a disk instead of 500 - big deal. What the harddisk manufacturers should focus on is increasing RPM. The speed of harddisks is one of the slowest growing figures in all of the IT business - if not *the* slowest, and with bloatware abundant this means sluggish boot-up and starting of programs.
The 320 GB disk is even only 5400 RPM - that's as slow as the good old Bigfoot disks of times forgotten (Remember? they were the size of a showbox and a lot heavier...).
I guess this disk is good news for the P2P community, but for serious use it is of very limited interest.
Actually luck is a factor in this
on
Draw!
·
· Score: 1
A very basic concept in chess computers is the min-maxing:
At turn n lets say the computer looks 4 turns ahead, including the current (very simple example): The algorithm chooses the most favorable move in its own turns (turn n and n+2), and the least favorable in the opponents turns (turn n+1 and n+3). The best achieveable result possible for a given move is hereby found and stored. When all moves has been examined the one with the highest min-maxing score is chosen.
A very important extension of this principle is pruning (yes it's called that....), which means that after the first move has been analysed completely and the min-maxing score found, the algorithm needs only check other moves until the opponent in his turn (n+1 and n+3 in the example) has the opportunity to chose a move with a worse min-max score. This cuts off whole "branches" of moves to search.
And although you can optimize the choice of which moves to search first (you'll want to search the best moves first, so that you can quickly prune later moves) you can never be sure. And that introduces an element of chance into an otherwise luck-free game.
Of course with the amount of iterations of the algorithm, it's very unlikely that this will not even out.
- Skov
Let's put an end to the rounding-confusion
on
Pet Bugs?
·
· Score: 1
There seems to be some uncertainty about all this rounding, so lets take it step by step:
We have a number n with a certain precision pn.
We need to represent it as a number r with a lower precision pr.
We want r to be as close to n as possible.
The rounding must not have a tendency to increase or decrease the value of n.
Whenever we round n to r the last pn - pb digits are thrown away as they are implicitly 0's in r (ie 5.1 = 5.1000000.....)
Unless the digits we throw away are all 0's, we are going to alter the value of our number, and it is this alteration we wish to minimize. We do that by rounding to the number of precision pr that is closest to n.
This is trivial for all numbers except those where the most significand digit lost in the rounding process is 5 and all the rest (if any) of the lost digits are 0's (f.ex. when rounding 4.5 or 4.50000 to the nearest whole number). These are not trivial as rounding up and down changes n by the same amount. In other words: Two numbers of precision pr are equally close to n.
The easy solution is then to just choose to always round up or down, and stick to that. It's an optimal solution when it comes to minimizing the difference between n and r, since there was no better solution than the one we chose (only one that was just as good). But does the solution have a tendency to either increase or decrease the value of n?
Look at all numbers of precision pn between two adjacent numbers, r1 and r2 of precision pr, adjacent meaning that there is no number rn of precision pr such that r1 < rn < r2. For every number n1 that is rounded towards r1 by an amount m there exists a number n2 that is rounded to r2 by the same amount, except where m = ((r1 - r2) / 2), which is excactly the non-trivial numbers discussed above. That means that in order for our rounding to be symmetrical, the rounding of the non-trivial numbers has to be symmetrical. And it is not symmetrical when we have chosen to always round up or down. And that is why other methods of rounding has been introduced - round to even being the most widely known.
I hope this can put an end to all the "count-the-roundings"-posts... I would have expected stuff like that from no-brainer journalists reporting on stuff they don't understand - but from inside the slashdot-crowd? never! Its devastating for our reputation for being brainy...
I really think you are missing the point with this question. The essence of all this, is that instead of using a fantazillion clockycycles on idle-threads every seconds, we can use them for whatever we want.
Dnet was the first distributed project I learned about. Now it's a funny little part of my life to drop in now and then and see how things are doing. To see how I'm doing, smiling at the-space.net making fun of the dutch cows when they overtook them for the first time and Killemall getting impatient.
After seeing the SETI@home client and all its bright colours, I've decided to shift to them when RC5 is over. Not because it is a worthier cause, but because it looks funny.
Now that, in my opinion, is the point of all this.
It's somewhat depressing that my favourite nerd-website surpasses me, even when it comes to being romantic. You know, I don't come here to get my inadequacies slapped in the face.
But what the heck... Congratulations!! Make it up to me by being very, very happy.
This post is so full of false claims, I just have to interfere:
A)
In addition, human populations are now being constantly mixed, again producing a blending that blocks evolutionary change. This increased mixing can be gauged by calculating the number of miles between a person's birthplace and his or her partner's, then between their parents' birthplaces, and finally, between their grandparents'.
In virtually every case, you will find that the number of miles drops dramatically the more that you head back into the past. Now people are going to universities and colleges where they meet and marry people from other continents. A generation ago, men and women rarely mated with anyone from a different town or city. Hence, the blending of our genes which will soon produce a uniformly brown-skinned population. Apart from that, there will be little change in the species.
Not only is this totally racist and white supremist horseshit, it is completely wrong.
Whatever qualification Prof. Steve Jones holds, he should probably take down his degree and wipe his arse with it, as it has turned out that is all it's good for.
It is beyond me how you find anything racist in this. Allow me to clarify: Something isn't racist just because it deals with racial issues - it must also discriminate. Stating the obvious facts that people mix more today than they used to, and that this will create a more homogenous world population is in no way racist.
B)
Evolution works by trying combinations. When one particular combination hits exactly right for the current conditions at the current moment in time the result is a sudden and exponential success.
Correct - but not enough. In addition to a lucky combination of genes, there is one more requirement for any evolutionary effect: selection. The point of the article is *not* that there is less chance of lucky combinations of genes - the point is that these fortunate new humans have no advantage to everybody else. Therefore the race as a whole will only benefit extremely marginally from this contribution to the gene-pool. And for every beneficial mutation, there's a thousand harmful. But without any selection each of these thousands of disadvantaged individuals will contribute as much to the next generation as the single lucky one. This adds up to a general degeneration.
Selection is not completely forgotten in the post though:
And with all this greater health, and wealth, and energy, they will produce A LOT MORE CHILDREN than the average person
No, no, no. As is widely known - and described in an earlier post - succes does *not* result in more kids. Quite the opposite actually. Those who don't win the nobel-prize or run a multi-billion company tend to produce more offspring instead, put bruntly. So we actually have a selection towards the lower end of the spectre.
Next up: Common colds second wind:
C)
Modern medicine knows no cure for the COMMON COLD!! How many more diseases are we completely at a loss to stop right now?? Can you imagine a cold strain escaping from Shanghai, or Calcutta?
The people living in those cities are the survivors. Every year simple diseases kill people in the developing world. The local population builds a resistance. The disease mutates and kills again. The local population builds more resistance. And so on and so forth.
Westerners, living in their sterile and hygenic conditions, eating denatured food full of salt, fat and sugar, won't have any resistance to these viscious new cold strains.
This is an evolutionary event just waiting to happen.
Let me get this straight: In Shanghai or Calcutta, a vicious variation of the common cold is contained completely from the rest of the world, and if it ever leaks out 99% of us are doomed because we have weakened resistances?
I suppose it's just dumb luck that none of the thousands of europeans and americans who visit these places every year, haven't caught this deadly flu yet? Of course not. When we go abroad, we get a stomach ache, because the local set of diseases are so unfamiliar to what we're used to, but thats it. The concept of any part of the world being isolated in respect to diseases is ludicrous. There are plenty of scary bio-hazardous scenarios to ponder about - but this is definitely not one of them.
Slashdot Mirror
the point made still has some value. Whenever individiuals take it upon them to enforce what is right and wrong, there's a risk of abuse or simply them doing a poor job.
But I guess (hope) that all who use these blacklists, monitor how it's working and are prepared to act if errors are being made. If the guy who runs your mailserver thinks it's added value for his users to block from servers on the lists, then he's going to use them - otherwise not.
As I state here. Increase in density is 2-dimensional (area) while transfer is 1-dimensional (line).
The speed of a drive consist of both raw transfer rate and seek time.
Let's look at transfer rate first:
If you double the density how will this affect transfer rate? Let's assume that the increase (it's doubled) in density is achieved by having sqrt(2) times more tracks and sqrt(2) times more bits in each track - a fair assumption IMO. The transfer rate of a new 5400 RPM disk compared to an old 7200 is then (5400*1.41)/7200 = 7636 / 7200 = 1.06. The 320 GB disk's transfer rate is 6% better - not very impressive.
The theoretical average seek time for the old 7200 RPM drive on the other hand is 1/(7200*2) = 6.9 ms compared to the new disk's 1/(5400*2) = 9.3 ms. That's 25% better - which I think is quite a lot.
In real life I think that you'll find an old 7200 RPM drive quite a bit snappier than a new 5400.
So now it's possible to fit a thousand ripped movies on a disk instead of 500 - big deal. What the harddisk manufacturers should focus on is increasing RPM. The speed of harddisks is one of the slowest growing figures in all of the IT business - if not *the* slowest, and with bloatware abundant this means sluggish boot-up and starting of programs.
The 320 GB disk is even only 5400 RPM - that's as slow as the good old Bigfoot disks of times forgotten (Remember? they were the size of a showbox and a lot heavier...).
I guess this disk is good news for the P2P community, but for serious use it is of very limited interest.
A very basic concept in chess computers is the min-maxing:
At turn n lets say the computer looks 4 turns ahead, including the current (very simple example): The algorithm chooses the most favorable move in its own turns (turn n and n+2), and the least favorable in the opponents turns (turn n+1 and n+3). The best achieveable result possible for a given move is hereby found and stored. When all moves has been examined the one with the highest min-maxing score is chosen.
A very important extension of this principle is pruning (yes it's called that....), which means that after the first move has been analysed completely and the min-maxing score found, the algorithm needs only check other moves until the opponent in his turn (n+1 and n+3 in the example) has the opportunity to chose a move with a worse min-max score. This cuts off whole "branches" of moves to search.
And although you can optimize the choice of which moves to search first (you'll want to search the best moves first, so that you can quickly prune later moves) you can never be sure. And that introduces an element of chance into an otherwise luck-free game.
Of course with the amount of iterations of the algorithm, it's very unlikely that this will not even out.
- Skov
We have a number n with a certain precision pn.
We need to represent it as a number r with a lower precision pr.
We want r to be as close to n as possible.
The rounding must not have a tendency to increase or decrease the value of n.
Whenever we round n to r the last pn - pb digits are thrown away as they are implicitly 0's in r (ie 5.1 = 5.1000000.....)
Unless the digits we throw away are all 0's, we are going to alter the value of our number, and it is this alteration we wish to minimize. We do that by rounding to the number of precision pr that is closest to n.
This is trivial for all numbers except those where the most significand digit lost in the rounding process is 5 and all the rest (if any) of the lost digits are 0's (f.ex. when rounding 4.5 or 4.50000 to the nearest whole number). These are not trivial as rounding up and down changes n by the same amount. In other words: Two numbers of precision pr are equally close to n.
The easy solution is then to just choose to always round up or down, and stick to that. It's an optimal solution when it comes to minimizing the difference between n and r, since there was no better solution than the one we chose (only one that was just as good). But does the solution have a tendency to either increase or decrease the value of n?
Look at all numbers of precision pn between two adjacent numbers, r1 and r2 of precision pr, adjacent meaning that there is no number rn of precision pr such that r1 < rn < r2. For every number n1 that is rounded towards r1 by an amount m there exists a number n2 that is rounded to r2 by the same amount, except where m = ((r1 - r2) / 2), which is excactly the non-trivial numbers discussed above. That means that in order for our rounding to be symmetrical, the rounding of the non-trivial numbers has to be symmetrical. And it is not symmetrical when we have chosen to always round up or down. And that is why other methods of rounding has been introduced - round to even being the most widely known.
I hope this can put an end to all the "count-the-roundings"-posts... I would have expected stuff like that from no-brainer journalists reporting on stuff they don't understand - but from inside the slashdot-crowd? never! Its devastating for our reputation for being brainy...
- Skov
I really think you are missing the point with this question. The essence of all this, is that instead of using a fantazillion clockycycles on idle-threads every seconds, we can use them for whatever we want.
Dnet was the first distributed project I learned about. Now it's a funny little part of my life to drop in now and then and see how things are doing. To see how I'm doing, smiling at the-space.net making fun of the dutch cows when they overtook them for the first time and Killemall getting impatient.
After seeing the SETI@home client and all its bright colours, I've decided to shift to them when RC5 is over. Not because it is a worthier cause, but because it looks funny.
Now that, in my opinion, is the point of all this.
It's somewhat depressing that my favourite nerd-website surpasses me, even when it comes to being romantic. You know, I don't come here to get my inadequacies slapped in the face.
But what the heck... Congratulations!! Make it up to me by being very, very happy.
A) It is beyond me how you find anything racist in this. Allow me to clarify: Something isn't racist just because it deals with racial issues - it must also discriminate. Stating the obvious facts that people mix more today than they used to, and that this will create a more homogenous world population is in no way racist.
B) Correct - but not enough. In addition to a lucky combination of genes, there is one more requirement for any evolutionary effect: selection. The point of the article is *not* that there is less chance of lucky combinations of genes - the point is that these fortunate new humans have no advantage to everybody else. Therefore the race as a whole will only benefit extremely marginally from this contribution to the gene-pool. And for every beneficial mutation, there's a thousand harmful. But without any selection each of these thousands of disadvantaged individuals will contribute as much to the next generation as the single lucky one. This adds up to a general degeneration.
Selection is not completely forgotten in the post though:
No, no, no. As is widely known - and described in an earlier post - succes does *not* result in more kids. Quite the opposite actually. Those who don't win the nobel-prize or run a multi-billion company tend to produce more offspring instead, put bruntly. So we actually have a selection towards the lower end of the spectre.
Next up: Common colds second wind:
C) Let me get this straight: In Shanghai or Calcutta, a vicious variation of the common cold is contained completely from the rest of the world, and if it ever leaks out 99% of us are doomed because we have weakened resistances?
I suppose it's just dumb luck that none of the thousands of europeans and americans who visit these places every year, haven't caught this deadly flu yet? Of course not. When we go abroad, we get a stomach ache, because the local set of diseases are so unfamiliar to what we're used to, but thats it. The concept of any part of the world being isolated in respect to diseases is ludicrous. There are plenty of scary bio-hazardous scenarios to ponder about - but this is definitely not one of them.