I'm not sure if it's 50m (I'm not the one making that claim), but you can pick the chunk size you want and work out how much damage it will do. Do also work out how many chunks you will get from your chosen source asteroid.
Being nuked by thousands of small nukes is also quite damaging (a 1000 km diameter asteroid should be able to produce a few thousand 50km diameter chunks).
Because deflecting an asteroid that's far away from earth is easier than reliably blasting it to small enough bits.
Think of how much force it takes to nudge a cue ball away from its original destination. Compare with the force it takes to blast a rock to small bits.
If a big asteroid is already too close to nudge away, we're screwed.
By posting here or anywhere else in this forum you agree to: 1) give me all your money, assets and all future earnings and assets. 2) whenever there is a full moon, stand in a public area on one foot and howl at the moon. 3) Say "boop" every 87.24 minutes.
You really think Yahoo will give Google and friends all the relevant details on all the users for cheap? Each of them will sell demographics (grouped), or detailed info on a particular user, but good luck buying the "everything" for a low price.
So it costs more to combine the baskets.
You don't use your real name on Slashdot. You don't have to use your real name on Yahoo or Hotmail.You don't have to use your real birthday, timezone, country, etc either.
If you just went "Google" (or Yahoo) for everything (mail, plus etc) they would have everything nice and neat, no need to do the extra work, no need to pay competitors etc for every bit of data. So even if you lied about who you are to Google, they would have everything on "user #11231323". So if one day someone says "hey tell me everything you know about this youtube user" and gives Google some money, google can sell everything they have about you.
In contrast if you split everything up, they'd have to go to Yahoo, Facebook, Microsoft etc one by one and say "tell me everything you know about this youtube user". And I bet they'd be charged a lot more, and might even be given a lot less.
You get more bang for your buck with read caching (more activities read from disk than write to it)
Even if you have 10 times more reads than writes, if your reads can be done at 50000/sec but your writes can only be done at 200-1000/sec (1-5 ms access time), then your writes remain a noticeable bottleneck.
And If you're still using XP you might be interested to read this: http://technet.microsoft.com/en-us/library/cc959914.aspx For every scan of a directory (even cached) XP by default will WRITE to the drive just to update the last access timestamps!
Linux had that problem too but see noatime and relatime.
KDE and GNOME suffer from the "Microsoft Windows disease".
What they should do is just come up with something that works very like Windows XP (ala ReactOS). So that when Microsoft finally kills XP (in 2014?), they suddenly get half the XP market share.
If it were up to me I'd cache the stuff that would make the disk head seek the most, and not cache sequential access (unless it happens often enough).
You don't cache stuff that goes: seek 1 millisecond, then sequential read for 10 seconds.You cache stuff that goes seek 1 millisecond, sequential read 5 milliseconds seek 1 millisecond, sequential read 10 milliseconds.
Cache the small stuff, and cache some filesystem metadata at higher priority. 20 years ago I wrote disk caching software to do similar stuff for the Apple IIGS. I think the hybrid drives better do something more intelligent than mere read caching, or most people aren't going to bother.
With all the genius programmers they have, they should use an algorithm that understands _time_ so if the drive is spending lots of time waiting to write/read something to/from the disk, then it should try to cache it. Maybe have a drive mood or something - so if the drive is waiting more, it mood changes to "cache more current stuff", otherwise it goes to "cache less of current stuff".
Seems to me the current hybrid drives don't do write-caching, they only do read-caching.
I can see why read-caching would be a lot simpler to implement, but I bet decent write-caching will really make hybrid drives as fast as SSDs for most desktop use.
Copying files from one location to another at SSD speeds till the write-cache fills up. Then while you do something else the drive flushes the cache to disk.
I'd thought they would move stuff around when that happens ( and that's why performance goes down if you don't have trim and the drive doesn't know which blocks can be overwritten without moving them).
It just means the people actually using that data, marketers, FBI, underpants gnomes, have to buy it from more than one source and then merge it.
I already said that.
And what are you going to do, run your own search engine? Run your own mail system for throwaway accounts? Run your own "facebook" with nobody else but you on it? Run your own "Amazon", "newegg", "ebay" with nobody buying or selling stuff on them?
The banks are just shifting more and more risks and responsibilities for losses to their customers.
They prefer to call stuff ID Theft rather than some sort of fraud. Since with ID Theft it's their customer's problem, whereas with fraud it might be their problem.
They also prefer debit cards. With credit cards, when "stuff happens", it's not my money that's gone, it's someone else's. They may try to get the money from me, but meanwhile I have my money. Whereas with debit cards, when "stuff happens" it's my money that's gone. In theory I might legally be entitled to get it back, but meanwhile I have lost my money.
Of course they are, but if Yahoo has part of the data and google has the some of it, facebook/etc the rest, they are less likely to share the data with each other. Better than Yahoo or Google having 100% of the data.
Make the various parties work and pay more for it.
If ReactOS actually becomes stable and usable I suspect a number of large US corporations might be interested too.
There are lots of corporations that are happy with Windows XP, and the only problems with it is the bugs are no longer going to be fixed, and MS will stop selling it.
I simply refuse to accept any argument that the markets are distinguishable from casino gambling (other than having an exceptionally toothless gambling commission).
Here's the difference:
In the financial world when you gamble with other people's money and lose big: a) you get a bailout and get to keep your bonuses and previous cut of the winnings. or b) your transactions get rolled back (see the HFT rollbacks).
In casinos when you gamble with other people's money and lose big, I don't think the outcomes are as rosy.
It's not mere toothlessness when they rollback HFT trades because of bugs in favoured _players_ systems (it's fine to rollback if it's due to bugs in the "casino" systems).
During colonial times, lobster was food for the poverty stricken, prisoners and indentured servants. In the Massachusetts colony that encompassed the land that became known as Maine, indentured servants protested and had instructions written in to their seven-year contracts that they would not be forced to eat lobster more than three times a week.
Far fewer people are allergic to fish, chicken, beef than they are to shrimp, crab, lobsters. Or even dust mites. So I wouldn't be surprised if many are also allergic to these "popular" arthropods.
Intel needs help breaking into the smartphone market and Nokia isn't tied as tightly to Qualcomm/ARM hardware as other vendors.
From what I see Nokia has no idea on how to break into the smartphone market. So why would buying them help Intel?
Just poach the few talented people they need from Nokia. Why pay lots of money to get saddled with all the crap and baggage that got Nokia in deep shit?
Nokia's patents are not what get you into the smartphone market. Otherwise Nokia would be in the smartphone market.
They could do an "On Live" sort of game service too, but with Linux as an option for the platform.
For people with weaker clients but high bandwidth low latency network connections it might be even faster - level load times can actually be shorter, frame rates could be higher.
If they are playing against other players on the same "cloud", the ping times won't really be that different. There's higher display lag - but that doesn't matter as much for some games/people.
On non-ARM systems, it is required to implement the ability to disable Secure Boot via firmware setup.
So despite Intel's efforts on getting power consumption down I guess there's not going to be a Windows phone or tablet on x86...;)
A Windows Server may also disable Secure Boot remotely using a strongly authenticated (preferably public-key based) out-of-band management connection, such as to a baseboard management controller or service processor.
I'm sure some hackers will be interested in investigations the various implementations of this feature.
Slashdot Mirror
http://impact.ese.ic.ac.uk/ImpactEffects/
http://simulator.down2earth.eu/
I'm not sure if it's 50m (I'm not the one making that claim), but you can pick the chunk size you want and work out how much damage it will do. Do also work out how many chunks you will get from your chosen source asteroid.
Being nuked by thousands of small nukes is also quite damaging (a 1000 km diameter asteroid should be able to produce a few thousand 50km diameter chunks).
Because deflecting an asteroid that's far away from earth is easier than reliably blasting it to small enough bits.
Think of how much force it takes to nudge a cue ball away from its original destination. Compare with the force it takes to blast a rock to small bits.
If a big asteroid is already too close to nudge away, we're screwed.
Julian May
http://en.wikipedia.org/wiki/Saga_of_Pliocene_Exile
By posting here or anywhere else in this forum you agree to:
1) give me all your money, assets and all future earnings and assets.
2) whenever there is a full moon, stand in a public area on one foot and howl at the moon.
3) Say "boop" every 87.24 minutes.
All legal right?
I think the real problem is if you are doing rm -rf bar/foo
and somehow accidents happened with the spacebar...
The scary ones I see are the < and > keys are close to each other on most US style keyboards, and so are the i and o keys (think dd).
So that's one of the reasons why I use cat instead of < and ignore those who harp about "useless use of cat".
You really think Yahoo will give Google and friends all the relevant details on all the users for cheap? Each of them will sell demographics (grouped), or detailed info on a particular user, but good luck buying the "everything" for a low price.
So it costs more to combine the baskets.
You don't use your real name on Slashdot. You don't have to use your real name on Yahoo or Hotmail.You don't have to use your real birthday, timezone, country, etc either.
If you just went "Google" (or Yahoo) for everything (mail, plus etc) they would have everything nice and neat, no need to do the extra work, no need to pay competitors etc for every bit of data. So even if you lied about who you are to Google, they would have everything on "user #11231323". So if one day someone says "hey tell me everything you know about this youtube user" and gives Google some money, google can sell everything they have about you.
In contrast if you split everything up, they'd have to go to Yahoo, Facebook, Microsoft etc one by one and say "tell me everything you know about this youtube user". And I bet they'd be charged a lot more, and might even be given a lot less.
You get more bang for your buck with read caching (more activities read from disk than write to it)
Even if you have 10 times more reads than writes, if your reads can be done at 50000/sec but your writes can only be done at 200-1000/sec (1-5 ms access time), then your writes remain a noticeable bottleneck.
And If you're still using XP you might be interested to read this: http://technet.microsoft.com/en-us/library/cc959914.aspx
For every scan of a directory (even cached) XP by default will WRITE to the drive just to update the last access timestamps!
Linux had that problem too but see noatime and relatime.
KDE and GNOME suffer from the "Microsoft Windows disease".
What they should do is just come up with something that works very like Windows XP (ala ReactOS). So that when Microsoft finally kills XP (in 2014?), they suddenly get half the XP market share.
SSD on laptop.
1Gbps LAN
RAID10 NAS (e.g. 4 x 2TB SATA).
Linux does RAID10 NAS pretty well, if it's always on, you do get full 1Gbps from stuff cached in RAM.
Copy the stuff you need to the laptop before your plane trip or whatever.
If it were up to me I'd cache the stuff that would make the disk head seek the most, and not cache sequential access (unless it happens often enough).
You don't cache stuff that goes: seek 1 millisecond, then sequential read for 10 seconds.You cache stuff that goes seek 1 millisecond, sequential read 5 milliseconds seek 1 millisecond, sequential read 10 milliseconds.
Cache the small stuff, and cache some filesystem metadata at higher priority. 20 years ago I wrote disk caching software to do similar stuff for the Apple IIGS. I think the hybrid drives better do something more intelligent than mere read caching, or most people aren't going to bother.
With all the genius programmers they have, they should use an algorithm that understands _time_ so if the drive is spending lots of time waiting to write/read something to/from the disk, then it should try to cache it. Maybe have a drive mood or something - so if the drive is waiting more, it mood changes to "cache more current stuff", otherwise it goes to "cache less of current stuff".
Seems to me the current hybrid drives don't do write-caching, they only do read-caching.
I can see why read-caching would be a lot simpler to implement, but I bet decent write-caching will really make hybrid drives as fast as SSDs for most desktop use.
Copying files from one location to another at SSD speeds till the write-cache fills up. Then while you do something else the drive flushes the cache to disk.
I'd thought they would move stuff around when that happens ( and that's why performance goes down if you don't have trim and the drive doesn't know which blocks can be overwritten without moving them).
It just means the people actually using that data, marketers, FBI, underpants gnomes, have to buy it from more than one source and then merge it.
I already said that.
And what are you going to do, run your own search engine? Run your own mail system for throwaway accounts? Run your own "facebook" with nobody else but you on it? Run your own "Amazon", "newegg", "ebay" with nobody buying or selling stuff on them?
So use them but don't put all your eggs in one basket, or go live in the wilderness (even if you don't use the Internet at all, the hypermarkets etc are also gathering data on your buying habits - Target figured out a girl was pregnant before her dad found out: http://www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/ ).
From what I see nearly all data-loss failures in SSDs are due to bugs or faults and not due to wear.
The banks are just shifting more and more risks and responsibilities for losses to their customers.
They prefer to call stuff ID Theft rather than some sort of fraud. Since with ID Theft it's their customer's problem, whereas with fraud it might be their problem.
They also prefer debit cards. With credit cards, when "stuff happens", it's not my money that's gone, it's someone else's. They may try to get the money from me, but meanwhile I have my money. Whereas with debit cards, when "stuff happens" it's my money that's gone. In theory I might legally be entitled to get it back, but meanwhile I have lost my money.
Of course they are, but if Yahoo has part of the data and google has the some of it, facebook/etc the rest, they are less likely to share the data with each other. Better than Yahoo or Google having 100% of the data.
Make the various parties work and pay more for it.
When can we stop eating corpses?
Over your dead body?
If ReactOS actually becomes stable and usable I suspect a number of large US corporations might be interested too.
There are lots of corporations that are happy with Windows XP, and the only problems with it is the bugs are no longer going to be fixed, and MS will stop selling it.
No I meant the cases where lots of trades got rolled back/cancelled.
I simply refuse to accept any argument that the markets are distinguishable from casino gambling (other than having an exceptionally toothless gambling commission).
Here's the difference:
In the financial world when you gamble with other people's money and lose big:
a) you get a bailout and get to keep your bonuses and previous cut of the winnings.
or
b) your transactions get rolled back (see the HFT rollbacks).
In casinos when you gamble with other people's money and lose big, I don't think the outcomes are as rosy.
It's not mere toothlessness when they rollback HFT trades because of bugs in favoured _players_ systems (it's fine to rollback if it's due to bugs in the "casino" systems).
Who apparently weren't too happy about it:
http://traveltips.usatoday.com/history-maine-lobster-21560.html
During colonial times, lobster was food for the poverty stricken, prisoners and indentured servants. In the Massachusetts colony that encompassed the land that became known as Maine, indentured servants protested and had instructions written in to their seven-year contracts that they would not be forced to eat lobster more than three times a week.
Thing that concerns me would be allergies.
Far fewer people are allergic to fish, chicken, beef than they are to shrimp, crab, lobsters. Or even dust mites. So I wouldn't be surprised if many are also allergic to these "popular" arthropods.
http://www.hollowtop.com/finl_html/allergies.htm
It's pointless too. From the summary:
Intel needs help breaking into the smartphone market and Nokia isn't tied as tightly to Qualcomm/ARM hardware as other vendors.
From what I see Nokia has no idea on how to break into the smartphone market. So why would buying them help Intel?
Just poach the few talented people they need from Nokia. Why pay lots of money to get saddled with all the crap and baggage that got Nokia in deep shit?
Nokia's patents are not what get you into the smartphone market. Otherwise Nokia would be in the smartphone market.
They could do an "On Live" sort of game service too, but with Linux as an option for the platform.
For people with weaker clients but high bandwidth low latency network connections it might be even faster - level load times can actually be shorter, frame rates could be higher.
If they are playing against other players on the same "cloud", the ping times won't really be that different. There's higher display lag - but that doesn't matter as much for some games/people.
On non-ARM systems, it is required to implement the ability to disable Secure Boot via firmware setup.
So despite Intel's efforts on getting power consumption down I guess there's not going to be a Windows phone or tablet on x86... ;)
A Windows Server may also disable Secure Boot remotely using a strongly authenticated (preferably public-key based) out-of-band management connection, such as to a baseboard management controller or service processor.
I'm sure some hackers will be interested in investigations the various implementations of this feature.