As long as you add an additional few seconds to update a database somewhere to the new name.
My company used to name machines for the primary user and come up with a department/building/location code for the rest. We do imaging of the machines, so naming the machine is pretty easy.
Most of our machines are Lenovos, though, and they come with a brief 6 or 7-character service tag number. They finally decided to save time and hassle and just name every machine after its service tag number about the time I started working for the company. Apparently the whole username thing was starting to fall apart at the seams what with everyone swapping machines around, exchanging offices and demanding that their personal computer be moved rather than just using the identical machine that was sitting there already, etc etc...
We do a lot of remote desktop here, so memorizing my service tag number to remote in to my machine was a major pain. But I eventually got it memorized just in time for them to issue me a laptop and make it irrelevant. However, I've finally got THAT one memorized and life is OK.
I'd prefer username, but I know when I call them with the service tag number they'll know exactly which machine I'm talking about and be able to do something. Anecdotes of remote wiping and reimaging the wrong machine have convinced me that something as clean and simple as using a number that is permanently attached to the physical machine = beauty.
You know, I'm a Comcast customer, and I use BitTorrent, and I agree with you - traffic shaping is a valid practice. Because I also use Vonage, and I understand that some packets are more important than others.
What I find invalid about what Comcast did was they actually shut down my BitTorrent connections completely. I would have been perfectly OK with them throttling the connections, though I'm a little miffed because they ALREADY throttle my upstream connection at 128kbps anyway, which is the fastest upload speed available where I am located. But I'd be OK with them throttling BitTorrent at 56kbps (modem speeds) or even slower when demand is high.
But they were simply disconnecting any connections made if deep-packet inspection revealed that the packets were BitTorrent. From my point of view, I saw connections being made, then going away.
So definition of terms is important. "Traffic shaping" good, "complete denial of entire classes of service" bad.
I imagine someone who implements this scheme would make sure the data has already been offloaded BEFORE peak demand.
There are a few points here:
(1) A data center broadcasting its data out to an alternate probably consumes little or no extra power than one that is just serving up web pages, other than the cost of transmission. All of the hard drives and fans are still spinning, you've just got a small incremental cost for the extra network traffic.
(2) Most of this can piggyback on already-existing algorithms that copy data from one data center to another for continuous operation and caching anyway. I don't see a great deal of increase in transmission based on this method - you do need extra data centers to cover for the ones that are shut down, but the net savings in "peak power" would more than make up for the extra use of "non-peak power".
In general, running a power grid is most efficient with a given range of output or load, the load it's designed for. If your power needs "peak and valley", you need to design your entire grid for the highest peak, which means your grid is running very inefficiently during the lowest valley because you have power sources that are running outside their optimum efficiency.
What this means is that the areas where these data centers are located can delay grid expansion projects (finding new power sources). These data centers have to be located SOMEWHERE, so building a half-dozen extras might mean that a few dozen cities where there are already data centers might be able to delay expanding their power grids, because demand at peak has been reduced.
"The team then devised a routing scheme designed to take advantage of daily and hourly fluctuations in electricity costs across the country. The resulting algorithm weighs up the physical distance needed to route information--because it's more expensive to move data further--against the likely cost savings from reduced energy use. Data collected from nine Akamai servers, covering 24 days of activity, provided a way to test the routing scheme using real-world data."
Your packets will not get routed through some router in Iceland because it's cheaper to buy (relatively expensive) electricity somewhere in the US than it is to pay for data transmission from overseas. The intent here is to put the data in a "relatively close" location where power is cheapest, balancing the cost of sending the data against the cost of carrying it locally. Even if you were PAID to consume electricity in Iceland, the transmission costs would gobble up any possible electrical savings.
Not to mention that there are some sacrifices that a company like Google would consider worth taking to save money. Reduced performance is not one of them. A mass exodus of ad-revenue-generating eyeballs just isn't within their business plan.
Had to re-read that. For a second, I thought you were referring to the Illudium Q-36 Explosive Space Modulator, and figured the US Government simply got the contract to supply them. And there would be a demand, since he seems to keep losing them all the time....
Do they have to be retained articles, or do deleted ones count? Because if you really wanted to, you could write a Wikipedia article about the number 3294199 and simply submit it 294199 times. Then your number is doubly significant.
Quick, get to Google Voice and see if you can get that number in your area code! Now THAT'S geek cred.
First, the clothes don't appear to be "emitting" anything. I don't speak Russian, so I can't get the technical details, but it appears that they are using some form of micromesh that shields the user from some of the radiation from a cell phone. The clothing itself does not emit any radiation, it's a passive block.
As to fixing the radiation, I can't say as I disagree, though I've heard many different stories on the results of studies, but better safe than sorry, I suppose. Does lead to a slight problem. I'd admittedly be hard-pressed to think of another thing other than radio waves to, well, carry radio signals. And I have yet to see an alternative to radio for wireless communication. I mean, if we all suddenly converted to smoke signals we'd start dying off of Black Lung or something.
Having said all that, I don't honestly see why a child should need a powered-on cell phone in a classroom setting. If the child needs a cell for some reason, or their parents want them to carry one, fine. But the best way to protect someone from the radiation the phones emit is to turn the damned things off until they are needed. If the parents need to reach the children, the appropriate way is to (a) leave a voicemail ot text and the kid can check during lunch or break, or if it's urgent (b) contact the school who will send someone to the classroom to give the message to the kid.
Yes, that's quite possible. I'm also sure that their measurement floor of.006 micrograms has a lot to do with it - I'm relatively certain that such amounts were all but undetectable in the 1980s for example.
And, for the record, I'm not (as the article is) suggesting that "contamination" = "use". The article is making a ridiculous assumption, on that we certainly agree.
I think the explanation is far simpler. Population increases (money changing hands faster), increases in detection equipment so we can detect increasingly tiny contamination, more machines handling money so the machines can get contaminated and spread the contamination further, etc etc etc.
Douglas Adams was right. Eliminating phone sanitizers is a really bad idea. Recall the "B" Ark!
I'd HOPE the Snopes article would be more informative and detailed.
Snopes = exhaustively researched articles done by people who take the time to get it right as much as they possibly can, with no need to be "first to post" to get credit.
Slashdot = geeks finding interesting articles and rushing to get them posted so they don't get deleted as duplicates. If a/. user waits until they find a good-quality well-researched article, it'll get smacked down as a duplicate because 500 slashdotters would have posted the crappy "breaking news" version hoping to be first.:)
Ummm... the article isn't introducing this as a new issue. Even the article summary makes that clear. The article is about the increases in contamination.
The point of the article is that the rate of contamination is increasing. The Snopes article makes mention of 1985 being 33 to 50% contamination rates. The article summary refers to a 2-year-old study that puts contamination closer to 67%, and now the most recent study puts it at around 90%. So the point is that the contamination has increased.
Nav4All is also a good turn-by-turn app for the BlackBerry, and while it isn't nearly as nice as TeleNav, it does have the advantage of being free.
TeleNav would be my choice if I traveled a lot and really needed turn-by-turn often, of course, but I might use turn-by-turn about 3-4 times a year, and $10 a month is just too rich for those odd occasions when I do.
Think about it. This is someone who is using SatNav on a handheld. Receiving a call is simply replacing one distraction with another.
Actually, dropping SatNav when the phone rings is probably a brilliant move. Imagine someone driving in traffic trying to hold a phone conversation AND watch their SatNav. It leaves absolutely no hands free to eat a burger.
Anyway, you're right. You really should park somewhere so you can have your conversation without being distracted by pesky pedestrians bouncing off your hood. Let me tell you, if those buggers survive their screams can really drown out a conversation. Especially the younger ones. I've been in the middle of closing a tough deal, been about to say "Yes", and had some asshat I'm about to hit yell "NO!" and screw up the whole deal, tens of thousands of dollars. It's annoying.
As a couple of the other posts have implied, a Google data center would likely get more expensive when demand is at peak.
Most commercial contracts have a provision for variable rates based on demand, and if they don't these sorts of data centers would be an ideal situation to introduce them. Google, Akamai, whomever could simply come into town and say "we want to build a building somewhere near a local Internet hub". The building would be fed plenty of power when demand is low, and that power could be sold to them very cheaply since demand is low at the time. When demand goes up, this building starts backing down its demand and freeing up capacity when the system needs it the most.
Given that most powerplants run most efficiently at a constant load, this could be a great way for the power company to offload "extra" energy during non-peak times but free up that capacity when a heat wave hits and everyone decides to turn their AC on.
Something like this actually makes it LESS likely that your power would go out during heavy load, simply because if it does the data center would probably be charged ruinous electric fees if they continued operating. So they'll shut down so you can turn your A/C up to "11" if you want.
Sure, Google for one will no doubt be happy to pass along the savings to the majority of their customers. I'm sure they'll pass along the entire 40%.
I'm a pretty typical Google customer, so let's add up the savings. Let's see, I use Google Voice, GMail, Google News, Google Maps / Google Earth, Google Documents, occasionally YouTube and probably a small handful of other services less often. My total bill comes to a whopping $0. $0 minus 40% is $0. See? They've passed along the entire savings.
Heck, maybe they'll be generous and give me a Googol % discount!
As far as the performance hit, I doubt it. It's already not easy to differentiate between data coming from a server in the US versus one overseas. If a server farm in the US gets shut down because of heat or high electric costs, I'd be willing to bet that for all practical purposes the switchover would be transparent to you, or nearly so.
Then you accept you've made a mistake, pay your $5-15, and move on. Stop buying lattes at StarDucks for a week and pocket the difference.
The instant a "money back" UNDO button goes into place, someone will just go back to tasting again. A well-intentioned system to allow people to avoid the consequences of their mistakes has allowed significant abuses of the system for the rest of us.
And by "rest of us", I will freely admit that I have acquired a couple of incorrect domain names in the past. Recently I got one for an informal camping/hiking/kayaking group I'm in, and one of the members said she didn't like the domain name I chose a couple of days later. So I just got a new domain name. My registrar charges $6 for the first year and $8 a year afterward, so it literally wasn't worth the time and effort of getting my money back for the mistake. I'll just let the domain lapse next year if I haven't found a use for it.
I think a lot of the need for an undo comes about simply because it exists. If people knew that a domain order was irrevocable, they'd probably spend an extra 5 seconds closely examining their domain name before hitting YES.
It could. As I posted previously, I experimented in college by refraining from sleep for about 5 days. I found that concentration on something helped me feel less tired, but that the effect lessened over time. I suspect that concentration was (in a way) replacing the dream state for me.
I also found that I felt less tired if I engaged in a brief period of random drawing - just taking some colored pencils or something and drawing shapes and scribbles at random, so that might have helped work out some of the daily issues.
But, at the end of 5 days and nights, I was pretty much done for, and my dreams were very vivid. So I expect there was some "pent up demand" at Brain Theater.:)
A lot of sleep is not spent dreaming, though. It would be interesting to see studies on how much REM sleep people get as their sleep decreases. If you get the same amount of time dreaming and/or in REM state, the loss of "less purposeful" sleep might not be significant.
Woulda been even funnier if you had made the same mistake as a lot of people did, and wondered how a sheep could be made shorter, then worrying about how that same effect would be applied to fruit.
Slashdot Mirror
As long as you add an additional few seconds to update a database somewhere to the new name.
My company used to name machines for the primary user and come up with a department/building/location code for the rest. We do imaging of the machines, so naming the machine is pretty easy.
Most of our machines are Lenovos, though, and they come with a brief 6 or 7-character service tag number. They finally decided to save time and hassle and just name every machine after its service tag number about the time I started working for the company. Apparently the whole username thing was starting to fall apart at the seams what with everyone swapping machines around, exchanging offices and demanding that their personal computer be moved rather than just using the identical machine that was sitting there already, etc etc...
We do a lot of remote desktop here, so memorizing my service tag number to remote in to my machine was a major pain. But I eventually got it memorized just in time for them to issue me a laptop and make it irrelevant. However, I've finally got THAT one memorized and life is OK.
I'd prefer username, but I know when I call them with the service tag number they'll know exactly which machine I'm talking about and be able to do something. Anecdotes of remote wiping and reimaging the wrong machine have convinced me that something as clean and simple as using a number that is permanently attached to the physical machine = beauty.
Clarification: True things that are this sad can't be funny. Oh, wait, maybe we need a "Gallows Humor" mod.
You know, I'm a Comcast customer, and I use BitTorrent, and I agree with you - traffic shaping is a valid practice. Because I also use Vonage, and I understand that some packets are more important than others.
What I find invalid about what Comcast did was they actually shut down my BitTorrent connections completely. I would have been perfectly OK with them throttling the connections, though I'm a little miffed because they ALREADY throttle my upstream connection at 128kbps anyway, which is the fastest upload speed available where I am located. But I'd be OK with them throttling BitTorrent at 56kbps (modem speeds) or even slower when demand is high.
But they were simply disconnecting any connections made if deep-packet inspection revealed that the packets were BitTorrent. From my point of view, I saw connections being made, then going away.
So definition of terms is important. "Traffic shaping" good, "complete denial of entire classes of service" bad.
If I had mod points, it would be a hard choice how to moderate your post.
"Funny", no, it's really not. True things can't be funny.
"Insightful" is probably closest.
What I really want is "True, but depressing."
Yes, I know, the correlation between contamination and use in the original article is unfounded. Agreed.
I imagine someone who implements this scheme would make sure the data has already been offloaded BEFORE peak demand.
There are a few points here:
(1) A data center broadcasting its data out to an alternate probably consumes little or no extra power than one that is just serving up web pages, other than the cost of transmission. All of the hard drives and fans are still spinning, you've just got a small incremental cost for the extra network traffic.
(2) Most of this can piggyback on already-existing algorithms that copy data from one data center to another for continuous operation and caching anyway. I don't see a great deal of increase in transmission based on this method - you do need extra data centers to cover for the ones that are shut down, but the net savings in "peak power" would more than make up for the extra use of "non-peak power".
In general, running a power grid is most efficient with a given range of output or load, the load it's designed for. If your power needs "peak and valley", you need to design your entire grid for the highest peak, which means your grid is running very inefficiently during the lowest valley because you have power sources that are running outside their optimum efficiency.
What this means is that the areas where these data centers are located can delay grid expansion projects (finding new power sources). These data centers have to be located SOMEWHERE, so building a half-dozen extras might mean that a few dozen cities where there are already data centers might be able to delay expanding their power grids, because demand at peak has been reduced.
From the article:
"The team then devised a routing scheme designed to take advantage of daily and hourly fluctuations in electricity costs across the country. The resulting algorithm weighs up the physical distance needed to route information--because it's more expensive to move data further--against the likely cost savings from reduced energy use. Data collected from nine Akamai servers, covering 24 days of activity, provided a way to test the routing scheme using real-world data."
Your packets will not get routed through some router in Iceland because it's cheaper to buy (relatively expensive) electricity somewhere in the US than it is to pay for data transmission from overseas. The intent here is to put the data in a "relatively close" location where power is cheapest, balancing the cost of sending the data against the cost of carrying it locally. Even if you were PAID to consume electricity in Iceland, the transmission costs would gobble up any possible electrical savings.
Not to mention that there are some sacrifices that a company like Google would consider worth taking to save money. Reduced performance is not one of them. A mass exodus of ad-revenue-generating eyeballs just isn't within their business plan.
OK, mea culpa. Would it help if I added "compared to your average Slashdot article"? :)
Had to re-read that. For a second, I thought you were referring to the Illudium Q-36 Explosive Space Modulator, and figured the US Government simply got the contract to supply them. And there would be a demand, since he seems to keep losing them all the time....
And this way you've ALWAYS got nukes on standby. It's a win-win.
Do they have to be retained articles, or do deleted ones count? Because if you really wanted to, you could write a Wikipedia article about the number 3294199 and simply submit it 294199 times. Then your number is doubly significant.
Quick, get to Google Voice and see if you can get that number in your area code! Now THAT'S geek cred.
First, the clothes don't appear to be "emitting" anything. I don't speak Russian, so I can't get the technical details, but it appears that they are using some form of micromesh that shields the user from some of the radiation from a cell phone. The clothing itself does not emit any radiation, it's a passive block.
As to fixing the radiation, I can't say as I disagree, though I've heard many different stories on the results of studies, but better safe than sorry, I suppose. Does lead to a slight problem. I'd admittedly be hard-pressed to think of another thing other than radio waves to, well, carry radio signals. And I have yet to see an alternative to radio for wireless communication. I mean, if we all suddenly converted to smoke signals we'd start dying off of Black Lung or something.
Having said all that, I don't honestly see why a child should need a powered-on cell phone in a classroom setting. If the child needs a cell for some reason, or their parents want them to carry one, fine. But the best way to protect someone from the radiation the phones emit is to turn the damned things off until they are needed. If the parents need to reach the children, the appropriate way is to (a) leave a voicemail ot text and the kid can check during lunch or break, or if it's urgent (b) contact the school who will send someone to the classroom to give the message to the kid.
That certainly explains why the machines are able to run so fast...
Sorry. I hope I'm not contributing to increasing percentages of currency contamination with all the /. fun I'm killing. :)
Yes, that's quite possible. I'm also sure that their measurement floor of .006 micrograms has a lot to do with it - I'm relatively certain that such amounts were all but undetectable in the 1980s for example.
And, for the record, I'm not (as the article is) suggesting that "contamination" = "use". The article is making a ridiculous assumption, on that we certainly agree.
I think the explanation is far simpler. Population increases (money changing hands faster), increases in detection equipment so we can detect increasingly tiny contamination, more machines handling money so the machines can get contaminated and spread the contamination further, etc etc etc.
Douglas Adams was right. Eliminating phone sanitizers is a really bad idea. Recall the "B" Ark!
I'd HOPE the Snopes article would be more informative and detailed.
Snopes = exhaustively researched articles done by people who take the time to get it right as much as they possibly can, with no need to be "first to post" to get credit.
Slashdot = geeks finding interesting articles and rushing to get them posted so they don't get deleted as duplicates. If a /. user waits until they find a good-quality well-researched article, it'll get smacked down as a duplicate because 500 slashdotters would have posted the crappy "breaking news" version hoping to be first. :)
Ummm... the article isn't introducing this as a new issue. Even the article summary makes that clear. The article is about the increases in contamination.
The point of the article is that the rate of contamination is increasing. The Snopes article makes mention of 1985 being 33 to 50% contamination rates. The article summary refers to a 2-year-old study that puts contamination closer to 67%, and now the most recent study puts it at around 90%. So the point is that the contamination has increased.
Nav4All is also a good turn-by-turn app for the BlackBerry, and while it isn't nearly as nice as TeleNav, it does have the advantage of being free.
TeleNav would be my choice if I traveled a lot and really needed turn-by-turn often, of course, but I might use turn-by-turn about 3-4 times a year, and $10 a month is just too rich for those odd occasions when I do.
Think about it. This is someone who is using SatNav on a handheld. Receiving a call is simply replacing one distraction with another.
Actually, dropping SatNav when the phone rings is probably a brilliant move. Imagine someone driving in traffic trying to hold a phone conversation AND watch their SatNav. It leaves absolutely no hands free to eat a burger.
Anyway, you're right. You really should park somewhere so you can have your conversation without being distracted by pesky pedestrians bouncing off your hood. Let me tell you, if those buggers survive their screams can really drown out a conversation. Especially the younger ones. I've been in the middle of closing a tough deal, been about to say "Yes", and had some asshat I'm about to hit yell "NO!" and screw up the whole deal, tens of thousands of dollars. It's annoying.
As a couple of the other posts have implied, a Google data center would likely get more expensive when demand is at peak.
Most commercial contracts have a provision for variable rates based on demand, and if they don't these sorts of data centers would be an ideal situation to introduce them. Google, Akamai, whomever could simply come into town and say "we want to build a building somewhere near a local Internet hub". The building would be fed plenty of power when demand is low, and that power could be sold to them very cheaply since demand is low at the time. When demand goes up, this building starts backing down its demand and freeing up capacity when the system needs it the most.
Given that most powerplants run most efficiently at a constant load, this could be a great way for the power company to offload "extra" energy during non-peak times but free up that capacity when a heat wave hits and everyone decides to turn their AC on.
Something like this actually makes it LESS likely that your power would go out during heavy load, simply because if it does the data center would probably be charged ruinous electric fees if they continued operating. So they'll shut down so you can turn your A/C up to "11" if you want.
Sure, Google for one will no doubt be happy to pass along the savings to the majority of their customers. I'm sure they'll pass along the entire 40%.
I'm a pretty typical Google customer, so let's add up the savings. Let's see, I use Google Voice, GMail, Google News, Google Maps / Google Earth, Google Documents, occasionally YouTube and probably a small handful of other services less often. My total bill comes to a whopping $0. $0 minus 40% is $0. See? They've passed along the entire savings.
Heck, maybe they'll be generous and give me a Googol % discount!
As far as the performance hit, I doubt it. It's already not easy to differentiate between data coming from a server in the US versus one overseas. If a server farm in the US gets shut down because of heat or high electric costs, I'd be willing to bet that for all practical purposes the switchover would be transparent to you, or nearly so.
Then you accept you've made a mistake, pay your $5-15, and move on. Stop buying lattes at StarDucks for a week and pocket the difference.
The instant a "money back" UNDO button goes into place, someone will just go back to tasting again. A well-intentioned system to allow people to avoid the consequences of their mistakes has allowed significant abuses of the system for the rest of us.
And by "rest of us", I will freely admit that I have acquired a couple of incorrect domain names in the past. Recently I got one for an informal camping/hiking/kayaking group I'm in, and one of the members said she didn't like the domain name I chose a couple of days later. So I just got a new domain name. My registrar charges $6 for the first year and $8 a year afterward, so it literally wasn't worth the time and effort of getting my money back for the mistake. I'll just let the domain lapse next year if I haven't found a use for it.
I think a lot of the need for an undo comes about simply because it exists. If people knew that a domain order was irrevocable, they'd probably spend an extra 5 seconds closely examining their domain name before hitting YES.
It could. As I posted previously, I experimented in college by refraining from sleep for about 5 days. I found that concentration on something helped me feel less tired, but that the effect lessened over time. I suspect that concentration was (in a way) replacing the dream state for me.
I also found that I felt less tired if I engaged in a brief period of random drawing - just taking some colored pencils or something and drawing shapes and scribbles at random, so that might have helped work out some of the daily issues.
But, at the end of 5 days and nights, I was pretty much done for, and my dreams were very vivid. So I expect there was some "pent up demand" at Brain Theater. :)
A lot of sleep is not spent dreaming, though. It would be interesting to see studies on how much REM sleep people get as their sleep decreases. If you get the same amount of time dreaming and/or in REM state, the loss of "less purposeful" sleep might not be significant.
Not true. I am a parent, and I sleep like a baby. Every 4 hours, I wake up crying.
Woulda been even funnier if you had made the same mistake as a lot of people did, and wondered how a sheep could be made shorter, then worrying about how that same effect would be applied to fruit.