Turn in your geek card. A short will develope and the cord will get hot and degrade the heat shrink and the now molten hot solder will work it's way to something vital such as the expensive rackmount server that some schmuck put under it. *banghead* Seems pretty damn unlikely to me, shorts tend to produce a very large spike of current followed by a tripped breaker or blown fuse
In my experiance ordinary scissors won't cut the thicker plastic clamshells. A good pair of side cutters will but they have an annoyingly short cut length. Trauma shears are good for opening them but most people don't keep those handy.
So people end up using sharp open blades with all the potential dangers those bring.
And then there are the sharp edges of the packaging itself, not enough to do any serious damage but certainly painfull.
But I think you're still learning the wrong lesson. Buying from two non-tier 1 providers ensures you don't have a direct hop to anyone important If you are multihomed there is nothing wrong with using tier 1 providers. Indeed it is possiblly one of the easiest ways to ensure you have two different tier 1 providers as eventual upstreams.
On the other hand if you are NOT multihomed and want reliability using a tier 1 is risky and using a wanabe tier 1 that likes to get into peering disputes (cogent) will practically gaurantee problems.
Buying transit from two major national "near-transit-free" providers is probably your safest option. Agreed with the provisio that you should still find out who thier transit providers are and make sure that between them they have transit to at least two tier 1 providers.
You consider it ridiculous to do some research to ensure your backup provider and main provider have between them upstreams to at least two different tier 1 providers? I would consider that basic due dilligence.
Of course the university are being even more stupid for picking cogent as thier sole upstream. This isn't the first time cogent have got into this sort of dispute and i'm 99% sure it won't be the last.
I'm not trying to put more blame on the victim I just disagree with you about the moral of the story, The moral IMO is not "don't multihome", it's "do your homework when picking providers for anything important". For internet providers doing your homework means finding out the route up the pyrimid and making sure it ends on at least two tier 1 providers (a collary of this is that single homing directly with a tier 1 provider is a bad idea).
End users and small buisnesses are as you say caught between a rock and a hard place in many cases since often the only decent broadband provider in the localitity is a tier 1 . That makes it even more important for anyone hosting anything important to do thier end properly or risk having thier users (who are probablly also customers) cut off.
Afaict ISPs tend to use hot potato routing for sending data to peers. that is when they need to send data to a peer and have multiple interconnection to that peer (at this level they will have multiple interconnections) they will do so at the interconnect closest to where the data came from.
This is generally a good strategy because it avoids stupid routing like sending the data over the atlantic twice once in each network but it does mean that the recipiant of traffic bears the majority of the cost of transfering it.
I presume this is the reason for the balanced transfer requirements in the agreements but it undoubtablly hurts content heavy networks like cogent.
Cogent is a Tier 1 network service provider (weather or not Sprint and L3 want it to be). That depends on your definition of tier 1. They don't buy transit but they do buy paid peering. It seems in this case they failed at a free peering trial and refused to pay for peering so eventually sprint cut them off.
The idea that Sprint doesn't get as much out of peering with Cogent as Cogent does peering with Sprint is absurd I'm not so sure about that.
Afaict when ISPs are connected at multiple places the sender will typically send the traffic out of the closest place to where it came from (since they don't know the geographic location of where it is going to). That means the recipiant bears the majority of the cost of moving the data. Afaict cogent is a hosting heavy network so they generate far more traffic than they recieve.
but i'm sure to some extent you are right and this is a power play, being a tier 1 is hugely profitable and companies like sprint really don't want a cheap competitor coming along and ruining things for them.
the problem is that the internet is hugely commercial nowadays, companies won't just allow traffic to take any old route only routes that someone is paying for.
In other works ISPs will route between two customers and they will route between a customer and a peer or a customer and thier upstream transit provider (if they have any) but they will not route between two peers or between a peer and their upstream transit.
Neither cogent nor sprint have any upstream transit providers so when they depeered each other single homed sprint customers (e.g end users on one of sprints services) wont be able to reach single homed cogent customers (e.g. a lot of bandwidth heavy services)
It then becomes analgous to employees going on strike. The depeering is bad for both cogent and sprint but whoever blinks first is going to get the worse end of the new peering deal.
For example, I know one site that was multi-homed. They had Sprint and a regional provider. The regional provider was de-peered by Cogent about a year ago, and the regional provider only buys transit from Sprint There is your problem, if you are going to multihome you need to research who your providers buy transit from and make sure as things move up the tree from both providers they end up at different tier 1 providers.
For end users things are harder, ideally the sites they connect to should be multihomed so it shouldn't be such an issue but where it is thier best option is probablly to use some sort of tunnel service bought from a third party until things clear up.
or a larger screen resolution. Unfortunately it doesn't, the 1000 series models have exactly the same screen resoloution as the 900 series as do most of thier competitors machines.
The only netbook i'm aware of with a screen more than 1024x600 is the HP 2133 mininote and that has an 8.9 inch screen! 1280x768 on a 8.9 inch is going too far in the tiny pixels direction for my liking.
Is anyone aware of a netbook with a 10 inch 1280x768 screen?
The success of these items helped them to bargain for a better price on Windows. More importantly IMO it allowed them to bargin for continued availability of XP.
Shipping netbooks with vista would have been the kiss of death for the whole netbook concept. Vista (at least prior to SP1, I hear things have improved significantly with SP1 but I don't use vista anymore so I can't say) is considerablly slower than XP on the hardware that was being put in netbooks and takes up so much disk space that the smaller netbooks would have great trouble accomodating it.
Shipping the first model with linux and instructions to install XP yourself then adding windows XP models to the lineup gives a clear threat of "let us ship windows XP or we will ship linux and leave the customers who want windows to pirate it themselves".
terminal velocity does not directly depend on density, it primerally depends on mass and the surface area presented to the air.
If all ratios, materials and postitions stay the same then mass will go as the cube of the dimensions while surface area presented to the air will only go as the square of the dimesions.
The problem is that illegal copies don't have DRM so making an illegal copy of an illegal copy is trivial. It's all digital so there is no generation loss.
Add that to the fact that there is this huge distribution network called the internet and one illegal copy can become a huge number in a relatively short space of time.
I suspect if they have any sense they would do a mixture, free peering with any local ISPs, paid peering to the major ISPs and then transit for whatever is left.
well a peering arrangement would only give them cheap bandwidth to sprints customers not to everyone else.
but anyway cogent is a wannabe tier 1. They don't buy transit but they do make use of paid peering arrangements. Add that to thier low prices which piss off the entrenched tier 1 providers and they are going to get into disputes over what if anything they should be paying for thier peering and if they de-peer thier single homed customers lose connectivity to whoever they de-peer with.
The bottom line is if you care about reliability don't be a single homed cogent customer.
I thought they were allowed to do it but if they did and the transaction was deemed fraudulant the retailer had to take the entire cost.
I have fairly recently had a card fail to work on chip and pin (chip and pin isn't very reliable in my experiance) and after a couple of attempts the retailer (it was one of the big supermarkets) put the transaction through as swipe and sign.
And then there are the times when a stores computer system goes down and they have to do card transactions the REALLY old fasioned way.
They are even getting less restrictive now, with no region locking on PS3 games Is that an official sony position or just the experiance of gamers?
I have noticed that at least the sony games I have bought have had a region marked on the box though the one game I bought which is not from my region did work.
What i'm saying is barring an official statement to the contary my suspicion is that region locking is optional and not being used right now to save on manufacturing costs (setting up for a blu-ray run isn't cheap afaict)
basically, the telescope must be a LOT cheaper than an equivalent satellite telescope. And it wouldn't surprise me if it was for one simple reason: reliability.
If you are going to put something in space it has to be extremely reliable since the only way to service it is with a very expensive shuttle mission (if indeed you can service it at all). That means you spend a huge ammount of time and money checking, double checking, triple checking and so on. Of course backup systems are an option but over time if the main fails the backup may well do so as well and backup systems add weight which is always at a premium on spacecraft.
On the other hand if it's mounted on a plane and something fails it's no big deal, switch to the backup and replace the primary when you are next in for servicing.
Which is impossible to achive since many of those addresses are not advertised anywhere public. Many of them probablly aren't even used anywhere.
Another definition would be anyone who buys transit from one of the widely acknolaged tier 1 ISPs and allows you to communicate with them via that route. Of course that then raises the question of cogents status (afaict they are a wannabe tier 1 who refuse to buy transit but are not respected enough to get free peering from all the tier 1 ISPs but i'm not an expert)
Anyone can print a check with any account and routing numbers they want. While checks are low-tech, and easy to copy, they're also very easy to trace. The fraudster's bank has identifying information for whomever cashed the fakes, which makes prosecution trivial. It seems to me that at least according to the article the problem is that theese numbers can be used internationally.
So your bank knows that your money has been transferred to some "bank" in $THIRD_WORLD_SHITHOLE . Said "bank" either does not have sufficiant evidence to figure out who picked the money up or refuses to cooperate.
And with credit cards, are you talking about making physical fake cards? Because that's not exactly something one can whip up with supplies from the local hardware store Afaict plastic card printers and magstripe writers are easy enough to get, Not a job for your local hardware store but plenty of places use ID cards that are very similar to credit cards so the printers are availible. You would probablly have to rig something up to do the embossing but that can't be terriblly difficult.
It's not a hardware store job but it's not out of reach of a reasonablly organised criminal with a few thousand pounds to spend and a location to get stuff delivered to.
Chip and pin cards are probablly much harder to fake but at least here in the UK most places will still put a transaction through with a swipe and sign if chip and pin fails or the card does not have a chip.
A "full archives" mirror is around 210 gigabytes and is slowly increasing. Archives mirrors are updated at least every 24 hours but no more then every 4 hours.
though you could cut that down quite a lot by using debmirror and only mirroring the releases "sections" and architectures you want
"I don't think that many of you put a photo of their keys online â" with their addresses" But many people do reveal thier place of work or study online. If you have that and a photograph you can simply follow them to find out where they live.
Turn in your geek card. A short will develope and the cord will get hot and degrade the heat shrink and the now molten hot solder will work it's way to something vital such as the expensive rackmount server that some schmuck put under it. *banghead*
Seems pretty damn unlikely to me, shorts tend to produce a very large spike of current followed by a tripped breaker or blown fuse
In my experiance ordinary scissors won't cut the thicker plastic clamshells. A good pair of side cutters will but they have an annoyingly short cut length. Trauma shears are good for opening them but most people don't keep those handy.
So people end up using sharp open blades with all the potential dangers those bring.
And then there are the sharp edges of the packaging itself, not enough to do any serious damage but certainly painfull.
But I think you're still learning the wrong lesson. Buying from two non-tier 1 providers ensures you don't have a direct hop to anyone important
If you are multihomed there is nothing wrong with using tier 1 providers. Indeed it is possiblly one of the easiest ways to ensure you have two different tier 1 providers as eventual upstreams.
On the other hand if you are NOT multihomed and want reliability using a tier 1 is risky and using a wanabe tier 1 that likes to get into peering disputes (cogent) will practically gaurantee problems.
Buying transit from two major national "near-transit-free" providers is probably your safest option.
Agreed with the provisio that you should still find out who thier transit providers are and make sure that between them they have transit to at least two tier 1 providers.
You consider it ridiculous to do some research to ensure your backup provider and main provider have between them upstreams to at least two different tier 1 providers? I would consider that basic due dilligence.
Of course the university are being even more stupid for picking cogent as thier sole upstream. This isn't the first time cogent have got into this sort of dispute and i'm 99% sure it won't be the last.
I'm not trying to put more blame on the victim I just disagree with you about the moral of the story, The moral IMO is not "don't multihome", it's "do your homework when picking providers for anything important". For internet providers doing your homework means finding out the route up the pyrimid and making sure it ends on at least two tier 1 providers (a collary of this is that single homing directly with a tier 1 provider is a bad idea).
End users and small buisnesses are as you say caught between a rock and a hard place in many cases since often the only decent broadband provider in the localitity is a tier 1 . That makes it even more important for anyone hosting anything important to do thier end properly or risk having thier users (who are probablly also customers) cut off.
Is that $4 per day? $4 per month? $4 per year? or what?
Afaict ISPs tend to use hot potato routing for sending data to peers. that is when they need to send data to a peer and have multiple interconnection to that peer (at this level they will have multiple interconnections) they will do so at the interconnect closest to where the data came from.
This is generally a good strategy because it avoids stupid routing like sending the data over the atlantic twice once in each network but it does mean that the recipiant of traffic bears the majority of the cost of transfering it.
I presume this is the reason for the balanced transfer requirements in the agreements but it undoubtablly hurts content heavy networks like cogent.
Cogent is a Tier 1 network service provider (weather or not Sprint and L3 want it to be).
That depends on your definition of tier 1. They don't buy transit but they do buy paid peering. It seems in this case they failed at a free peering trial and refused to pay for peering so eventually sprint cut them off.
The idea that Sprint doesn't get as much out of peering with Cogent as Cogent does peering with Sprint is absurd
I'm not so sure about that.
Afaict when ISPs are connected at multiple places the sender will typically send the traffic out of the closest place to where it came from (since they don't know the geographic location of where it is going to). That means the recipiant bears the majority of the cost of moving the data. Afaict cogent is a hosting heavy network so they generate far more traffic than they recieve.
but i'm sure to some extent you are right and this is a power play, being a tier 1 is hugely profitable and companies like sprint really don't want a cheap competitor coming along and ruining things for them.
$400-1000 for a 100mb connection with no caps etc
$400=$1000 per what?
the problem is that the internet is hugely commercial nowadays, companies won't just allow traffic to take any old route only routes that someone is paying for.
In other works ISPs will route between two customers and they will route between a customer and a peer or a customer and thier upstream transit provider (if they have any) but they will not route between two peers or between a peer and their upstream transit.
Neither cogent nor sprint have any upstream transit providers so when they depeered each other single homed sprint customers (e.g end users on one of sprints services) wont be able to reach single homed cogent customers (e.g. a lot of bandwidth heavy services)
It then becomes analgous to employees going on strike. The depeering is bad for both cogent and sprint but whoever blinks first is going to get the worse end of the new peering deal.
For example, I know one site that was multi-homed. They had Sprint and a regional provider. The regional provider was de-peered by Cogent about a year ago, and the regional provider only buys transit from Sprint
There is your problem, if you are going to multihome you need to research who your providers buy transit from and make sure as things move up the tree from both providers they end up at different tier 1 providers.
For end users things are harder, ideally the sites they connect to should be multihomed so it shouldn't be such an issue but where it is thier best option is probablly to use some sort of tunnel service bought from a third party until things clear up.
or a larger screen resolution.
Unfortunately it doesn't, the 1000 series models have exactly the same screen resoloution as the 900 series as do most of thier competitors machines.
The only netbook i'm aware of with a screen more than 1024x600 is the HP 2133 mininote and that has an 8.9 inch screen! 1280x768 on a 8.9 inch is going too far in the tiny pixels direction for my liking.
Is anyone aware of a netbook with a 10 inch 1280x768 screen?
Umm according to HP the 2133 has a 8.9 inch screen. I'm thinking that at 1280x768 that would be uncomfortablly small.
I really wish asus would do a model in the same case as the 1000 series but with a higher res screen.
The success of these items helped them to bargain for a better price on Windows.
More importantly IMO it allowed them to bargin for continued availability of XP.
Shipping netbooks with vista would have been the kiss of death for the whole netbook concept. Vista (at least prior to SP1, I hear things have improved significantly with SP1 but I don't use vista anymore so I can't say) is considerablly slower than XP on the hardware that was being put in netbooks and takes up so much disk space that the smaller netbooks would have great trouble accomodating it.
Shipping the first model with linux and instructions to install XP yourself then adding windows XP models to the lineup gives a clear threat of "let us ship windows XP or we will ship linux and leave the customers who want windows to pirate it themselves".
terminal velocity does not directly depend on density, it primerally depends on mass and the surface area presented to the air.
If all ratios, materials and postitions stay the same then mass will go as the cube of the dimensions while surface area presented to the air will only go as the square of the dimesions.
The problem is that illegal copies don't have DRM so making an illegal copy of an illegal copy is trivial. It's all digital so there is no generation loss.
Add that to the fact that there is this huge distribution network called the internet and one illegal copy can become a huge number in a relatively short space of time.
I suspect if they have any sense they would do a mixture, free peering with any local ISPs, paid peering to the major ISPs and then transit for whatever is left.
well a peering arrangement would only give them cheap bandwidth to sprints customers not to everyone else.
but anyway cogent is a wannabe tier 1. They don't buy transit but they do make use of paid peering arrangements. Add that to thier low prices which piss off the entrenched tier 1 providers and they are going to get into disputes over what if anything they should be paying for thier peering and if they de-peer thier single homed customers lose connectivity to whoever they de-peer with.
The bottom line is if you care about reliability don't be a single homed cogent customer.
I thought they were allowed to do it but if they did and the transaction was deemed fraudulant the retailer had to take the entire cost.
I have fairly recently had a card fail to work on chip and pin (chip and pin isn't very reliable in my experiance) and after a couple of attempts the retailer (it was one of the big supermarkets) put the transaction through as swipe and sign.
And then there are the times when a stores computer system goes down and they have to do card transactions the REALLY old fasioned way.
They are even getting less restrictive now, with no region locking on PS3 games
Is that an official sony position or just the experiance of gamers?
I have noticed that at least the sony games I have bought have had a region marked on the box though the one game I bought which is not from my region did work.
What i'm saying is barring an official statement to the contary my suspicion is that region locking is optional and not being used right now to save on manufacturing costs (setting up for a blu-ray run isn't cheap afaict)
basically, the telescope must be a LOT cheaper than an equivalent satellite telescope.
And it wouldn't surprise me if it was for one simple reason: reliability.
If you are going to put something in space it has to be extremely reliable since the only way to service it is with a very expensive shuttle mission (if indeed you can service it at all). That means you spend a huge ammount of time and money checking, double checking, triple checking and so on. Of course backup systems are an option but over time if the main fails the backup may well do so as well and backup systems add weight which is always at a premium on spacecraft.
On the other hand if it's mounted on a plane and something fails it's no big deal, switch to the backup and replace the primary when you are next in for servicing.
Which is impossible to achive since many of those addresses are not advertised anywhere public. Many of them probablly aren't even used anywhere.
Another definition would be anyone who buys transit from one of the widely acknolaged tier 1 ISPs and allows you to communicate with them via that route. Of course that then raises the question of cogents status (afaict they are a wannabe tier 1 who refuse to buy transit but are not respected enough to get free peering from all the tier 1 ISPs but i'm not an expert)
Anyone can print a check with any account and routing numbers they want. While checks are low-tech, and easy to copy, they're also very easy to trace. The fraudster's bank has identifying information for whomever cashed the fakes, which makes prosecution trivial.
It seems to me that at least according to the article the problem is that theese numbers can be used internationally.
So your bank knows that your money has been transferred to some "bank" in $THIRD_WORLD_SHITHOLE . Said "bank" either does not have sufficiant evidence to figure out who picked the money up or refuses to cooperate.
What then?
And with credit cards, are you talking about making physical fake cards? Because that's not exactly something one can whip up with supplies from the local hardware store
Afaict plastic card printers and magstripe writers are easy enough to get, Not a job for your local hardware store but plenty of places use ID cards that are very similar to credit cards so the printers are availible. You would probablly have to rig something up to do the embossing but that can't be terriblly difficult.
It's not a hardware store job but it's not out of reach of a reasonablly organised criminal with a few thousand pounds to spend and a location to get stuff delivered to.
Chip and pin cards are probablly much harder to fake but at least here in the UK most places will still put a transaction through with a swipe and sign if chip and pin fails or the card does not have a chip.
from http://www.ubuntu.com/getubuntu/mirror/1
A "full archives" mirror is around 210 gigabytes and is slowly increasing. Archives mirrors are updated at least every 24 hours but no more then every 4 hours.
though you could cut that down quite a lot by using debmirror and only mirroring the releases "sections" and architectures you want
"I don't think that many of you put a photo of their keys online â" with their addresses"
But many people do reveal thier place of work or study online. If you have that and a photograph you can simply follow them to find out where they live.