mmm, in much of the UK we have a middle ground which seems to work fairly well. What people loosely reffer to as "taxi's" are split into "hackney carridges" and "private hire cars". The fomer are heavilly regulated and limited in number (so the streets aren't flooded with taxi's looking for jobs). The latter are regulated more lightly and AIUI they are not limited in number but they are only allowed to take journeys pre-booked through an operator.
Specifically, RIPE's policy is that each LIR can get one/22 from the final/8, and that's it. The idea is to make sure that new LIRs can at least get some v4 space to run NAT64/CGNAT on.
I wonder if there are there any stats for number of new LIRs registered and whether said stats show an uptick when RIPE introduced this policy?
It can be a good soloution but it does add extra complexity both in maintaining the proxy itself and in modifying abuse-management code in the application to understand the concept of a trusted proxy.
The cost of man/hours to do any software it way higher comparativelly.
Sure, if you are paying for those man hours or doing the software development when you would otherwise have been doing paid work.
If you are doing it as a hobby (instead of watching TV, posting on/. or whatever) then those man hours cost nothing. On the other hand being expected to pay for the privilage of posting free apps is deeply offputting.
Mounting a BGA is one thing (probablly easier than mounting a QFN), fabbing a PCB to mount one on is quite another. It's not so much etching the pads for the chip (BGA ball pitches at least for FPGAs are often larger than QFP or QFN pitches), it's making the connections to get the inner pads of the BGA out. That means lots of really small vias and for all but the smallest chips more than two layers.
The pictures show that he had a VERY high voltage source, high enough to produce visible sparks of significant length (so probablly 10kV or more). It's not clear how he applied it to the devices but I would guess either between two ports or between a port and mains ground (applying it between two pins on the same connector has the problem of how do you stop the connector arcing over to itself).
Isolating transformers are useful things but they do have their limits. The ones in ethernet are designed to deal with mains wiring related faults, not lightning or people with deliberate high voltage sources.
120V (or even 240V) in common mode should do nothing. 120V in differential mode will cause massive overcurrents in the circuitry adjacent to the port and possiblly in the switch as a whole but it's unlikely to be able to go in one port and come out another while still having enough strength to do damage to devices beyond the switch.
However TFA was clearly using a LOT more than mains voltage. A sufficiently large common mode voltage (you can't really apply very high voltages in differential mode because your connector will just flash over) will either punch through or flash over the isolating transformers and find it's way to ground by any means nessacery frying stuff along the way.
Cheap switches are probablly going to be worse than expensive ones because they are less likely to be grounded (meaning the only earth path is to go beyond the switch into devices behind it) and more likely to have isolation transformers that don't actually meet the requirements.
Most PC connectors are non-isolated and referenced to the PCs ground. Apply a large voltage in common mode and it will find it's way to ground through all sorts of paths, many of them likely destructive. Ethernet on the other hand has isolation transformers designed to survive a strength test of arround 1.5KV*. 120V (or even 240V) AC in common mode on an ethernet port should have no affect if the device is not defective.
120 VAC in differential mode will definately fry the port, it may fry the rest of the device but it's unlikely to "jump" an ethernet switch to devices behind. The path of least resistance is a short loop through the transciever, not going all over the board, via another transceiver and back out of another port.
* There are several different strength tests with different combinations of voltage and duration of which the standard requires at least one to be passed.
The real question is how "self contained" will those legal entities have to be before governments/courts consider it unreasonable to punish one entity in the conglomerate for the actions of another? will seperate companies owned by a common parent company suffice? will they have to be owned and controlled by different people? if they had to be controlled/owened by different people how independent do those people have to be? will they have to be using different brands?
Linux may run on x86, but it also runs on MIPS, PPC, ARM etc which Windows generally does not
The amusing thing is it *HAS* run on all those architectures but for whatever reason MS decided to abandon their mips and ppc ports and cripple the arm port.
Your browser gets a list of trusted root certificates and will accept any valid certificate issued by these CAs. On my windows 8 box there are 53. Any of these providers could issue certificates for any number of domains.
Worse those providers can issue "intermediate certificates" which also have the power to issue certificates for any number of domains. They can and do issue those intermediate certificates to third parties. So the list of root certs in your browser is not a complete list of entities who can issue certs your browser will trust.
There was recently an extension added to allow intermediate certs to be limited to certain ranges of names but that only helps in clients new enough to recognise the extension.
There was also recently an extension added for "key pinning" which makes bogus certs less useful.
Google is their own certificate authority.
At least when I go to google and check the cert I get a cert that has a google intermediate and a geotrust root. I don't see any evidence of name constraints on said intermediate cert though:(
There are many ISPs but nearly all of them rely on BT openreach and many of them also rely on BT wholesale. So what you get depends largely on what the BT network looks like in your area. It can vary massively between areas of similar development levels. Either you are quoting advertised rather than actual rates or you got lucky with the areas you chose to live in.
If you are on a small phone exchange then most of the ISPs will be relying on BT wholesale and congestion and/or caps are likely. If you are a long way from the exchange or the wiring happens to be of crap quality then your ADSL speeds will suck.
The recent introduction of FTTC+VDSL services (presumabblly what you are reffering to when you say "fiber") has improved things for many people but it isn't deployed everywhere yet. can still be problematic in the case of long or shitty wiring from cabinet to user and in some areas there have been oversubscription issues (when installing FTTC equipment openreach assumes that there will be initally only a small number of users on a given cabinet, this isn't always true). There was also an issue where interference concerns prevented installation of VDSL equipment in the exchange thereby preventing people on "exchange only" lines from getting VDSL though I belive BT are working on inserting cabinets to solve that one.
If you live in a virgin media area you have the option of using their network, unfortunately using their network means you are stuck with them as an ISP. Upstream speeds also suck.
There have also been a few upstarts offering very fast service to the lucky few but afaict having a negligable impact on the overall picture.
From what I understand, for gaming the latency (RTT) is more important than the speed of the connection.
When you are actually playing that is true (though it does vary by game, some games use a lot more bandwidth than others).
When you are trying to get into the game it's another matter. Many games require you to be fully up to date before you can even connect to the matchmaking servers and many games can download additional content when connecting to a game.
And if you have multiple gamers in the household having more bandwidth tends to reduce the latency spikes when one gamer is in-map while another is downloading content.
I see a couple of practical problems with taking the bitcoin blockchain concept and attempting to apply it to balances denominated in government currency (or any other extenal asset)
1: the bitcoin blockchain gains it's security against modification of history from having lots of computing power behind it. That computing power is currently "paid for" by issuing new bitcoins*. The beauty of this system is that as the userbase grows and the value of a bitcoin grows the effective reward and hence the ammount of computing power devoted to keeping the system secure grows with it. A system where the balances are exchangable at a fixed rate for "governement currency" has to be far more careful about creating balances out of thin air if it doesn't want to go bankrupt in short order. 2: For a blockchain balance demoninated in "government currency" to be anything more than a fiction there has to be an entity that can actually make the exchanges both taking "government currency" and exchanging it for a balance in the blockchain (which means they need the ability to create balances in the blockchain out of thin air) and taking a balance in the blockchain and exchanging it for government money. That means you have a central entity for government regulators who don't like anonymous transfers to attack.
*Bitcoin does eventually intend to transition for a scheme where the "miners" are paid through transaction fees but it remains to be seen if that transaction will be a success or if mining power will drop off to the point where a 51% attack becomes accessible to a far wider range of adversatories.
It doesn't need to. Once you have the data in digital form you can keep migrating it to whatever the best storage device of the day is without incurring any further quality loss.
The term stable has a couple of subtuly different meanings. There is the "doesn't crash very often" meaning and the "what worked today will work in the future" meaning.
The only reasonable way to get security updates for debian testing/unstable is to keep rolling forward with them. That means you are close to the bleeding edge but it also means you are building on something that is distaincly a moving target and that can and will break your stuff. The same applies to the development releases of ubuntu.
At the other end of the scale you have Debian stable and Ubuntu LTS. Lots of stability but also old software. The release cycles are not synced so sometimes it makes sense to choose between them based on which of the two had a release more recently.
What Ubuntu offers that Debian doesn't is the 6-monthly releases, these offer something of a middle ground on the stability/up to dateness scale.
Yes webp is a google creation. It's basically a single still frame from the vp8 video codec (as used by webm). Being based on modern techniques it gives a better quality for a given size (or smaller size for a given quality) than JPEG and if you have support for webm then implementing webp as well requires very little extra code.
However it has failed to catch on more widely. Afaict chrome is the only major browser that supports it. There is a bug requesting supporting in firefox but it doesn't seem to be going anywhere. IE and safari seem even less likely to adopt it.
I have no idea whether they will bother to implement this service on superhubs in modem mode but I don't see any obvious reason why they couldn't.
Public wifi services that rely on piggybacking on end-user routers are always going to be spotty. Especially in a market like the UK where there are many competing ISPs. I doubt a few more users dropping out because they happen to have their superhub in an unusual mode probablly won't matter much in the grand scheme of things.
In particular the standard userland interfaces (libc and such) on 32-bit ports of linux still use 32-bit time_t. For embedded distros that can afford to sacrifice binary compatibility with both older versions of themselves and regular linux systems this is fairly easy to fix but for more general purpose distros that care about binary compatibility it is much harder and people aren't sure if it's worth doing it.
While the beagleboard line are familiar to many of us linux geeks they don't have anything like the general recognition that tablets or even raspberry pi's do and they have been somewhat uncompetitive for a while (especially since the pi2 and odroid c1 showed up). So they aren't an especially good target for drawing marketing comparisions with.
OTOH creating the electronic equivilent of a ratty photocopy is pretty easy.
While dealing with the electronic equivilent of a ratty photocopy is not exactly pleasant it's usually better than not having any documentation at all.
Slashdot Mirror
mmm, in much of the UK we have a middle ground which seems to work fairly well. What people loosely reffer to as "taxi's" are split into "hackney carridges" and "private hire cars". The fomer are heavilly regulated and limited in number (so the streets aren't flooded with taxi's looking for jobs). The latter are regulated more lightly and AIUI they are not limited in number but they are only allowed to take journeys pre-booked through an operator.
Specifically, RIPE's policy is that each LIR can get one /22 from the final /8, and that's it. The idea is to make sure that new LIRs can at least get some v4 space to run NAT64/CGNAT on.
I wonder if there are there any stats for number of new LIRs registered and whether said stats show an uptick when RIPE introduced this policy?
It can be a good soloution but it does add extra complexity both in maintaining the proxy itself and in modifying abuse-management code in the application to understand the concept of a trusted proxy.
There's technical reasons for the length as by assigning humongous blocks at a time routing is greatly simplified.
OTOH by discouraging NAT they encourage people to apply for PI space. PI space is the enemy of efficient routing.
If each site is on a seperate VM then either each site gets a public v4 IP or you add the additional complications of reverse proxies to the system.
The cost of man/hours to do any software it way higher comparativelly.
Sure, if you are paying for those man hours or doing the software development when you would otherwise have been doing paid work.
If you are doing it as a hobby (instead of watching TV, posting on /. or whatever) then those man hours cost nothing. On the other hand being expected to pay for the privilage of posting free apps is deeply offputting.
Mounting a BGA is one thing (probablly easier than mounting a QFN), fabbing a PCB to mount one on is quite another. It's not so much etching the pads for the chip (BGA ball pitches at least for FPGAs are often larger than QFP or QFN pitches), it's making the connections to get the inner pads of the BGA out. That means lots of really small vias and for all but the smallest chips more than two layers.
The pictures show that he had a VERY high voltage source, high enough to produce visible sparks of significant length (so probablly 10kV or more). It's not clear how he applied it to the devices but I would guess either between two ports or between a port and mains ground (applying it between two pins on the same connector has the problem of how do you stop the connector arcing over to itself).
Isolating transformers are useful things but they do have their limits. The ones in ethernet are designed to deal with mains wiring related faults, not lightning or people with deliberate high voltage sources.
120V (or even 240V) in common mode should do nothing. 120V in differential mode will cause massive overcurrents in the circuitry adjacent to the port and possiblly in the switch as a whole but it's unlikely to be able to go in one port and come out another while still having enough strength to do damage to devices beyond the switch.
However TFA was clearly using a LOT more than mains voltage. A sufficiently large common mode voltage (you can't really apply very high voltages in differential mode because your connector will just flash over) will either punch through or flash over the isolating transformers and find it's way to ground by any means nessacery frying stuff along the way.
Cheap switches are probablly going to be worse than expensive ones because they are less likely to be grounded (meaning the only earth path is to go beyond the switch into devices behind it) and more likely to have isolation transformers that don't actually meet the requirements.
Most PC connectors are non-isolated and referenced to the PCs ground. Apply a large voltage in common mode and it will find it's way to ground through all sorts of paths, many of them likely destructive. Ethernet on the other hand has isolation transformers designed to survive a strength test of arround 1.5KV*. 120V (or even 240V) AC in common mode on an ethernet port should have no affect if the device is not defective.
120 VAC in differential mode will definately fry the port, it may fry the rest of the device but it's unlikely to "jump" an ethernet switch to devices behind. The path of least resistance is a short loop through the transciever, not going all over the board, via another transceiver and back out of another port.
* There are several different strength tests with different combinations of voltage and duration of which the standard requires at least one to be passed.
The real question is how "self contained" will those legal entities have to be before governments/courts consider it unreasonable to punish one entity in the conglomerate for the actions of another? will seperate companies owned by a common parent company suffice? will they have to be owned and controlled by different people? if they had to be controlled/owened by different people how independent do those people have to be? will they have to be using different brands?
Linux may run on x86, but it also runs on MIPS, PPC, ARM etc which Windows generally does not
The amusing thing is it *HAS* run on all those architectures but for whatever reason MS decided to abandon their mips and ppc ports and cripple the arm port.
Your browser gets a list of trusted root certificates and will accept any valid certificate issued by these CAs. On my windows 8 box there are 53. Any of these providers could issue certificates for any number of domains.
Worse those providers can issue "intermediate certificates" which also have the power to issue certificates for any number of domains. They can and do issue those intermediate certificates to third parties. So the list of root certs in your browser is not a complete list of entities who can issue certs your browser will trust.
There was recently an extension added to allow intermediate certs to be limited to certain ranges of names but that only helps in clients new enough to recognise the extension.
There was also recently an extension added for "key pinning" which makes bogus certs less useful.
Google is their own certificate authority.
At least when I go to google and check the cert I get a cert that has a google intermediate and a geotrust root. I don't see any evidence of name constraints on said intermediate cert though :(
There are many ISPs but nearly all of them rely on BT openreach and many of them also rely on BT wholesale. So what you get depends largely on what the BT network looks like in your area. It can vary massively between areas of similar development levels. Either you are quoting advertised rather than actual rates or you got lucky with the areas you chose to live in.
If you are on a small phone exchange then most of the ISPs will be relying on BT wholesale and congestion and/or caps are likely. If you are a long way from the exchange or the wiring happens to be of crap quality then your ADSL speeds will suck.
The recent introduction of FTTC+VDSL services (presumabblly what you are reffering to when you say "fiber") has improved things for many people but it isn't deployed everywhere yet. can still be problematic in the case of long or shitty wiring from cabinet to user and in some areas there have been oversubscription issues (when installing FTTC equipment openreach assumes that there will be initally only a small number of users on a given cabinet, this isn't always true). There was also an issue where interference concerns prevented installation of VDSL equipment in the exchange thereby preventing people on "exchange only" lines from getting VDSL though I belive BT are working on inserting cabinets to solve that one.
If you live in a virgin media area you have the option of using their network, unfortunately using their network means you are stuck with them as an ISP. Upstream speeds also suck.
There have also been a few upstarts offering very fast service to the lucky few but afaict having a negligable impact on the overall picture.
From what I understand, for gaming the latency (RTT) is more important than the speed of the connection.
When you are actually playing that is true (though it does vary by game, some games use a lot more bandwidth than others).
When you are trying to get into the game it's another matter. Many games require you to be fully up to date before you can even connect to the matchmaking servers and many games can download additional content when connecting to a game.
And if you have multiple gamers in the household having more bandwidth tends to reduce the latency spikes when one gamer is in-map while another is downloading content.
I see a couple of practical problems with taking the bitcoin blockchain concept and attempting to apply it to balances denominated in government currency (or any other extenal asset)
1: the bitcoin blockchain gains it's security against modification of history from having lots of computing power behind it. That computing power is currently "paid for" by issuing new bitcoins*. The beauty of this system is that as the userbase grows and the value of a bitcoin grows the effective reward and hence the ammount of computing power devoted to keeping the system secure grows with it. A system where the balances are exchangable at a fixed rate for "governement currency" has to be far more careful about creating balances out of thin air if it doesn't want to go bankrupt in short order.
2: For a blockchain balance demoninated in "government currency" to be anything more than a fiction there has to be an entity that can actually make the exchanges both taking "government currency" and exchanging it for a balance in the blockchain (which means they need the ability to create balances in the blockchain out of thin air) and taking a balance in the blockchain and exchanging it for government money. That means you have a central entity for government regulators who don't like anonymous transfers to attack.
*Bitcoin does eventually intend to transition for a scheme where the "miners" are paid through transaction fees but it remains to be seen if that transaction will be a success or if mining power will drop off to the point where a 51% attack becomes accessible to a far wider range of adversatories.
It doesn't need to. Once you have the data in digital form you can keep migrating it to whatever the best storage device of the day is without incurring any further quality loss.
The term stable has a couple of subtuly different meanings. There is the "doesn't crash very often" meaning and the "what worked today will work in the future" meaning.
The only reasonable way to get security updates for debian testing/unstable is to keep rolling forward with them. That means you are close to the bleeding edge but it also means you are building on something that is distaincly a moving target and that can and will break your stuff. The same applies to the development releases of ubuntu.
At the other end of the scale you have Debian stable and Ubuntu LTS. Lots of stability but also old software. The release cycles are not synced so sometimes it makes sense to choose between them based on which of the two had a release more recently.
What Ubuntu offers that Debian doesn't is the 6-monthly releases, these offer something of a middle ground on the stability/up to dateness scale.
Yes webp is a google creation. It's basically a single still frame from the vp8 video codec (as used by webm). Being based on modern techniques it gives a better quality for a given size (or smaller size for a given quality) than JPEG and if you have support for webm then implementing webp as well requires very little extra code.
However it has failed to catch on more widely. Afaict chrome is the only major browser that supports it. There is a bug requesting supporting in firefox but it doesn't seem to be going anywhere. IE and safari seem even less likely to adopt it.
I prefer 1280x1024 to 1440x900.
But 1920x1080 is better than either and in my experiance also cheaper.
I have no idea whether they will bother to implement this service on superhubs in modem mode but I don't see any obvious reason why they couldn't.
Public wifi services that rely on piggybacking on end-user routers are always going to be spotty. Especially in a market like the UK where there are many competing ISPs. I doubt a few more users dropping out because they happen to have their superhub in an unusual mode probablly won't matter much in the grand scheme of things.
In particular the standard userland interfaces (libc and such) on 32-bit ports of linux still use 32-bit time_t. For embedded distros that can afford to sacrifice binary compatibility with both older versions of themselves and regular linux systems this is fairly easy to fix but for more general purpose distros that care about binary compatibility it is much harder and people aren't sure if it's worth doing it.
Note: my information is based on http://www.slideshare.net/lina... if anyone has more recent information i'd like to hear it.
I question the practicality of this technique for ATMs; you still need a clone of the card to use the PIN.
Or just steal the card.
While the beagleboard line are familiar to many of us linux geeks they don't have anything like the general recognition that tablets or even raspberry pi's do and they have been somewhat uncompetitive for a while (especially since the pi2 and odroid c1 showed up). So they aren't an especially good target for drawing marketing comparisions with.
OTOH creating the electronic equivilent of a ratty photocopy is pretty easy.
While dealing with the electronic equivilent of a ratty photocopy is not exactly pleasant it's usually better than not having any documentation at all.