No, an idle pipe *should* cost the same as one running 100%... But that's simply not the case, you pay a fixed amount for the physical connection, then you pay based on how much of it you used. Search for "IP Transit" and see how the major ISPs charge for it.
It's not as bad as it used to be... There is no longer a "local rate", and national calls cost what the local rate used to. They aren't free, but there are plenty of bundles getting you national calls bundled in with you're line rental etc. Local calls in the US aren't free either, they are included in the line rental.
They are typically charged on a 95th percentile basis, that is they ignore the top 5% of usage and go for the highest from the remaining 95% over a 1 month period. Lots of heavy users push this up... but a heavy user during an otherwise idle period makes little difference. In the UK it's not just the transit bandwidth to the internet tho, providers have to pay for bandwidth usage on the DSL connections between the ISP and the customer.
ISPs already adapt to the increased demand, by offering more expensive services to satisfy that demand. The cheap services only work on the basis that they are mostly idle. If those connections are active all the time, it costs the ISP more to provide. For that to be sustainable, they have to charge more for the service. If you don't like it, don't rag on the tier-2 isp's like plusnet, complain to the tier-1's who charge for bandwidth usage, and to the local loop providers who charge for bandwidth usage between customer and isp. They cant provide a service for less than it costs them, go to the source of the services.
Bytes do cost money, because ISP's charge money for them. As noted in the summary, Plusnet is a tier-2 ISP... Tier-1's charge smaller companies for their traffic use. Tho they may have gigabit capable cables, they will pay depending how much capacity is used.
When it comes to ADSL in the UK, BT charge you for the physical connection, then they charge you per line, then they charge you based on the amount of data transferred. Having a bunch of idle customers is much cheaper than a bunch of p2p users. This is also why ISPs impose transfer limits, if everyone ran their connection flat out 24/7 it would be unsustainable for the ISP... That's why dedicated leased lines cost so much more - you're paying for the bandwidth usage... Similarly, a 10mb line is cheaper than a 100mb line, even tho they will be delivered over the same fibre, consume the same amount of power etc.
Sure, bytes *shouldnt* cost money, but currently they do.
The trouble is, there is no single entity in control of the internet. Changing a currency in a country, or turning off analog tv is pretty easy. The government can simply order it to be done. This won't work on the internet, any country forcing the use of V6 will effectively be cut off from everywhere else. What could be done however, is require that any device with a V4 address also has a V6 address. Any networking device being sold must support V6. Any ISP supplying connectivity must supply V6 etc... Force everyone to go dual stack, and then phase out V4 at a later date. It's not until everyone uses v6 that anyone will even consider dropping V4.
The current state is terrible. I don't know of any consumer level broadband routers which support V6, I had to buy a Cisco device to get V6 support. I only know of 2 ISPs which provide IPv6 over DSL, and none who provide it over cable... And relatively few who provide it for hosted services in a datacenter. And having spoken to a friend at the ISP i use, of roughly 1000 users connected to a DSL access node, 2 have IPv6 fully active and routed (him and me), and 3 more have IPv6 capable devices but aren't using the capability.
And despite the pair of/8 blocks HP have, they still use significant sized blocks outside of these blocks too: www.compaq.com has address 161.114.23.244 And they used to have all their webservers outside of the main/8 blocks too...
Apple also, used to have most of their stuff outside but it seems now at least their forward facing stuff has been moved to 17.x. Then theres companies like GE and Ford who only use the IPs internally, and don't route any part of their/8 blocks to the Internet. dns005.ford.com has address 136.8.159.21
etc... Their website is hosted by akamai now, but i'm sure it had a 136.x address when i looked before.
Really they need to take back the large ipblocks that were allocated to companies years ago, but which aren't even being used . Ford has a/8 ipblock (16777216 addresses) that they use internally and dont route to the internet, why cant they use 10.0.0.0/8 internally like everyone else?
IPv6 is not new, it's been around for years... Some 10 year old kit does support it fine. Backbone routers from 10 years ago however, there aren't too many left, at least on major backbones... What was considered fast 10 years ago is pretty slow these days, major backbones have long since upgraded.
As for what supports V6, most enterprise level kit does, most consumer oriented routers etc, don't.. Every major OS now supports V6 too.
The trouble with siteadvisor, is that it's quite easy to identify when it's hitting your site. With that in mind, it's fairly easy to serve up a different site to siteadvisor, or just not serve the malware.
To see an example of a site that does this, look up www.acunetix.com on siteadvisor, notice how siteadvisor has downloaded some programs from their site and verified them malware free, look also how siteadvisor has submitted its email address to the site and not received any email. Now go to acunetix.com, and try to download the same programs, you will need to submit you're email to do so... You will quite quickly receive an automated mailing.. Now why didn't siteadvisor get this automated mail?
Actually there are lots of open source applications which can do shared calendaring, and there has been for years. (rpc.cmsd, ical, caldav). The problem is that you have to migrated *everything* at once... Exchange won't talk any kind of standard protocols for anything but mail. Outlook won't talk any standard protocols for anything but mail. If you want to migrate you have to switch all your clients and servers at once, you cant switch them gradually. Also you lose all you're data, because exchange doesn't store the data in a format that can easily be converted to any of the standard formats used by other apps. If people could migrate gradually and with less disruption, a lot more people would migrate away from exchange.
Most microsoft protocols/formats are only intended to be used with one program, or one server / one client, and look at the problems it makes them:
Entourage (OSX outlook equivalent) cannot talk natively to exchange, it hooks over the web interface. MS Publisher cannot read/write word documents properly, importing sometimes works, exporting rarely does, its import/export is vastly inferior to openoffice actually.
Had they been documented properly internally, they would have been able to implement them in other products much more easily.
Well, surely it would have been good practice to actually document formats/protocols/specs/etc *before* you implement... You create a spec of what you need, then you implement it and adjust the spec if you really need to... That way the code gets written more efficiently, you dont end up writing a big chunk of code than having to rewrite it again because its fundamentally incompatible with something you later found out you need but hadn't thought of while writing it. And you could at least document as you go along, it makes maintaining the code a lot easier in future, especially if different people will be expected to read it.
A T1 may only be 1.5mb, but its 1.5mb both ways. Also you're far more likely to get other useful services with a T1 like multiple IPs, reverse DNS, an SLA offering guaranteed uptime, and the ability to use the full 1.5mb in both directions day in day out with no fear of disconnection, contention or shaping. Consumer ISPs will usually slow down (due to shaping or congestion) during busy periods, and they might terminate your connection if you use obscene levels of bandwidth. With a T1, you are paying for the 1.5mb dedicated to you, not shared with anyone else.
I can understand not showing the ads to people in comcast service areas, as they wouldn't be potential customers anyway... But only showing the ads to existing customers? How do they intend to attract new customers?
Their service costs too much? It's pretty cheap for what you get, and the only way they can keep this price point sustainable is by shaping heavy users. It works based on the fact that most users use only a fraction of the available bandwidth. Even paying for a 6Mbps guaranteed rate would cost you considerably more, remember 6Mbps is 4x T1 speed, although with 4 T1 lines you'd get better upstream.
While true for a lot of smaller projects, most of the important and widely used ones are decently audited (things like openssl etc). That said, if something was really business critical it's not a big step to contract someone in to take a look at it either, any business of significant size will get external contractors in to audit any important systems anyway, having the source means they can do a more thorough job.
Depends what the peers are... I run bittorrent on a fast colo box (100mb link), and for popular well seeded torrents often get download speeds upwards of 80mbit. Other people do the same, so even with a small number of seeds i can see very good speeds. Things like linux iso's tend to be seeded by fast servers (the same servers that host fast ftp mirrors usually), and its not unusual to get several mb/sec from a single peer.
The trouble is modern bloated graphics/flash laden websites, that would load pathetically slowly on dialup... Because many people think they need broadband, many people get it, resulting in website authors thinking everyone has it. Hopefully the prevalence of mobiles with limited browsers and slow connections will help webmasters see the error of their ways.
By donating to charity you will buy good PR, which is otherwise quite expensive. You can also donate goods to charity, and claim a tax break relative to what they would have been sold at... This is designed for goods where there is a tangible cost to produce them and a small margin, and the tax break means that the company can afford to donate more goods for the same cost. But when it comes to software, which is virtually 100% profit, such a company actually directly profits from "giving" it to charity.
Also, for all the money the gates foundation (and other similar organizations) spends on medical research, how much of this research goes into the public domain, and how much goes to pharmaceutical companies owned by the very same people who own the foundations? Similarly, how many of their donations come with strings attached, like "heres $1 million for drugs, but you have to buy all you're drugs from a specific company"... So the entire $1mil goes back to said drugs company, as does other money that came from other sources - a net win for the owners of the foundations. Similarly gates has been known to make "donations" on condition that various schools etc use microsoft software exclusively.
Genuine philanthropists would hand over money without any strings attached, and often do so anonymously, some big charities like oxfam receive large anonymous donations at times.
Well, just because you may not know too much about C or encryption... I'm not really inclined to trust some company that says product X is secure, but i'm far more likely to trust a string of unconnected individuals, especially if some of those individuals are recognised cryptography experts or have at least studied cryptography at a reputable establishment. Sure it's not perfect, but its a huge step in the right direction. The only perfect solution would be to study cryptography and programming (in whatever language) yourself first.
Slashdot Mirror
No, an idle pipe *should* cost the same as one running 100%...
But that's simply not the case, you pay a fixed amount for the physical connection, then you pay based on how much of it you used. Search for "IP Transit" and see how the major ISPs charge for it.
It's not as bad as it used to be...
There is no longer a "local rate", and national calls cost what the local rate used to.
They aren't free, but there are plenty of bundles getting you national calls bundled in with you're line rental etc.
Local calls in the US aren't free either, they are included in the line rental.
They are typically charged on a 95th percentile basis, that is they ignore the top 5% of usage and go for the highest from the remaining 95% over a 1 month period.
Lots of heavy users push this up... but a heavy user during an otherwise idle period makes little difference.
In the UK it's not just the transit bandwidth to the internet tho, providers have to pay for bandwidth usage on the DSL connections between the ISP and the customer.
ISPs already adapt to the increased demand, by offering more expensive services to satisfy that demand.
The cheap services only work on the basis that they are mostly idle. If those connections are active all the time, it costs the ISP more to provide. For that to be sustainable, they have to charge more for the service.
If you don't like it, don't rag on the tier-2 isp's like plusnet, complain to the tier-1's who charge for bandwidth usage, and to the local loop providers who charge for bandwidth usage between customer and isp.
They cant provide a service for less than it costs them, go to the source of the services.
Bytes do cost money, because ISP's charge money for them.
As noted in the summary, Plusnet is a tier-2 ISP... Tier-1's charge smaller companies for their traffic use. Tho they may have gigabit capable cables, they will pay depending how much capacity is used.
When it comes to ADSL in the UK, BT charge you for the physical connection, then they charge you per line, then they charge you based on the amount of data transferred. Having a bunch of idle customers is much cheaper than a bunch of p2p users. This is also why ISPs impose transfer limits, if everyone ran their connection flat out 24/7 it would be unsustainable for the ISP...
That's why dedicated leased lines cost so much more - you're paying for the bandwidth usage...
Similarly, a 10mb line is cheaper than a 100mb line, even tho they will be delivered over the same fibre, consume the same amount of power etc.
Sure, bytes *shouldnt* cost money, but currently they do.
The trouble is, there is no single entity in control of the internet.
Changing a currency in a country, or turning off analog tv is pretty easy. The government can simply order it to be done. This won't work on the internet, any country forcing the use of V6 will effectively be cut off from everywhere else.
What could be done however, is require that any device with a V4 address also has a V6 address. Any networking device being sold must support V6. Any ISP supplying connectivity must supply V6 etc... Force everyone to go dual stack, and then phase out V4 at a later date. It's not until everyone uses v6 that anyone will even consider dropping V4.
The current state is terrible. I don't know of any consumer level broadband routers which support V6, I had to buy a Cisco device to get V6 support. I only know of 2 ISPs which provide IPv6 over DSL, and none who provide it over cable... And relatively few who provide it for hosted services in a datacenter.
And having spoken to a friend at the ISP i use, of roughly 1000 users connected to a DSL access node, 2 have IPv6 fully active and routed (him and me), and 3 more have IPv6 capable devices but aren't using the capability.
And despite the pair of /8 blocks HP have, they still use significant sized blocks outside of these blocks too: /8 blocks too...
/8 blocks to the Internet.
www.compaq.com has address 161.114.23.244
And they used to have all their webservers outside of the main
Apple also, used to have most of their stuff outside but it seems now at least their forward facing stuff has been moved to 17.x.
Then theres companies like GE and Ford who only use the IPs internally, and don't route any part of their
dns005.ford.com has address 136.8.159.21
etc... Their website is hosted by akamai now, but i'm sure it had a 136.x address when i looked before.
Really they need to take back the large ipblocks that were allocated to companies years ago, but which aren't even being used . Ford has a /8 ipblock (16777216 addresses) that they use internally and dont route to the internet, why cant they use 10.0.0.0/8 internally like everyone else?
IPv6 is not new, it's been around for years... Some 10 year old kit does support it fine.
Backbone routers from 10 years ago however, there aren't too many left, at least on major backbones... What was considered fast 10 years ago is pretty slow these days, major backbones have long since upgraded.
As for what supports V6, most enterprise level kit does, most consumer oriented routers etc, don't.. Every major OS now supports V6 too.
Altavista used to... Back when it was run by DEC.
See:
http://www.ipv6.org/v6-www.html
Microsoft research have a v6 site too...
My site (www.ev4.org) is also available on v6, just incase anyone cares.
The trouble with siteadvisor, is that it's quite easy to identify when it's hitting your site.
With that in mind, it's fairly easy to serve up a different site to siteadvisor, or just not serve the malware.
To see an example of a site that does this, look up www.acunetix.com on siteadvisor, notice how siteadvisor has downloaded some programs from their site and verified them malware free, look also how siteadvisor has submitted its email address to the site and not received any email.
Now go to acunetix.com, and try to download the same programs, you will need to submit you're email to do so... You will quite quickly receive an automated mailing.. Now why didn't siteadvisor get this automated mail?
Actually there are lots of open source applications which can do shared calendaring, and there has been for years. (rpc.cmsd, ical, caldav).
The problem is that you have to migrated *everything* at once...
Exchange won't talk any kind of standard protocols for anything but mail.
Outlook won't talk any standard protocols for anything but mail.
If you want to migrate you have to switch all your clients and servers at once, you cant switch them gradually. Also you lose all you're data, because exchange doesn't store the data in a format that can easily be converted to any of the standard formats used by other apps.
If people could migrate gradually and with less disruption, a lot more people would migrate away from exchange.
Most microsoft protocols/formats are only intended to be used with one program, or one server / one client, and look at the problems it makes them:
Entourage (OSX outlook equivalent) cannot talk natively to exchange, it hooks over the web interface.
MS Publisher cannot read/write word documents properly, importing sometimes works, exporting rarely does, its import/export is vastly inferior to openoffice actually.
Had they been documented properly internally, they would have been able to implement them in other products much more easily.
Well, surely it would have been good practice to actually document formats/protocols/specs/etc *before* you implement...
You create a spec of what you need, then you implement it and adjust the spec if you really need to... That way the code gets written more efficiently, you dont end up writing a big chunk of code than having to rewrite it again because its fundamentally incompatible with something you later found out you need but hadn't thought of while writing it.
And you could at least document as you go along, it makes maintaining the code a lot easier in future, especially if different people will be expected to read it.
A T1 may only be 1.5mb, but its 1.5mb both ways.
Also you're far more likely to get other useful services with a T1 like multiple IPs, reverse DNS, an SLA offering guaranteed uptime, and the ability to use the full 1.5mb in both directions day in day out with no fear of disconnection, contention or shaping. Consumer ISPs will usually slow down (due to shaping or congestion) during busy periods, and they might terminate your connection if you use obscene levels of bandwidth. With a T1, you are paying for the 1.5mb dedicated to you, not shared with anyone else.
Well, with TCP window scaling you should be able to get reasonable throughput even with fairly high latency...
I can understand not showing the ads to people in comcast service areas, as they wouldn't be potential customers anyway...
But only showing the ads to existing customers? How do they intend to attract new customers?
Their service costs too much?
It's pretty cheap for what you get, and the only way they can keep this price point sustainable is by shaping heavy users. It works based on the fact that most users use only a fraction of the available bandwidth.
Even paying for a 6Mbps guaranteed rate would cost you considerably more, remember 6Mbps is 4x T1 speed, although with 4 T1 lines you'd get better upstream.
Bittorrent keeps a lot of connections open, low end routers can't keep so many active connections in their state tables so they end up falling over.
Strictly speaking they are only "forging" the identity of an IP address they own.
While true for a lot of smaller projects, most of the important and widely used ones are decently audited (things like openssl etc).
That said, if something was really business critical it's not a big step to contract someone in to take a look at it either, any business of significant size will get external contractors in to audit any important systems anyway, having the source means they can do a more thorough job.
Depends what the peers are...
I run bittorrent on a fast colo box (100mb link), and for popular well seeded torrents often get download speeds upwards of 80mbit. Other people do the same, so even with a small number of seeds i can see very good speeds. Things like linux iso's tend to be seeded by fast servers (the same servers that host fast ftp mirrors usually), and its not unusual to get several mb/sec from a single peer.
The trouble is modern bloated graphics/flash laden websites, that would load pathetically slowly on dialup...
Because many people think they need broadband, many people get it, resulting in website authors thinking everyone has it. Hopefully the prevalence of mobiles with limited browsers and slow connections will help webmasters see the error of their ways.
By donating to charity you will buy good PR, which is otherwise quite expensive.
You can also donate goods to charity, and claim a tax break relative to what they would have been sold at... This is designed for goods where there is a tangible cost to produce them and a small margin, and the tax break means that the company can afford to donate more goods for the same cost. But when it comes to software, which is virtually 100% profit, such a company actually directly profits from "giving" it to charity.
Also, for all the money the gates foundation (and other similar organizations) spends on medical research, how much of this research goes into the public domain, and how much goes to pharmaceutical companies owned by the very same people who own the foundations?
Similarly, how many of their donations come with strings attached, like "heres $1 million for drugs, but you have to buy all you're drugs from a specific company"... So the entire $1mil goes back to said drugs company, as does other money that came from other sources - a net win for the owners of the foundations. Similarly gates has been known to make "donations" on condition that various schools etc use microsoft software exclusively.
Genuine philanthropists would hand over money without any strings attached, and often do so anonymously, some big charities like oxfam receive large anonymous donations at times.
Well, just because you may not know too much about C or encryption...
I'm not really inclined to trust some company that says product X is secure, but i'm far more likely to trust a string of unconnected individuals, especially if some of those individuals are recognised cryptography experts or have at least studied cryptography at a reputable establishment.
Sure it's not perfect, but its a huge step in the right direction. The only perfect solution would be to study cryptography and programming (in whatever language) yourself first.