You've removed my pivotal question in your quote and replied to something much less meaningful. I'm happy to have an interesting discussion if you want to continue, because I think it's an interesting subject and not because I have an agenda to push (because frankly nothing we say here is going to impact existing policy either way) but that conversation is going to be much less interesting if we're answering half-questions.
You made a bold statement that said (and I'm paraphrasing) that because an innocent person could be put to death that no people should be put to death. I'm not agreeing nor disagreeing with that statement, but I'm asking if it's really any different than saying that because an innocent person could spend their entire life in deplorable jail conditions that no person should be put in jail? The burden of proof is higher in a capital case so how does one justify innocent people spending life in prison while at the same time condemning execution?
I'm not 100% convinced that letting an innocent person sit in jail for 60 years is any different, and many would argue is worse, than dying. If that is true then what makes us think that incarceration is morally superior to execution? Irreparable mistakes will be made either way.
As for the guilty, doesn't worry me a bit. As long as they don't get out, society has been equally improved.
Has it really been "equally improved"? Or would the $1.5 million dollars ($30k/year for 50 years) it costs to incarcerate that person for life have a greater social benefit spent in reducing future crime such early childhood education, community outreach programs and mental heath? Further for those that are released after murder, rape, etc, after x years and re-enter the society as their "debt paid" and then re-offend, what is the cost to that, both financially as well as impact to social improvement?
A quick (and horrifying from current social acceptance) math exercise shows that executing the 159,000 people serving life sentences (in US prisons as of 2012) would provide roughly $5 billion dollars a year (or a quarter of a trillion dollars over the period of time those people are serving their sentences) that could be directed to social welfare. If there is one thing we do not have a shortage of on this planet, it's people.
This is like trying to say that because some people die in traffic accidents, it's ok if your kid drowned in the neighbor's pool with no fence. The one has NOTHING to do with the other.
The point I was trying to make was that with any policy there is going to be mistakes, whether those mistakes are in execution, or whether the mistake is in non-execution - those mistakes are going to cause innocent people to die either way.
Is letting the innocent rot in jail for 60+ years until they die from (un)natural causes is any better for that person, or does it just make society feel morally superior? Or, are you really suggesting that "And we must not put people who are not actually guilty in jail. Ever."...?
The sad reality is that every day people die from mistakes, sometimes those mistakes are their own and sometimes they are the mistakes of others.
Except that it's the misused version of 'begets', and not the actual concept of 'begs'. It's about as correct as 'wratched' being used currently in place of 'wretched', but it's never been about 'begging'.
It's easy to remember 20+ web passwords if they mean something to you:
I Use Gmail For Sending Email = IUGFSE.
My Money Is Safe At Toronto Dominion Bank = MMISATDB.
I Love To Eat Pizza At Joe's Pizzeria = ILTEPAJP....add a sequence or some other memorable number, perhaps a standard special character as the 2nd or 3rd character, and capitalize the even, odd, or 4th and last characters or whatever makes sense to you and you end up with:
Really? From TFA: "We are singling out spammers on our network and blocking port 25," said Mitch Bowling, Comcast's vice president of operations. "We don't think it's the right approach to blanket port 25. The right approach is to seek out people who are spamming our network and others."...so, any spammers they find, instead of terminating the account they block port 25. Of course everyone else they don't 'find' can still spam away...
If you aren't able to get a proper reverse DNS entry for your public outbound mail server then you probably shouldn't be running one. If you have a real static IP (as opposed to "my IP doesn't seem to change") - then it shouldn't be a problem getting the DNS setup correctly.
To answer the original question about "what should you do", the answer is simple - if the ISP won't issue a PTR record because of the type of connection being used then the customer should smart-host their mail through the ISP mail servers to ensure global reachability. As you say, often the edge device is a swiss-army knife and in many cases the admin isn't competent enough to properly secure/maintain it. This is exactly what blocking outbound SMTP from dynamic space is meant to accomplish and I'm pleasantly surprised to hear that Comcast/Verizon have finally started to implement what every other responsible ISP has been doing for a decade.
To play devil's advocate using your example it'd be the same as selling "child poison" and saying there are plenty of other things you could do with it.:) I don't disagree with you that tools can be abused for non-intended purposes, but this software is being promoted for its intended purpose. The fix is stronger security protocols of course, but I couldn't resist the analogy - sorry.
Visa doesn't do hex, and they certainly wouldn't convert spaces to binary. Most of the credit card import systems use formatted flat files. The problem is more likely an alignment issue and the credit card number itself was 4231 4885 5308 1845.
Not enough people willing to pay the true costs of having 5Mbps connections running full out 24x7. The rates we pay now are based on a specific usage pattern. When that pattern changes disproportionately based on a handful of users, there are three options. 1) Raise rates, 2) Implement caps, 3) Implement throttling.
> a datapipe is not a water pipe. It's serial, not parallel. > During a burst, basically the pipe is utterly full. > Nothing can get in until enough stuff gets out first.
You make it sound like once someone initiates a large transfer their entire session is 'put in line ahead of anything else'. That is not the way it works however as packets are interleaved as they arrive at the router. Excluding any QoS tagging, generally speaking every packet has the exact same chance of making it through.
My head hurts from reading that. I'm not sure where you are pulling numbers from but they don't smell very good. The reason adsl "gets slow" when busy is the asymetrical part, specifically when you saturate the slower outbound direction with general traffic the ack's to tcp packets can't be sent fast enough to also saturate the faster inbound direction which leads to the download effectively slowing to the same speed as the upload.
I don't disagree that speed restrictions should be protocol independent.:)
There are a couple of problems with the 'being able to use the bandwidth if it is available' though, ie most ISP's have larger capacity links than what they pay for, ie they will have a couple of 1000Mbps connections but pay based on peak usage of that circuit, say for example 300Mbps on each, so if there no restrictions then they flood those circuits and quadruple their costs. Or if they are flat rate, then the concept of being able to use all the bandwidth when it's available still doesn't work for fair sharing, ie someone doing 500GB will get a hundred times the priority of someone doing 5GB so ultimately when that link is saturated it's still impacting the speeds of the majority for the benefit of the few.
Ultimately caps are going to have to become widespread and tiered service the norm. Once the bigger companies realize that the backlash they get from introducing it is from a demographic they don't want anyways, then everyone else will be better off. The average user will be able to continue to receive faster speeds at the same price they are paying now and 500GB users will be shunted to ISP's that don't limit usage and oversubscribe their networks by unmanagable amounts.
Sure, lets stick with food for a moment. The light customer might take 1 plate, the heavy takes 10 plates - it used to be that way with internet as well, ie 10 dialup hours for light customers, 100 hours for heavy customers, then 1GB traffic and 10GB. This is something the business can model, ie for every 95 customers that take 1 plate, there will be 5 customers that take 10 plates giving an average of 1.5 plates per customer. Now those heavy customers are taking 500 plates for an average of 26 plates per customer. See the problem? Feel free to substitute whatever numbers you want but a very small population can cause that average cost of doing business to skyrocket and the average customer doesn't want to pay 10 times the price. So it's not the ISP's that don't want to change speeds as you proclaim rather it is the majority of the other customers that don't want the ISP's to change the speeds, prices or quality to satisify a few very very unprofitable users.
Is it still unlimited, absolutely, within the speed they decide to serve certain packets to you. I have yet to find any ISP that doesn't include the words "up to" when describing their bandwidth speeds. You seem to advocate that by saying they can give you bursts of speed, but then seem to expect that anything below that burst is an unfair limitation?
Try it with the food though, go to your local buffet and take 500 plates of food, eat 10 of them, stuff 240 of them into a storage bin you'll never open again and share the other 250 plates to strangers outside. I expect you'll find some sort of plate limitation speed in effect. Repeat monthly.
Sorry for the double post but I missed the main point. You stated that the ISP could slow everyones traffic down to accomodate things without having to upgrade anything. My point is why should the rest of the users have their connections slowed down when their patterns haven't changed? Why not only slow the traffic of those that would otherwise monopolize their connections? What you are suggesting is to penalize everyone for the actions of a few.
Slashdot Mirror
It certainly is if
You've removed my pivotal question in your quote and replied to something much less meaningful. I'm happy to have an interesting discussion if you want to continue, because I think it's an interesting subject and not because I have an agenda to push (because frankly nothing we say here is going to impact existing policy either way) but that conversation is going to be much less interesting if we're answering half-questions.
You made a bold statement that said (and I'm paraphrasing) that because an innocent person could be put to death that no people should be put to death. I'm not agreeing nor disagreeing with that statement, but I'm asking if it's really any different than saying that because an innocent person could spend their entire life in deplorable jail conditions that no person should be put in jail? The burden of proof is higher in a capital case so how does one justify innocent people spending life in prison while at the same time condemning execution?
I'm not 100% convinced that letting an innocent person sit in jail for 60 years is any different, and many would argue is worse, than dying. If that is true then what makes us think that incarceration is morally superior to execution? Irreparable mistakes will be made either way.
As for the guilty, doesn't worry me a bit. As long as they don't get out, society has been equally improved.
Has it really been "equally improved"? Or would the $1.5 million dollars ($30k/year for 50 years) it costs to incarcerate that person for life have a greater social benefit spent in reducing future crime such early childhood education, community outreach programs and mental heath? Further for those that are released after murder, rape, etc, after x years and re-enter the society as their "debt paid" and then re-offend, what is the cost to that, both financially as well as impact to social improvement?
A quick (and horrifying from current social acceptance) math exercise shows that executing the 159,000 people serving life sentences (in US prisons as of 2012) would provide roughly $5 billion dollars a year (or a quarter of a trillion dollars over the period of time those people are serving their sentences) that could be directed to social welfare. If there is one thing we do not have a shortage of on this planet, it's people.
This is like trying to say that because some people die in traffic accidents, it's ok if your kid drowned in the neighbor's pool with no fence. The one has NOTHING to do with the other.
The point I was trying to make was that with any policy there is going to be mistakes, whether those mistakes are in execution, or whether the mistake is in non-execution - those mistakes are going to cause innocent people to die either way.
Is letting the innocent rot in jail for 60+ years until they die from (un)natural causes is any better for that person, or does it just make society feel morally superior? Or, are you really suggesting that "And we must not put people who are not actually guilty in jail. Ever."...?
The sad reality is that every day people die from mistakes, sometimes those mistakes are their own and sometimes they are the mistakes of others.
dupe: Forgot to login first.
Except that it's the misused version of 'begets', and not the actual concept of 'begs'. It's about as correct as 'wratched' being used currently in place of 'wretched', but it's never been about 'begging'.
It's easy to remember 20+ web passwords if they mean something to you:
I Use Gmail For Sending Email = IUGFSE.
My Money Is Safe At Toronto Dominion Bank = MMISATDB.
I Love To Eat Pizza At Joe's Pizzeria = ILTEPAJP. ...add a sequence or some other memorable number, perhaps a standard special character as the 2nd or 3rd character, and capitalize the even, odd, or 4th and last characters or whatever makes sense to you and you end up with:
iu@GfsE54
mm@IsatdB54
il@TepajP54
...easy to remember, and pretty strong passwords.
And no, our military power couldn't stop an armed populace. The military wouldn't have a chance...unless they wanted to just kill everyone.
Same holds true for an unarmed populace.
+8
It's still British food.
Very rarely are static and dynamic IP's issued from the same subnet. If you can't get out of a blocked range and won't switch ISP's, then smart-host.
Really? From TFA: "We are singling out spammers on our network and blocking port 25," said Mitch Bowling, Comcast's vice president of operations. "We don't think it's the right approach to blanket port 25. The right approach is to seek out people who are spamming our network and others." ...so, any spammers they find, instead of terminating the account they block port 25. Of course everyone else they don't 'find' can still spam away...
If you aren't able to get a proper reverse DNS entry for your public outbound mail server then you probably shouldn't be running one. If you have a real static IP (as opposed to "my IP doesn't seem to change") - then it shouldn't be a problem getting the DNS setup correctly.
To answer the original question about "what should you do", the answer is simple - if the ISP won't issue a PTR record because of the type of connection being used then the customer should smart-host their mail through the ISP mail servers to ensure global reachability. As you say, often the edge device is a swiss-army knife and in many cases the admin isn't competent enough to properly secure/maintain it. This is exactly what blocking outbound SMTP from dynamic space is meant to accomplish and I'm pleasantly surprised to hear that Comcast/Verizon have finally started to implement what every other responsible ISP has been doing for a decade.
5) Stop trying to run a mail server from a dynamic IP address and wondering why the rest of the world doesn't want to accept your mail.
Thank you!
To play devil's advocate using your example it'd be the same as selling "child poison" and saying there are plenty of other things you could do with it. :) I don't disagree with you that tools can be abused for non-intended purposes, but this software is being promoted for its intended purpose. The fix is stronger security protocols of course, but I couldn't resist the analogy - sorry.
Interesting idea on the both hands....
Only in MDU's, otherwise Fibe 16 is the top end which is provisioned with ADSL2+.
Waves of Mass Destruction...
Visa doesn't do hex, and they certainly wouldn't convert spaces to binary. Most of the credit card import systems use formatted flat files. The problem is more likely an alignment issue and the credit card number itself was 4231 4885 5308 1845.
No problem. Oh, you want it for $40 still?
Not enough people willing to pay the true costs of having 5Mbps connections running full out 24x7. The rates we pay now are based on a specific usage pattern. When that pattern changes disproportionately based on a handful of users, there are three options. 1) Raise rates, 2) Implement caps, 3) Implement throttling.
> a datapipe is not a water pipe. It's serial, not parallel.
> During a burst, basically the pipe is utterly full.
> Nothing can get in until enough stuff gets out first.
You make it sound like once someone initiates a large transfer their entire session is 'put in line ahead of anything else'. That is not the way it works however as packets are interleaved as they arrive at the router. Excluding any QoS tagging, generally speaking every packet has the exact same chance of making it through.
My head hurts from reading that. I'm not sure where you are pulling numbers from but they don't smell very good. The reason adsl "gets slow" when busy is the asymetrical part, specifically when you saturate the slower outbound direction with general traffic the ack's to tcp packets can't be sent fast enough to also saturate the faster inbound direction which leads to the download effectively slowing to the same speed as the upload.
Do you _really_ think your ISP has 100Mbps of dedicated internet _per customer_ at euro75?
I don't disagree that speed restrictions should be protocol independent. :)
There are a couple of problems with the 'being able to use the bandwidth if it is available' though, ie most ISP's have larger capacity links than what they pay for, ie they will have a couple of 1000Mbps connections but pay based on peak usage of that circuit, say for example 300Mbps on each, so if there no restrictions then they flood those circuits and quadruple their costs. Or if they are flat rate, then the concept of being able to use all the bandwidth when it's available still doesn't work for fair sharing, ie someone doing 500GB will get a hundred times the priority of someone doing 5GB so ultimately when that link is saturated it's still impacting the speeds of the majority for the benefit of the few.
Ultimately caps are going to have to become widespread and tiered service the norm. Once the bigger companies realize that the backlash they get from introducing it is from a demographic they don't want anyways, then everyone else will be better off. The average user will be able to continue to receive faster speeds at the same price they are paying now and 500GB users will be shunted to ISP's that don't limit usage and oversubscribe their networks by unmanagable amounts.
Sure, lets stick with food for a moment. The light customer might take 1 plate, the heavy takes 10 plates - it used to be that way with internet as well, ie 10 dialup hours for light customers, 100 hours for heavy customers, then 1GB traffic and 10GB. This is something the business can model, ie for every 95 customers that take 1 plate, there will be 5 customers that take 10 plates giving an average of 1.5 plates per customer. Now those heavy customers are taking 500 plates for an average of 26 plates per customer. See the problem? Feel free to substitute whatever numbers you want but a very small population can cause that average cost of doing business to skyrocket and the average customer doesn't want to pay 10 times the price. So it's not the ISP's that don't want to change speeds as you proclaim rather it is the majority of the other customers that don't want the ISP's to change the speeds, prices or quality to satisify a few very very unprofitable users.
Is it still unlimited, absolutely, within the speed they decide to serve certain packets to you. I have yet to find any ISP that doesn't include the words "up to" when describing their bandwidth speeds. You seem to advocate that by saying they can give you bursts of speed, but then seem to expect that anything below that burst is an unfair limitation?
Try it with the food though, go to your local buffet and take 500 plates of food, eat 10 of them, stuff 240 of them into a storage bin you'll never open again and share the other 250 plates to strangers outside. I expect you'll find some sort of plate limitation speed in effect. Repeat monthly.
Sorry for the double post but I missed the main point. You stated that the ISP could slow everyones traffic down to accomodate things without having to upgrade anything. My point is why should the rest of the users have their connections slowed down when their patterns haven't changed? Why not only slow the traffic of those that would otherwise monopolize their connections? What you are suggesting is to penalize everyone for the actions of a few.