> Email is already a best-effort service without guaranteed delivery. > If the S&R team actually needs a particular piece of information delivered immediately, they should choose a service that is optimized for that purpose.
You are mistaken. Most streaming protocols are best effort and layered on top of UDP, which is explicitly defined as a best-effort service. Packets may or may not arrive, and may arrive out of order.
SMTP is defined as a reliable service, and runs over TCP, also defined as a reliable service. Packets that don't make it the first time will be retransmitted and they will be placed in the correct order. Emails that cannot be delivered will be retried for several days. If, after five days or so, the email absolutely cannot be delivered, the sender will be notified of the failure.
So tell me again why an organization using a service that is defined to be reliable, for life-safety purposes, shouldn't be allowed to buy a slower but more reliable internet connection, one unlike Netflix , who wants lots of bandwidth but doesn't care if 1% of packets are dropped and lost forever? Why exactly do we all have to be forced to use a fast, unreliable network, (or all be forced to use a slow, reliable one), rather than being able to choose and buy the service with the SLA we want?
Your point is certainly right, and your explanation of why is pretty much on target. One very minor point:
> for VOIP it is essential that each packet be delivered, in order, and as quickly as possible. Delays of even 100ms create audible distortion. For email...
For EMAIL, "it is essential that each packet be delivered". You wouldn't tolerate an email service that dropped words out of the middle of your email. For VOIP or other live streaming, you want the packets NOT be delivered if they can't be delivered in order and on time. A phone call that drops out for split second is slightly annoying, a call that changes "ex Stacy" into "sexy Stac" is intolerable.
While horrible latency, over 200 ms, is bad with VOIP, the main issue is that you want CONSISTENT latency. You don't want "each packet as quickly as possible", you want each packet delivered as quickly or slowly as the last packet. If we represent silence with an underscore, the difference can be seen as:
___________This_message_post_by_me. (High, but consistent latency) Th__is__mess__agepo__ostb__yme. (Low, but inconsistent latency)
> I don't see how "the ISP should treat every packet the same" is unreasonable.... > If the user needs a stronger guarantee, they should get a better connection with a better SLA. None of this is illegal or unreasonable.
In your first sentence you made it ILLEGAL to sell New York City a more reliable connection with a better for emergency services to use for emergency commications. Everyone has to have the same SLA that Netflix does. Remember, every packet has to be treated the same.
It truly is a non-trivial issue, and any simplistic bumper sticker policy will be wrong. Of course, restraint of trade is ALREADY unlawful, and throttling Netflix in an anti-competitive way may very well be restraint of trade. So maybe the proper response is to use our pre-existing laws which already account for ambiguity rather than enact new simplistic laws that are guaranteed to be harmful in many cases.
> everyone knows that net neutrality doesn't mean every connection has the same latency and bandwidth. It means you can't prioritize Company A's traffic over Company B's traffic
You and I know somewhat what a REASONABLE set of rules of rules might be, but GP is right as to the draft language. It basically said every packet has to be treated the same. As to company A and company B, if company A is a hospital and company B is a Nigerian prince, that's a difficult situation to write legislation for. Is it okay to deprioritize email from known spammers and allow the email from a search and rescue team to go through first? That's not allowed if the rule is "all users must be treated the same."
How about ads? On a slow wireless link, is it okay to deliver the text of a web page before the ads from DoubleClick ? They are both http web traffic.
Administrators making case-by-case decisions can make reasonable decisions in most cases. Coming up with simple rules deciding what admins must do in all cases for the next 20 years is much trickier, especially for bureaucrats who don't know the tech as well.
The Japanese already developed ship killing airplanes, making most large navy ships no better than targets. Unfortunately that won't change until you get the WW1-tactics trained admirals retired.
Ftfy to match what they said before WWII. Then some smart person put airplanes ON the ships. So it was mobile airfields off of Japan's coast vs fixed land-based airfields in Japan. No fighting in the U.S. once we went to war. If we could figure out a way to put missiles on ships, it could be our missiles on mobile platforms off their coast vs their missiles in their territory. Again keeping the fight several thousand miles away from the US.
I'm just curious if you have a citation for that. I would think that being unprepared and likely to fail would make someone nervous. I know that for me personally, I never worry about tests BECAUSE I know the material, so I know I'll pass. The main unknowns that affect my score are a) whether I mismark one question or two by hurrying through and b) how many wrong/nonsensical/ stupid questions there are. I don't worry precisely because I know the material or can figure it out, so I'll score at least 85%.
My last sentence does point out something about tests that is either a weakness or a strength, depending on how you look at it. While doing QA for courses where I don't know the subject, I can frequently barely pass by just applying intelligence/ common sense such as "any true/false question with the word 'always' or 'never' is probably false - most rules have exceptions ". That means I can get some questions without knowing the material ahead of time, BUT the same is true in real life - the same logic allows me to fix software written in a language I've never used, or evaluate products I'm unfamiliar with. In that way, the test measures how well I'd perform in that field apart from the knowledge I have in the field.
I've been jealous and I've been angry, but I've never killed anyone. Just because you have a feeling doesn't mean you have to do whatever comes into your head. I've thought "I'd like to fuck her" and "I'd like to punch him in the mouth", and chose to do neither. Reptiles may not have the ability to make such decisions, but mammals do.
> Everyone knows plenty of smart people who are "terrible at taking tests." Yet often these people are able to run circles around those good at taking tests when it comes to applied work in class.
At first I nodded my head in agreement, but then I started wondering- is that really true, or is that something said to avoid acknowledging that the person who consistently gets poor grades truly isn't that bright (or hasn't learned the material) . Thinking about people I know, there does seem to be a strong correlation between people who can successfully answer test questions and those who can answer "the mail server is down, what should we do?" Those who don't know what an exponential algorithm is vs a constant- time algorithm do in fact tend to use exponential algorithms.
That said, there are also a lot of poor test questions, mostly used by people who don't know how to use the discriminative index to find poor quality questions. Better education regarding how to maintain a good test bank would do quite a bit toward having better quality tests. There are also a few test-taking skills, but those can be easily learned by grade 5. Running statistics on our online courses, those who spend more time studying the material do better on the test, as do those who get started sooner rather than putting it off until the last minute. Common sense suggests that doing those two things will help you learn the material, so maybe the tests really do measure knowledge of the material, and some people truly don't learn it well. Maybe they're not "bad at tests", they're "bad at learning ".
There's a word for someone who has many opinions, all based on ignorance. Then there are informed opinions, based on learning the facts FIRST and learning on to explain and relate them.
> This prediction was made on nonsensical assumptions
The article said "IEA Executive Director Maria van der Hoeven stressed in a statement that her agency's two reports do not represent a forecast. "
The reports are not predictions of what will happen. They are a statement of "if bureaucrats wanted to increase the use of solar electric, here's how they could try, here's how much magical technology would be required, and here's what some of the (disastrous) consequences would be. It doesn't say that anyone will, can, or should attempt such a thing.
From the article. "IEA Executive Director Maria van der Hoeven stressed in a statement that her agency's two reports do not represent a forecast. "
The report said "if you wanted to try to have more solar, here's what you would try, and here's what the (devastating) consequences would be. They absolutely did not in any way say that would happen or should happen.
> That is where the company cut costs. That is where perverse incentives come in. The top honchos will have a policy directive that says "you must follow these procedures to handle viruses classified as ABC". Then do not hire enough people to enforce the policy
Biohazards are classified as level 1 through level 4. Polio is a level 2. Procedures are set by international agreements.
I see that while you know something about polio, you know nothing about how biohazards are classified or how the procedures are developed, much less how those standards were enforced at this lab. Based on your utter and complete lack of knowledge about the situation, you've completely made things up in your own head and you're ready to crucify the first person you see.
It appears that someone accidentally dumped the wrong bucket down the drain . From that, you infer:
> create incentive systems that encourage the violation of the same policies, and claim immunity,... Nominal financial penalties for those who were negligent are in order. But extraordinary penalties, amounting to all the pay and bonuses collected by the upper management in the last five or ten years should be assessed.
At this point, we have no idea what policies were in place, what the incentives were, or how upper management is going to respond. As far as we know, upper management could have had monthly safety audits, with large bonuses to staff every time they got a perfect score on the audit, and clear penalties for any infraction. As far as we know, while management was doing a superb job, one of the staff scientists came in with a horrible hangover and the first thing they did was clean up their work area by dumping out the "cleaning solution " they had been using the day before. Or maybe it was the opposite. We don't know. We really have no idea what happened at this point. In many workplaces I'm familiar with , the most likely cause would be that management chose policies that involved being so extremely careful that it got to be a pain in the ass, so staff started to ignore some of the policies because it was annoying to spend so much time on it, with 30 minutes of safety procedures required to perform a two minute task.
I've certainly noticed that. Midrange value-oriented places frequently include a continental breakfast too, whereas high-end places want you to buy their overpriced breakfast.
Sometimes I enjoy employing certain Priceline biding tactics to get a $200 room for $81, but other than the appearance the less-expensive places are often just as good or better.
Well that's an unexpectedly gracious and humble reply to a post which included the words "maybe you're stupid". I'm sorry for including those words. Maybe you're a gentleman and a scholar, fine sir.
Yes, if you're using a full 100GbE link to serve empty files nothing but headers, and you have an Atom CPU, that CPU will probably be the bottleneck. You'd want to upgrade that CPU.
On the other hand, if you're serving files, where the body of the response is much larger than the headers, you'd have only 100 Mbps of headers and your Atom CPU could keep up.
You're not under the impression that you are the first person to think about the potential tradeoff, are you? People much smarter than either of us calculated many different scenarios years ago, then actually tested it in the real world with a few billion requests between Android or Chrome and the various Google-owned sites. Based on analyzing a few billion data points in the real world, this is the optimal architecture for performance.
> ! You shouldn't be able to obtain a static number and then have the right to start charging money. . The story just highlights how wrong the current system is.
No, it doesn't. The story has absolutely nothing whatsoever to do with credit cards. That's a completely unrelated topic. Maybe credit card numbers are stupid, maybe Obama is stupid, maybe you're stupid. This story has absolutely nothing to do with any of those things.
That was at 1.2 Ghz. It's likely that most devices used in the next 15 years will run at 1.2 Ghz or faster.
> Did the net result of header decompression along with the easier parsing of the binary header take more or fewer CPU resources then the older uncompressed, ASCII header?
You keep conflating two completely separate things, but faster + faster = faster. Compressed is faster than uncompressed. Note that's the end of a sentence.
Also, and completely separately, binary is faster than ascii. Both are improvements, in terms of speed.
I just did a quick test and found that decompression runs at about 800 Mbps. Therefore, if the network connection is slower than 800 Mbps, it makes sense to transfer compressed data over the network, then decompress it.
The fact that binary data doesn't need the same parsing like ascii does is kind of an unrelated issue.
So if I type my comments into a web form, I'm not writing. If I use a quill pen and parchment, that's writing, I presume? What about a mechanical typewriter?
> You are being the useful idiot, allowing someone else to have their say over and over again. At the very least such astroturfing should be ignored by politicians.
It's called AGREEING. Often, my personal position on an issue is a expressed well by an EFF author, who also took the time to cite verifiable facts. It's not fraud or misrepresentation to say "I agree with this statement ". I think it's important for our representatives to know that position is held by many people, not just the one person who wrote down what we're all thinking. I might therefore sign the letter which represents my thoughts, while adding any additional comments that I wish to express.
Most often, I write my own separate comments rather than signing a letter I agree with, but that's just because I enjoy doing my own research and citing the sources that I think are best. If someone else agrees with me and wants to add their name to what I wrote that accurately represents their opinion on the matter.
That's kind of what you do. Why would you send a letter or email to only one of your two senators? Every online form I've ever seen lets you write to your two senators, your house rep, and frequently also the white house or other applicable office. They take your input on the online form and either print and mailvit or aggregate it and send the comments and signatures to the people's congressional representatives.
Slashdot Mirror
> Email is already a best-effort service without guaranteed delivery.
> If the S&R team actually needs a particular piece of information delivered immediately, they should choose a service that is optimized for that purpose.
You are mistaken. Most streaming protocols are best effort and layered on top of UDP, which is explicitly defined as a best-effort service. Packets may or may not arrive, and may arrive out of order.
SMTP is defined as a reliable service, and runs over TCP, also defined as a reliable service. Packets that don't make it the first time will be retransmitted and they will be placed in the correct order. Emails that cannot be delivered will be retried for several days. If, after five days or so, the email absolutely cannot be delivered, the sender will be notified of the failure.
So tell me again why an organization using a service that is defined to be reliable, for life-safety purposes, shouldn't be allowed to buy a slower but more reliable internet connection, one unlike Netflix , who wants lots of bandwidth but doesn't care if 1% of packets are dropped and lost forever? Why exactly do we all have to be forced to use a fast, unreliable network, (or all be forced to use a slow, reliable one), rather than being able to choose and buy the service with the SLA we want?
Your point is certainly right, and your explanation of why is pretty much on target. One very minor point:
> for VOIP it is essential that each packet be delivered, in order, and as quickly as possible. Delays of even 100ms create audible distortion. For email ...
For EMAIL, "it is essential that each packet be delivered". You wouldn't tolerate an email service that dropped words out of the middle of your email. For VOIP or other live streaming, you want the packets NOT be delivered if they can't be delivered in order and on time. A phone call that drops out for split second is slightly annoying, a call that changes "ex Stacy" into "sexy Stac" is intolerable.
While horrible latency, over 200 ms, is bad with VOIP, the main issue is that you want CONSISTENT latency. You don't want "each packet as quickly as possible", you want each packet delivered as quickly or slowly as the last packet. If we represent silence with an underscore, the difference can be seen as:
___________This_message_post_by_me. (High, but consistent latency)
Th__is__mess__agepo__ostb__yme. (Low, but inconsistent latency)
> I don't see how "the ISP should treat every packet the same" is unreasonable. ...
> If the user needs a stronger guarantee, they should get a better connection with a better SLA. None of this is illegal or unreasonable.
In your first sentence you made it ILLEGAL to sell New York City a more reliable connection with a better for emergency services to use for emergency commications. Everyone has to have the same SLA that Netflix does. Remember, every packet has to be treated the same.
It truly is a non-trivial issue, and any simplistic bumper sticker policy will be wrong. Of course, restraint of trade is ALREADY unlawful, and throttling Netflix in an anti-competitive way may very well be restraint of trade. So maybe the proper response is to use our pre-existing laws which already account for ambiguity rather than enact new simplistic laws that are guaranteed to be harmful in many cases.
> everyone knows that net neutrality doesn't mean every connection has the same latency and bandwidth. It means you can't prioritize Company A's traffic over Company B's traffic
You and I know somewhat what a REASONABLE set of rules of rules might be, but GP is right as to the draft language. It basically said every packet has to be treated the same. As to company A and company B, if company A is a hospital and company B is a Nigerian prince, that's a difficult situation to write legislation for. Is it okay to deprioritize email from known spammers and allow the email from a search and rescue team to go through first? That's not allowed if the rule is "all users must be treated the same."
How about ads? On a slow wireless link, is it okay to deliver the text of a web page before the ads from DoubleClick ? They are both http web traffic.
Administrators making case-by-case decisions can make reasonable decisions in most cases. Coming up with simple rules deciding what admins must do in all cases for the next 20 years is much trickier, especially for bureaucrats who don't know the tech as well.
The Japanese already developed ship killing airplanes, making most large navy ships no better than targets. Unfortunately that won't change until you get the WW1-tactics trained admirals retired.
Ftfy to match what they said before WWII. Then some smart person put airplanes ON the ships. So it was mobile airfields off of Japan's coast vs fixed land-based airfields in Japan. No fighting in the U.S. once we went to war. If we could figure out a way to put missiles on ships, it could be our missiles on mobile platforms off their coast vs their missiles in their territory. Again keeping the fight several thousand miles away from the US.
> smart people have a high rate of test anxiety
I'm just curious if you have a citation for that. I would think that being unprepared and likely to fail would make someone nervous. I know that for me personally, I never worry about tests BECAUSE I know the material, so I know I'll pass. The main unknowns that affect my score are a) whether I mismark one question or two by hurrying through and b) how many wrong/nonsensical/ stupid questions there are. I don't worry precisely because I know the material or can figure it out, so I'll score at least 85%.
My last sentence does point out something about tests that is either a weakness or a strength, depending on how you look at it. While doing QA for courses where I don't know the subject, I can frequently barely pass by just applying intelligence/ common sense such as "any true/false question with the word 'always' or 'never' is probably false - most rules have exceptions ". That means I can get some questions without knowing the material ahead of time, BUT the same is true in real life - the same logic allows me to fix software written in a language I've never used, or evaluate products I'm unfamiliar with. In that way, the test measures how well I'd perform in that field apart from the knowledge I have in the field.
> revoke the big cable companies' business licenses?
Or maybe just not renew their franchise, the monopoly those governments enforce.
I've been jealous and I've been angry, but I've never killed anyone. Just because you have a feeling doesn't mean you have to do whatever comes into your head. I've thought "I'd like to fuck her" and "I'd like to punch him in the mouth", and chose to do neither. Reptiles may not have the ability to make such decisions, but mammals do.
> Everyone knows plenty of smart people who are "terrible at taking tests." Yet often these people are able to run circles around those good at taking tests when it comes to applied work in class.
At first I nodded my head in agreement, but then I started wondering- is that really true, or is that something said to avoid acknowledging that the person who consistently gets poor grades truly isn't that bright (or hasn't learned the material) . Thinking about people I know, there does seem to be a strong correlation between people who can successfully answer test questions and those who can answer "the mail server is down, what should we do?" Those who don't know what an exponential algorithm is vs a constant- time algorithm do in fact tend to use exponential algorithms.
That said, there are also a lot of poor test questions, mostly used by people who don't know how to use the discriminative index to find poor quality questions. Better education regarding how to maintain a good test bank would do quite a bit toward having better quality tests. There are also a few test-taking skills, but those can be easily learned by grade 5. Running statistics on our online courses, those who spend more time studying the material do better on the test, as do those who get started sooner rather than putting it off until the last minute. Common sense suggests that doing those two things will help you learn the material, so maybe the tests really do measure knowledge of the material, and some people truly don't learn it well. Maybe they're not "bad at tests", they're "bad at learning ".
There's a word for someone who has many opinions, all based on ignorance. Then there are informed opinions, based on learning the facts FIRST and learning on to explain and relate them.
> This prediction was made on nonsensical assumptions
The article said "IEA Executive Director Maria van der Hoeven stressed in a statement that her agency's two reports do not represent a forecast. "
The reports are not predictions of what will happen. They are a statement of "if bureaucrats wanted to increase the use of solar electric, here's how they could try, here's how much magical technology would be required, and here's what some of the (disastrous) consequences would be. It doesn't say that anyone will, can, or should attempt such a thing.
From the article. "IEA Executive Director Maria van der Hoeven stressed in a statement that her agency's two reports do not represent a forecast. "
The report said "if you wanted to try to have more solar, here's what you would try, and here's what the (devastating) consequences would be. They absolutely did not in any way say that would happen or should happen.
> That is where the company cut costs. That is where perverse incentives come in. The top honchos will have a policy directive that says "you must follow these procedures to handle viruses classified as ABC". Then do not hire enough people to enforce the policy
Biohazards are classified as level 1 through level 4. Polio is a level 2. Procedures are set by international agreements.
I see that while you know something about polio, you know nothing about how biohazards are classified or how the procedures are developed, much less how those standards were enforced at this lab. Based on your utter and complete lack of knowledge about the situation, you've completely made things up in your own head and you're ready to crucify the first person you see.
It appears that someone accidentally dumped the wrong bucket down the drain .
From that, you infer:
> create incentive systems that encourage the violation of the same policies, and claim immunity, ... Nominal financial penalties for those who were negligent are in order. But extraordinary penalties, amounting to all the pay and bonuses collected by the upper management in the last five or ten years should be assessed.
At this point, we have no idea what policies were in place, what the incentives were, or how upper management is going to respond.
As far as we know, upper management could have had monthly safety audits, with large bonuses to staff every time they got a perfect score on the audit, and clear penalties for any infraction. As far as we know, while management was doing a superb job, one of the staff scientists came in with a horrible hangover and the first thing they did was clean up their work area by dumping out the "cleaning solution " they had been using the day before. Or maybe it was the opposite. We don't know. We really have no idea what happened at this point. In many workplaces I'm familiar with , the most likely cause would be that management chose policies that involved being so extremely careful that it got to be a pain in the ass, so staff started to ignore some of the policies because it was annoying to spend so much time on it, with 30 minutes of safety procedures required to perform a two minute task.
I've certainly noticed that. Midrange value-oriented places frequently include a continental breakfast too, whereas high-end places want you to buy their overpriced breakfast.
Sometimes I enjoy employing certain Priceline biding tactics to get a $200 room for $81, but other than the appearance the less-expensive places are often just as good or better.
Well that's an unexpectedly gracious and humble reply to a post which included the words "maybe you're stupid".
I'm sorry for including those words. Maybe you're a gentleman and a scholar, fine sir.
Yes, if you're using a full 100GbE link to serve empty files nothing but headers, and you have an Atom CPU, that CPU will probably be the bottleneck. You'd want to upgrade that CPU.
On the other hand, if you're serving files, where the body of the response is much larger than the headers, you'd have only 100 Mbps of headers and your Atom CPU could keep up.
You're not under the impression that you are the first person to think about the potential tradeoff, are you? People much smarter than either of us calculated many different scenarios years ago, then actually tested it in the real world with a few billion requests between Android or Chrome and the various Google-owned sites. Based on analyzing a few billion data points in the real world, this is the optimal architecture for performance.
> ! You shouldn't be able to obtain a static number and then have the right to start charging money. . The story just highlights how wrong the current system is.
No, it doesn't. The story has absolutely nothing whatsoever to do with credit cards. That's a completely unrelated topic. Maybe credit card numbers are stupid, maybe Obama is stupid, maybe you're stupid. This story has absolutely nothing to do with any of those things.
That was at 1.2 Ghz. It's likely that most devices used in the next 15 years will run at 1.2 Ghz or faster.
> Did the net result of header decompression along with the easier parsing of the binary header take more or fewer CPU resources then the older uncompressed, ASCII header?
You keep conflating two completely separate things, but faster + faster = faster. Compressed is faster than uncompressed. Note that's the end of a sentence.
Also, and completely separately, binary is faster than ascii. Both are improvements, in terms of speed.
> The web form has no area for typing your own comments.
False. You like to just completely make things up out of thin air, don't you? See that box labeled "Add your public comments" on their forms?:
http://action.americancommitme...
I just did a quick test and found that decompression runs at about 800 Mbps. Therefore, if the network connection is slower than 800 Mbps, it makes sense to transfer compressed data over the network, then decompress it.
The fact that binary data doesn't need the same parsing like ascii does is kind of an unrelated issue.
You forgot the first word of "embrace, extend, extinguish".
Take html for example:
1) We now have browser that renders html, just like Netscape. ... years later ...
2) "Best viewed in Internet Explorer"
3) Requires IE 4
Of fuck, Firefox is kicking our ass. Time to return to step 1.
> Why have Visa and Mastercard not changed their purchase validation system?
This story has nothing whatsoever to do with credit cards. The bad guys got a list of names and email addresses - a phone book.
> a web form is not you writing letters.
So if I type my comments into a web form, I'm not writing. If I use a quill pen and parchment, that's writing, I presume? What about a mechanical typewriter?
> You are being the useful idiot, allowing someone else to have their say over and over again. At the very least such astroturfing should be ignored by politicians.
It's called AGREEING. Often, my personal position on an issue is a expressed well by an EFF author, who also took the time to cite verifiable facts. It's not fraud or misrepresentation to say "I agree with this statement ". I think it's important for our representatives to know that position is held by many people, not just the one person who wrote down what we're all thinking. I might therefore sign the letter which represents my thoughts, while adding any additional comments that I wish to express.
Most often, I write my own separate comments rather than signing a letter I agree with, but that's just because I enjoy doing my own research and citing the sources that I think are best. If someone else agrees with me and wants to add their name to what I wrote that accurately represents their opinion on the matter.
That's kind of what you do. Why would you send a letter or email to only one of your two senators?
Every online form I've ever seen lets you write to your two senators, your house rep, and frequently also the white house or other applicable office. They take your input on the online form and either print and mailvit or aggregate it and send the comments and signatures to the people's congressional representatives.