Looks like the ISP has some problems with their bandwidth controls. That isn't a QoS issue at all.
Why are your neighbour allowed to download at such a high speed that it prevents you from watching video. The opposite question could of course also be, Why are you allowed to download video at such a high rate that it interfers with your opponents p2p traffic.
An ISP should ensure that it gives each custome adequate bandwidth. What the customer does with that bandwidth is their own business.
Know the counter argument from QoS proponents is of course, What if I want to use p2p, VoIP and play an online game at the same time. Without QoS this won't work.
This is true, but the only place where the QoS needs to take place is on the local router (or using software like cFosSpeed). The ISP doesn't have to be involved at all. I have personally cFosSpeed and played online games while using p2p, and it works just fine.
Religion is the act of following faith instead of reason.
There are two types of atheism, Strong and Weak. Among those I would only classify Strong atheism, the certainty that there is no god, as a religion.
Weak atheism however is simply the philosophical conclusion that in lack of evidence there is no reason to believe in a god, just as you don't believe in other things you don't have evidence for.
The difference between weak atheism and agnosticism is a matter of philosophy. Both views recognize that we can't know if god exists or not. The difference is in how to interpet that knowledge. I personally prefer weak atheism nowadays (used to be agnostic), because I found the agnostic worldview to be less productive.
I am from Sweden, so I was interested about the subject and searched on the Internet. I found a very good link on the European Commission website that contains some simplified information about the different justice systems in the european union. Below is the link to the english version of the chapter discussing evidence and proof in Sweden (There are also quick links to pages discussion the same for each and every country in the EU)
8. Can evidence that has been acquired in an unlawful manner be referred to as evidence?
The principle of admissibility of evidence means that there are only certain rare exceptions where it is forbidden to use certain types of evidence. That evidence has been acquired in an unlawful manner does not therefore, in principle, prevent the proof being referred to during the trial. This can, however, be of significance in the weighing of evidence.
Bittorrent does have authentication of content built in. One of the big advantages over FTP and HTTP btw. When I have recieved something via bittorrent, I am guaranteed to have recieved exactly what the publisher of the torrent file wanted me to recieve, and nothing else. As long as you trust the publisher of the original torrent file, you can trust the data your recieve.
Of course, if you use sites like the piratebay, you can't trust the publisher (because anyone can publish), so you can't be sure if what you download is fake or not.
You are correct that the GPL is a license and not a contract. However, that actually furthers my original point. From http://lwn.net/Articles/61292/ , quote by "Professor Moglen":
"The GPL, however, is a true copyright license: a unilateral permission, in which no obligations are reciprocally required by the licensor. Copyright holders of computer programs are given, by the Copyright Act, exclusive right to copy, modify and redistribute their programs. The GPL, reduced to its essence, says: 'You may copy, modify and redistribute this software, whether modified or unmodified, freely. But if you redistribute it, in modified or unmodified form, your permission extends only to distribution under the terms of this license. If you violate the terms of this license, all permission is withdrawn."
So the original copyright holder never has any obligation at all to the GPL. He can do whatever he want (as long as he is the sole owner of the original code).
If I distribute a program I have written myself and include the GPL license, but refuse to give out the source code, I am doing nothing wrong. I already have distribution rights to the binary, so I don't have to agree to the GPL. At the same time, anyone getting a hold of the binary, can't legally redistribute it because they don't have the source code, so they can't furfill the requirements needed to get a license for redistribution.
"I'm very curious to see if the GPL is even a legally valid document. How can you agree not to sue "somebody?" With commercial software there is at least an implication that the license has been granted to a specific customer, not "everybody in general.""
GPL is also only granted to a specfic person. To be specific, the one who is planning to redistribute the program. If you are just using the program, you aren't actually operating under the GPL license. This however doesn't really matter. Licenses are just a grant to the user so that he can do something he otherwise legally couldn't. Because it isn't a contract, there is no need to target it to specific individuals.
What I question more are EULAs. If they aren't contracts, but only licenses, their only power should be to allow the user to do something he couldn't otherwise do (in exchange for possible restrictions). Agreeing to an EULA should however not constitute a contract.
Even if I click the yes box in agreement to an EULA, it shouldn't affect me as long as I am only doing things that I can legally do without the EULA. It is only if I need to do something that is illegal, but granted by the EULA that it should start to affect me.
What I stated is correct as far as I can interpet the GPL. I think you are failing to see the the tree because of the forest.
The GPL doesn't say that the licensor has to provide you with the source code. It says that you have to provide the source code if you redistribute. It is a minor difference in wording, but legal it is a huge difference, shifting the burden from the licensor to the licensee. For most people it doesn't make any difference, because they are both licensors and licensees, but the original copyright holder is only a licensor.
The original copyright holder never has to agree to the GPL (as a licensee) because he has distribution rights anyway (as long as he wrote all the code himself). I don't think it is even legally possible, because it would constitute a contract with oneself. If someone else adds to the source code, it is a different case, because in that case the original copyright holder has to agree to the GPL to use the modified code.
AS sais above, the only time the original copyright holder (having written all the code himself) is affected by the license is as a licensor. In other words, when someone wants to redistribute the program/source code, and in that case the only thing the licenser does is giving up the exclusive copyright in exchange for any derivative being under the GPL and what that entails.
But the agreement the original copyright holder enters just states that he has no right to complain if the licensee modifies and redistributes the source code. It doesn't say anything about providing the source code in the first place.
The only one who has to provide source code is someone who licenses the code under the GPL, and the owner doesn't need a license because he already has the additional rights that the GPL grants, so there is no need to obtain a license from himself. He just have the obligation of not suing anyone who redistributes the source code that he never provided to them.
Of course, that only applies as long as all the source code is produced by the original copyright holder, and doesn't contain any GPL code licensed by someone else.
You will probably find opera 9.5 much better. I have been using opera for the last 8 years or so, and I agree that opera is often slower than firefox. (large slashdot pages is one example)
However, after downloading and testing opera 9.5 on a few of the troublesome sites, it is apparent that the problems have been fixed. The user experience now matches the theoretical benchmark values, because of performance improvements to a few key functions. (tables, javascript, images and interface threading).
ACDSee Classic version 2.43 is a very fast image viewer.
It is also renders directly onto the visible screen and doesn't lock the controls while doing so, which makes it the perfect viewer for quickly browsing images, especially when enlarged in full screen mode.
Also, by just pressing enter you switch between the image viewer and a simple but effective directory browser (that also has an image preview part).
The only features I miss are unicode support and rebinding the keyboard and viewing zip archives as an image directory. Oh, and it doesn't have png transparency support either, which doesn't matter to me much, because I mostly view jpgs.
If I could find an image viewer as fast as ACDSee Classic (including the ability to see partially rendered images) with any of the missing features I mentioned I would probably switch, but I havn't. And I have looked at a lot of viewers.
Just tested opera 9.5. Went directly to slashhdot, logged in to get my user settings (low threshhold, display full comments), and located the first article I could find with 500+ comments.
I am happy to say that opera has finally fixed the biggest flaw in their rendering engine. Not only is the rendering of slashdot much faster, it also no longer locks up the browser while it does it. (Most likely the table improvements you mentioned)
Tested a few other sites, and the overall user performance seems to have been improved. Both, with faster rendering/DOM manipulation/javascript and with better multithreading so you no longer have to wait for a page to complete loading, before using it. That is in my opinion far more important than artifical benchmarks.
Sure, the BSD developers may be hypocritical if they complain about someone taking their code without sharing back, but that doesn't exclude the GPL developers from also being hypocritical.
The complaint about the hypocrisy of some GPL developers is very much valid. A GPL developer simply can't integrate ANY non-GPL code into his project unless he dual licenses it back. (and the dual licensing back may also be hypocritical depending on how fundamentalistically he follows GPL)
The problem is that as soon as a GPL developer takes any non-GPL code, without licensing it back, he is breaking the fundamental moral GPL rule that all code should be shared back. The only way to avoid this is by not touching the non-GPL code in the first place.
The swedish equivalent to the SAT uses such questions for the math part. Here is a practice example:
A,B and C are three different positive whole numbers. What is the mean average of the numbers?
(1) The sum of the two largest numbers are 130. (2) The sum of the two smallest numbers are 110.
Enough information to solve the problem can be obtained from (a) 1 but not 2 (b) 2 but not 1 (c) 1 together with 2 (d) 1 or 2 separatly (e) Not from the two statements
I called myself agnostic until someone posed the question if I was agnostic in my the belief in Santa Claus, the tooth fairy and the flying spaghetti monster.
At that point I realized that calling oneself agnostic because there is a very tiny possibility that a god exists is just playing with the definition of the word agnostic. For all practical purposes I am an atheist.
While I understand why the parent is understandably marked as a troll, he has a point. Why are ISPs allowed to exclusivly limit p2p protocols. I wouldn't mind as much if they throttled bandwidth for those accounts that use lots of bandwidth. That would simply be fair and sound practice to allow everyone to use their share of the bandwidth.
Throttling specific protocols however is just the first step to allowing the ISPs to control what is and what is not allowed on the internet.
Why should false advertising laws care what the contract says? Of course in the us where it is legal for a "news"/propaganda agency to lie, I doubt that false advertising laws have much power.
And I am in favor of throttling email or banning it completly, because it is just a bunch of spam that takes up precious bandwidth. Oh, and ban VOIP while you are at it. If people want to phone each other, they should use the existing telephone system instead of wasting bandwidth.
Protocol throttling is nothing more hypocrisy coming from those that use other protocols. If they need to throttle, they should have the guts to throttle by bandwidth instead of protocol. It would still hurt those that use bittorrent, but those that downloading lots via other protocols would also feel it. Focusing on a single protocol simply isn't fair in any sense of the word.
The problem is that if you play the computer or even just the same opponent most of the time, you will learn to play moves that the opponent doesn't punish but other humans will. In other words, you will get better at beating the specific opponent at the cost of getting worse at go itself.
Btw, gnugo 2.7 is ancient. The latest stable is 3.6 and the development is 3.7.10. And those are beginning to get old also.
With 90+% of the raid paladins being specced holy for healing you are correct.
Of course with the above statement you can also conclude that paladins aren't hybrids. They may discuise themselves as hybrids at lower levels, but in the raid environment paladins are pretty much the most pure healer class in the endgame.
The only time I have latency problems on my connection is when I am trying to do gaming at the same time as I am using p2p. (Others using up their bandwidth doesn't seem to affect me, probably because my ISP seems to shape total bandwidth of users -- which is the neutral way to shape)
Anyway, there are a couple of ways to make sure that my gaming experience is smooth.
1. The ISP prioritizes the gaming packets or depreoritizes other packets. The problem with this method is that the ISP gets to decide which data is important and which is not. As seen with the post I first responded to, ISPs simply can't be trusted with that responsibility.
2. I don't use up all my availible bandwidth on file sharing, and use speed limiting that is built into the file sharing program. This is an easy way to fix the problem which I usually use. By always keeping some spare bandwidth, latency sensitive services work as expected. The problem with this method is that I can't use all my bandwidth. Also, if the availible bandwidth varies too much, I have to keep changing the p2p speed limit.
3. Programs/Hardware requiring low latency should use a Packet Scheduler built into the Operating System/Local Router. That way the user has the power to decide what is important instead of the ISP.
Sounds like the real problem you are having is with your ISP and not with blizzard. Only evil packet (protocol) shaping would prevent you from downloading a mere 4 MB file in a few minutes, even if you aren't uploading anything. Fortunally blizzard has alternative ways of downloading patches, but it really shouldn't be nescessary. Bittorrent is not much different than a http/ftp download, except that it is also possible for clients to exchange parts of the file/files between themselves when the server is overloaded.
This is why real net neutrality is so important (and I am talking about real net neutrality, not the fake one that some are advocating that still allows packet shaping).
Except of course that GPL software isn't licensed. It is free to download and use at your leasure. You can of course not redistribute it, but that is because of copyright law, not because of GPL.
GPL does give you the possibility to make a deal (contract) that gives you some rights of redistribution and modification. This deal is a mutually beneficial contract where both parties gain something.
It is a common misconception that only weather is a chaotic system, and that climate isn't because it is an average of the weather. That is however an incorrect conclusion. All it means is that climate and weather doesn't share the same chaotic system. Climate instead has its own chaotic system.
As for an IPCC link. Directly from Third Assessment Report (TAR) "Climate Change 2001: The Scientific Basis" (Fourth assessment report is due out this year):
I have no problem with bandwidth shaping, which makes sure that everyone gets their fair share of the cake (bandwidth). I do however have a big problem with packet shaping. Packet shaping gives more of the cake to those who use "approved protocols", or talk with "approved services".
Of course someone will undoubtably think to mention VoiP to which I will pre-emptivly respond client QoS-shaping. It is a simple solution that has existed for ages and doesn't involve the ISP deciding what is good and bad content.
I have never understood the density argument. A lower population density should mostly affect the base cost of connecting someone to the internet, as well as the connection quality.
Packet shaping however is about ISPs don't wanting to upgrade upstream routers to handle the higher bandwidth (Atleast I assume ISPs were intelligent enough to dig enough bandwidth into the ground the first time they dug. If they weren't, they deserve to die a painful death). The cost difference of upgrading routers between a low population and a high population area should be far less than the cost difference of digging in the first place.
Of course, I am not a network expert, so if someone wants to correct me, please do.
Claiming that opera is faster than firefox on slashdot? Opera completly sucks at rendering slashdot quickly. What takes 2 seconds for firefox to render can take opera half a minute, if not more (This is with big comment pages, and with a low threshold on displaying the full comments. Add moderation drop down boxes for extra slow speed). When I browse slashdot, I usually open all the interesting tabs as quickly as possible and then do something else for a couple of minutes.
I still mainly use opera because firefox doesn't support full page zoom, but in my experience firefox is much more stable when it comes to speed. Opera may be the fastest on some pages, but when opera is slow it is really slow.
Good post, with believable numbers. I just wanted to ad my slightly more pessemistic view of advertising in general.
Important to remember is that the numbers you are using estimates the business cost for advertising, which is completly different from the society cost.
First of all, the 14% you estimated is larger than the real cost to society, because a part of it is simple money transfers. This is best seen in TV Ads where the majority of money goes to pay for the construction of tv shows and not to pay for the administration and airing of the ads. The opposite type of ad would be handing out flyers. With that kind of advertising, the business and society cost nearly coincide.
It is hard to estimate the exact society cost that comes from administrating and distributing ads, but I would guess at somewhere between 5 and 10 percent (which is a bit lower than 14%, but still represents maybe half an hour of your workday)
That is however only the direct cost of creating advertising. The big part of society cost is still unmentioned, and that is the cost of people having to/being forced to view ads. 2-3 hours of tv per day (average american from what I understand) equals 34-51 minutes of advertising. Some of it may be used for bathroom breaks but that is non-continous time so it has limited utility. Having to spend time separating ads from useful information when searching on the internet is another time spender.
There is also the big and difficult to measure fact that "disinformation" (or whatever you would like to call it) in advertising decreases the ability for consumers to make informed decisions, and as anyone who has read even the basics about market economy should know, the key to a well functioning market driven society is informed consumers.
All in all, the society cost of advertising is pretty big. A part of it is probably unavoidable though. Communication is needed to make informed decisions and where there is communication there will be advertising or the likes of it.
Slashdot Mirror
Looks like the ISP has some problems with their bandwidth controls. That isn't a QoS issue at all.
Why are your neighbour allowed to download at such a high speed that it prevents you from watching video. The opposite question could of course also be, Why are you allowed to download video at such a high rate that it interfers with your opponents p2p traffic.
An ISP should ensure that it gives each custome adequate bandwidth. What the customer does with that bandwidth is their own business.
Know the counter argument from QoS proponents is of course, What if I want to use p2p, VoIP and play an online game at the same time. Without QoS this won't work.
This is true, but the only place where the QoS needs to take place is on the local router (or using software like cFosSpeed). The ISP doesn't have to be involved at all. I have personally cFosSpeed and played online games while using p2p, and it works just fine.
Religion is the act of following faith instead of reason.
There are two types of atheism, Strong and Weak. Among those I would only classify Strong atheism, the certainty that there is no god, as a religion.
Weak atheism however is simply the philosophical conclusion that in lack of evidence there is no reason to believe in a god, just as you don't believe in other things you don't have evidence for.
The difference between weak atheism and agnosticism is a matter of philosophy. Both views recognize that we can't know if god exists or not. The difference is in how to interpet that knowledge. I personally prefer weak atheism nowadays (used to be agnostic), because I found the agnostic worldview to be less productive.
I am from Sweden, so I was interested about the subject and searched on the Internet. I found a very good link on the European Commission website that contains some simplified information about the different justice systems in the european union. Below is the link to the english version of the chapter discussing evidence and proof in Sweden (There are also quick links to pages discussion the same for each and every country in the EU)
http://ec.europa.eu/civiljustice/evidence/evidence_swe_en.htm/8. Can evidence that has been acquired in an unlawful manner be referred to as evidence?
The principle of admissibility of evidence means that there are only certain rare exceptions where it is forbidden to use certain types of evidence. That evidence has been acquired in an unlawful manner does not therefore, in principle, prevent the proof being referred to during the trial. This can, however, be of significance in the weighing of evidence.
Bittorrent does have authentication of content built in. One of the big advantages over FTP and HTTP btw. When I have recieved something via bittorrent, I am guaranteed to have recieved exactly what the publisher of the torrent file wanted me to recieve, and nothing else. As long as you trust the publisher of the original torrent file, you can trust the data your recieve.
Of course, if you use sites like the piratebay, you can't trust the publisher (because anyone can publish), so you can't be sure if what you download is fake or not.
You are correct that the GPL is a license and not a contract. However, that actually furthers my original point. From http://lwn.net/Articles/61292/ , quote by "Professor Moglen":
"The GPL, however, is a true copyright license: a unilateral permission, in which no obligations are reciprocally required by the licensor. Copyright holders of computer programs are given, by the Copyright Act, exclusive right to copy, modify and redistribute their programs. The GPL, reduced to its essence, says: 'You may copy, modify and redistribute this software, whether modified or unmodified, freely. But if you redistribute it, in modified or unmodified form, your permission extends only to distribution under the terms of this license. If you violate the terms of this license, all permission is withdrawn."
So the original copyright holder never has any obligation at all to the GPL. He can do whatever he want (as long as he is the sole owner of the original code).
If I distribute a program I have written myself and include the GPL license, but refuse to give out the source code, I am doing nothing wrong. I already have distribution rights to the binary, so I don't have to agree to the GPL. At the same time, anyone getting a hold of the binary, can't legally redistribute it because they don't have the source code, so they can't furfill the requirements needed to get a license for redistribution.
"I'm very curious to see if the GPL is even a legally valid document. How can you agree not to sue "somebody?" With commercial software there is at least an implication that the license has been granted to a specific customer, not "everybody in general.""
GPL is also only granted to a specfic person. To be specific, the one who is planning to redistribute the program. If you are just using the program, you aren't actually operating under the GPL license. This however doesn't really matter. Licenses are just a grant to the user so that he can do something he otherwise legally couldn't. Because it isn't a contract, there is no need to target it to specific individuals.
What I question more are EULAs. If they aren't contracts, but only licenses, their only power should be to allow the user to do something he couldn't otherwise do (in exchange for possible restrictions). Agreeing to an EULA should however not constitute a contract.
Even if I click the yes box in agreement to an EULA, it shouldn't affect me as long as I am only doing things that I can legally do without the EULA. It is only if I need to do something that is illegal, but granted by the EULA that it should start to affect me.
What I stated is correct as far as I can interpet the GPL. I think you are failing to see the the tree because of the forest.
The GPL doesn't say that the licensor has to provide you with the source code. It says that you have to provide the source code if you redistribute. It is a minor difference in wording, but legal it is a huge difference, shifting the burden from the licensor to the licensee. For most people it doesn't make any difference, because they are both licensors and licensees, but the original copyright holder is only a licensor.
The original copyright holder never has to agree to the GPL (as a licensee) because he has distribution rights anyway (as long as he wrote all the code himself). I don't think it is even legally possible, because it would constitute a contract with oneself. If someone else adds to the source code, it is a different case, because in that case the original copyright holder has to agree to the GPL to use the modified code.
AS sais above, the only time the original copyright holder (having written all the code himself) is affected by the license is as a licensor. In other words, when someone wants to redistribute the program/source code, and in that case the only thing the licenser does is giving up the exclusive copyright in exchange for any derivative being under the GPL and what that entails.
But the agreement the original copyright holder enters just states that he has no right to complain if the licensee modifies and redistributes the source code. It doesn't say anything about providing the source code in the first place.
The only one who has to provide source code is someone who licenses the code under the GPL, and the owner doesn't need a license because he already has the additional rights that the GPL grants, so there is no need to obtain a license from himself. He just have the obligation of not suing anyone who redistributes the source code that he never provided to them.
Of course, that only applies as long as all the source code is produced by the original copyright holder, and doesn't contain any GPL code licensed by someone else.
You will probably find opera 9.5 much better. I have been using opera for the last 8 years or so, and I agree that opera is often slower than firefox. (large slashdot pages is one example)
However, after downloading and testing opera 9.5 on a few of the troublesome sites, it is apparent that the problems have been fixed. The user experience now matches the theoretical benchmark values, because of performance improvements to a few key functions. (tables, javascript, images and interface threading).
ACDSee Classic version 2.43 is a very fast image viewer.
It is also renders directly onto the visible screen and doesn't lock the controls while doing so, which makes it the perfect viewer for quickly browsing images, especially when enlarged in full screen mode.
Also, by just pressing enter you switch between the image viewer and a simple but effective directory browser (that also has an image preview part).
The only features I miss are unicode support and rebinding the keyboard and viewing zip archives as an image directory. Oh, and it doesn't have png transparency support either, which doesn't matter to me much, because I mostly view jpgs.
If I could find an image viewer as fast as ACDSee Classic (including the ability to see partially rendered images) with any of the missing features I mentioned I would probably switch, but I havn't. And I have looked at a lot of viewers.
Just tested opera 9.5. Went directly to slashhdot, logged in to get my user settings (low threshhold, display full comments), and located the first article I could find with 500+ comments.
I am happy to say that opera has finally fixed the biggest flaw in their rendering engine. Not only is the rendering of slashdot much faster, it also no longer locks up the browser while it does it. (Most likely the table improvements you mentioned)
Tested a few other sites, and the overall user performance seems to have been improved. Both, with faster rendering/DOM manipulation/javascript and with better multithreading so you no longer have to wait for a page to complete loading, before using it. That is in my opinion far more important than artifical benchmarks.
Sure, the BSD developers may be hypocritical if they complain about someone taking their code without sharing back, but that doesn't exclude the GPL developers from also being hypocritical.
The complaint about the hypocrisy of some GPL developers is very much valid. A GPL developer simply can't integrate ANY non-GPL code into his project unless he dual licenses it back. (and the dual licensing back may also be hypocritical depending on how fundamentalistically he follows GPL)
The problem is that as soon as a GPL developer takes any non-GPL code, without licensing it back, he is breaking the fundamental moral GPL rule that all code should be shared back. The only way to avoid this is by not touching the non-GPL code in the first place.
The swedish equivalent to the SAT uses such questions for the math part. Here is a practice example:
A,B and C are three different positive whole numbers. What is the mean average of the numbers?
(1) The sum of the two largest numbers are 130.
(2) The sum of the two smallest numbers are 110.
Enough information to solve the problem can be obtained from
(a) 1 but not 2
(b) 2 but not 1
(c) 1 together with 2
(d) 1 or 2 separatly
(e) Not from the two statements
(The answer is e btw)
I called myself agnostic until someone posed the question if I was agnostic in my the belief in Santa Claus, the tooth fairy and the flying spaghetti monster.
At that point I realized that calling oneself agnostic because there is a very tiny possibility that a god exists is just playing with the definition of the word agnostic. For all practical purposes I am an atheist.
Othwerwise, I agree with what you said.
While I understand why the parent is understandably marked as a troll, he has a point. Why are ISPs allowed to exclusivly limit p2p protocols. I wouldn't mind as much if they throttled bandwidth for those accounts that use lots of bandwidth. That would simply be fair and sound practice to allow everyone to use their share of the bandwidth.
Throttling specific protocols however is just the first step to allowing the ISPs to control what is and what is not allowed on the internet.
Why should false advertising laws care what the contract says? Of course in the us where it is legal for a "news"/propaganda agency to lie, I doubt that false advertising laws have much power.
And I am in favor of throttling email or banning it completly, because it is just a bunch of spam that takes up precious bandwidth. Oh, and ban VOIP while you are at it. If people want to phone each other, they should use the existing telephone system instead of wasting bandwidth.
Protocol throttling is nothing more hypocrisy coming from those that use other protocols. If they need to throttle, they should have the guts to throttle by bandwidth instead of protocol. It would still hurt those that use bittorrent, but those that downloading lots via other protocols would also feel it. Focusing on a single protocol simply isn't fair in any sense of the word.
The problem is that if you play the computer or even just the same opponent most of the time, you will learn to play moves that the opponent doesn't punish but other humans will. In other words, you will get better at beating the specific opponent at the cost of getting worse at go itself.
Btw, gnugo 2.7 is ancient. The latest stable is 3.6 and the development is 3.7.10. And those are beginning to get old also.
With 90+% of the raid paladins being specced holy for healing you are correct.
Of course with the above statement you can also conclude that paladins aren't hybrids. They may discuise themselves as hybrids at lower levels, but in the raid environment paladins are pretty much the most pure healer class in the endgame.
The only time I have latency problems on my connection is when I am trying to do gaming at the same time as I am using p2p. (Others using up their bandwidth doesn't seem to affect me, probably because my ISP seems to shape total bandwidth of users -- which is the neutral way to shape)
Anyway, there are a couple of ways to make sure that my gaming experience is smooth.
1. The ISP prioritizes the gaming packets or depreoritizes other packets. The problem with this method is that the ISP gets to decide which data is important and which is not. As seen with the post I first responded to, ISPs simply can't be trusted with that responsibility.
2. I don't use up all my availible bandwidth on file sharing, and use speed limiting that is built into the file sharing program. This is an easy way to fix the problem which I usually use. By always keeping some spare bandwidth, latency sensitive services work as expected. The problem with this method is that I can't use all my bandwidth. Also, if the availible bandwidth varies too much, I have to keep changing the p2p speed limit.
3. Programs/Hardware requiring low latency should use a Packet Scheduler built into the Operating System/Local Router. That way the user has the power to decide what is important instead of the ISP.
Sounds like the real problem you are having is with your ISP and not with blizzard. Only evil packet (protocol) shaping would prevent you from downloading a mere 4 MB file in a few minutes, even if you aren't uploading anything. Fortunally blizzard has alternative ways of downloading patches, but it really shouldn't be nescessary. Bittorrent is not much different than a http/ftp download, except that it is also possible for clients to exchange parts of the file/files between themselves when the server is overloaded.
This is why real net neutrality is so important (and I am talking about real net neutrality, not the fake one that some are advocating that still allows packet shaping).
Except of course that GPL software isn't licensed. It is free to download and use at your leasure. You can of course not redistribute it, but that is because of copyright law, not because of GPL.
GPL does give you the possibility to make a deal (contract) that gives you some rights of redistribution and modification. This deal is a mutually beneficial contract where both parties gain something.
It is a common misconception that only weather is a chaotic system, and that climate isn't because it is an average of the weather. That is however an incorrect conclusion. All it means is that climate and weather doesn't share the same chaotic system. Climate instead has its own chaotic system.
As for an IPCC link. Directly from Third Assessment Report (TAR) "Climate Change 2001: The Scientific Basis" (Fourth assessment report is due out this year):
http://www.grida.no/climate/ipcc_tar/wg1/504.htm
It explicitly mentions the difficulty in long term prediction of climate, due to its chaotic components.
I have no problem with bandwidth shaping, which makes sure that everyone gets their fair share of the cake (bandwidth). I do however have a big problem with packet shaping. Packet shaping gives more of the cake to those who use "approved protocols", or talk with "approved services".
Of course someone will undoubtably think to mention VoiP to which I will pre-emptivly respond client QoS-shaping. It is a simple solution that has existed for ages and doesn't involve the ISP deciding what is good and bad content.
I have never understood the density argument. A lower population density should mostly affect the base cost of connecting someone to the internet, as well as the connection quality.
Packet shaping however is about ISPs don't wanting to upgrade upstream routers to handle the higher bandwidth (Atleast I assume ISPs were intelligent enough to dig enough bandwidth into the ground the first time they dug. If they weren't, they deserve to die a painful death). The cost difference of upgrading routers between a low population and a high population area should be far less than the cost difference of digging in the first place.
Of course, I am not a network expert, so if someone wants to correct me, please do.
Claiming that opera is faster than firefox on slashdot? Opera completly sucks at rendering slashdot quickly. What takes 2 seconds for firefox to render can take opera half a minute, if not more (This is with big comment pages, and with a low threshold on displaying the full comments. Add moderation drop down boxes for extra slow speed). When I browse slashdot, I usually open all the interesting tabs as quickly as possible and then do something else for a couple of minutes.
I still mainly use opera because firefox doesn't support full page zoom, but in my experience firefox is much more stable when it comes to speed. Opera may be the fastest on some pages, but when opera is slow it is really slow.
Good post, with believable numbers. I just wanted to ad my slightly more pessemistic view of advertising in general.
Important to remember is that the numbers you are using estimates the business cost for advertising, which is completly different from the society cost.
First of all, the 14% you estimated is larger than the real cost to society, because a part of it is simple money transfers. This is best seen in TV Ads where the majority of money goes to pay for the construction of tv shows and not to pay for the administration and airing of the ads. The opposite type of ad would be handing out flyers. With that kind of advertising, the business and society cost nearly coincide.
It is hard to estimate the exact society cost that comes from administrating and distributing ads, but I would guess at somewhere between 5 and 10 percent (which is a bit lower than 14%, but still represents maybe half an hour of your workday)
That is however only the direct cost of creating advertising. The big part of society cost is still unmentioned, and that is the cost of people having to/being forced to view ads. 2-3 hours of tv per day (average american from what I understand) equals 34-51 minutes of advertising. Some of it may be used for bathroom breaks but that is non-continous time so it has limited utility. Having to spend time separating ads from useful information when searching on the internet is another time spender.
There is also the big and difficult to measure fact that "disinformation" (or whatever you would like to call it) in advertising decreases the ability for consumers to make informed decisions, and as anyone who has read even the basics about market economy should know, the key to a well functioning market driven society is informed consumers.
All in all, the society cost of advertising is pretty big. A part of it is probably unavoidable though. Communication is needed to make informed decisions and where there is communication there will be advertising or the likes of it.