A high speed trader does not increase liquidity, because for a high speed trader to work it has to know that It can buy something now and sell it moments later. This means the initial seller and the final buyer already existed before the high speed trader got involved, they just hadn't found each other yet. The item being sold was already as liquid as it was going to get.
I hope you are being ironic but I can't tell. If you're serious then it requires rebuttal. If a museum doesn't have interesting artifacts, then they don't attract visitors. If they don't attract visitors they don't have admissions income (or in the case of free museums have a hard time justifying the public funding they receive). Without income, they can't acquire interesting artifacts. It is a catch 22. If museums had to be run as a business and pull themselves up by their bootstraps, we wouldn't have any museums. All the great museums owe their existence to gift or public grant: The Louvre, The British Museum, The Smithsonian, American Museum of Natural History.
If these items are currently NASA property then transferring an asset from one government body to another has zero cost and the museum should not have to pay to acquire them. If these are not NASA property then there are one of two possibilities. 1) They are stolen US Government property. 2) NASA was wrong to transfer them to private ownership in the first place.
Gee if only we had a government body charged with the preservation of important historical documents. Oh wait! We do! I don't understand why these items aren't going to the National Archives. Its not like they are gonna raise enough money for a rocket or anything. The Smithsonian Institution would be a better home than some private collection.
Item A is the only one that really matters. It doesn't matter if it is (C) made in the UK, by (D) EEA employees, if the end product is just an imitation of US culture. Item B is meant to prevent it from just being an non-diverse outsiders caricature of British culture, which is what a game about warm beer, bad cooking and drinking tea would be.
Your right that compressed air is a less energy efficient storage medium than Li-ion batteries, but only for the first couple of years. Li-ion battery storage capacity decreases at about 20% a year because of natural degradation. Consider the cost to frequently manufacture, replace and dispose of batteries compared to the wear cycles of a compressed air container which is probably measured in decades.
My point here is that the maintenance cost for compressed air energy storage is quite low compared to other options. You also have to consider the cost of making the storage devices. Steel tanks are mostly hollow and we are already really good at making them. We are good at recycling steel too. Air storage, unlike fuel cells or batteries options which consume lots of metals and require complex electronics to regulate, compressed air is extremely cheap and simple.
If our choice is cheap simple but supposedly inefficient storage of 50% via compressed air or storing 0% via other supposedly more efficient but unaffordable and unsustainable methods the choice is pretty simple.
Bad Science and Science that does not agree with the reader are not the same thing. Bad research is defined by inaccurate models, failure to consider important variables, or just plain bad math. Bad research is not defined by what its results show.
My wife's PhD thesis was a Meta-Analysis, and I helped her create some new tools for doing the math behind the analysis so I feel like I have a pretty good grasp on the topic. The process (greatly simplified) is this. Dig through hundreds of articles published in peer reviewed journals on the topic you are examining, and find as many as you can that test the specific theory you are studying. All the articles included in the meta-analysis must test the same theory. Next you need to reverse engineer the numbers reported in the article. This can be a bit tricky since each article may have reported their result using different statistical tests. Occasionally some articles don't have all the relevant numbers and you have to contact the author. Once you have all that data together the math is relatively straight forward.
Presuming that all the other articles that you feed into the process are based on high quality research, then a Meta-Analysis can give you an insight to the overall strength of the results of the theory being tested. As you might imagine this process can easily be a Garbage In Garbage Out sort of situation. The researcher performing the meta-analysis must have the ability to identify bad studies that overlooked key moderating variables, or were simply done poorly and remove these bad studies from their analysis. If you want to attack this meta-analysis, attack the articles it was based off of. A meta-analysis by itself is not 'conclusive' just because of the method it represents. The analysis itself must be performed on many many well done studies in order to have any credibility of its own.
If you are being cost aware, or just prefer to use public transportation, get an Oyster Card. An Oyster card is a prepaid fair card for the Underground and buses. It costs a £3 deposit, plus any fair you choose to put on it. The system automatically calculates the cheapest ticket combination for whatever travel you do. Therefore if you only use it for a single trip in day, it debits for that trip, however if you use it several times, it debits up to the cost of a day pass which is much cheaper than single tickets. When you leave you can return the card and get refunded the deposit and any fair on the card. If you are fly in and out of Heathrow Airport, there are kiosks in the airport for the cards. If you are flying into Gatwick, you can take the train from the airport to Victoria Station which will also have kiosks for the Oyster card.
Also don't bother using currency exchanges. No one ever seems to change enough, and if you change too much you can get burnt converting the extra back. You can use your bank's ATM card in the cash machines here. Don't take out too much cash, just a little pocket money for the occasional taxi or ice cream cone. For all other purchases I recommend using a credit card for purchases for the fraud protection (usual warnings about knowing your financial limits apply.) Some ATM, Debit, and Credit card providers block foreign transactions by default. You can usually contact them and explain you will be traveling to specific countries on specific days and will remove the block for you. This usually isn't a problem in the UK, but its worth checking anyway. Definitely check with your banks if you plan on making a side trip to France; everybody seems to block French transactions.
If you sum up the figures given in the article, it only accounts for 75.9% of the contributions. I am going to speculate that this missing quarter is contributed by many who contribute infrequently. IE, IT staff in companies that use Linux and find the occasional bug and submit a patch to correct it. If this speculation is correct, the largest group that contributes is 'Everyone Else'.
One of the best theories I've heard about declining bee populations is that humans have been selecting our crops for traits that we desire such as larger fruit, and may have inadvertently selected out traits that bees desire such as flower nectar. In this scenario pollinating our crops becomes a bigger job with ever smaller return on the work for the bees. I think a piece of information that might support this theory is to examine how wild bees near undeveloped areas have been affected. Presumably an undeveloped area would still contain wild flowering plants that would still have normal nectar levels making it easier for bees to survive in that area than near say a great big old corn field.
The US spends more than twice as much per capita on health care than the UK. Thats because in the US we pay the medical staff (Doctors, Nurses, Orderlies, Janitors, etc) _AND_ the insurance salesmen, claims adjusters, case management workers, collections departments, the CEOs of insurance companies, advertising for hard-ons-in-a-jar....
I would be interested to know just how much of each dollar spent goes to actual health care, and how much goes to all those other things.
When I first moved to the UK and found that all my choices for ISPs had a metered usage plan, I was against it at first. My major complaint was that I had no way to predict how much data transfer I was going to use, so I didn't know what tier I should sign up for. Now that I've been on such a system for a couple years now I really do say that its more fair. The provider I am with now (plus.net) has a pretty good system I set a fixed monthly cost. For each £ I prepay I get so many GB of transfer. If I go over limit, I can choose to have my speed capped at 128K (Still plenty for email and most surfing), or optionally choose to pay a per GB charge that is slightly more expensive than the prepaid rate. Additionally They make a distinction between peak and off peak hours. So only transfers during peak hours actually count to my monthly transfer. The result is that I've learned to schedule my large downloads into Off Peak Hours. I have a had a few months where my home transfer was nearly 100 GB. However 80+ GB of that was Off peak usage which I did not pay for directly. Whats the result of all this? My ISP gets to manage their network performance during peak hours so all users have a pleasant experience. I still get big downloads, and I pay whats fair for what I use. The limits on my account are clearly defined. There is no mysterious 'use too much and we'll cut you off' amount.
I am very happy with this system, but to be clear, the reason why I am happy with this system is my ISP has provided choices. If Time Warner fails to provide similar choice then it will be awful.
I had, er, still have but no longer use, a laptop that had a Stop Security Plate on it. The basic idea is that it has contact information on it that anyone can call and verify the ownership of the property. If the plate is removed, it leaves a permanent mark on the property indicating it is stolen property. The nice thing about this product is that it is not software based, so it will work on anything with a large enough flat spot to glue the plate to.
They also gave out free (although horribly admin-locked) laptops to students.
No, Not free. There was a $600 a semester line item on my bill over 4 semesters. Students are buying a laptop for $2400. Oh, and if you drop out after the third semester, you had to pay the last $600 or give it back. The school doesn't pay anything for the laptops. The cost goes right to the tuition.
On the otherhand, in fall 2002 when I was issued my Thinkpad T-30, there was no more powerful laptop on the market, and $2400 was slightly below market prices for that particular piece of equipment.
Some of the other commenters have mentioned that the school likes to be bleeding edge and its true. I went there for a two year stint from fall of 02 to spring of 04. They hit a lot of firsts. First school with a mandatory laptop program (you could not enroll in a CIS major without buying or providing a laptop.) First school with campus wide wireless. Yes you could get a signal on any part of school property (Even out in the equestrian program's barns.) The only trouble with the original wireless networks is that because they adopted so early, the existing network was 802.11a. As many of you may know, its getting harder and harder to find and support 802.11a hardware.
Additionally they removed all the copper Ethernet from the dorms so using the Internet from the dorms was horrible. There really was not enough bandwidth to go around, and lots of concrete and metal furniture didn't help either. This was also at the time when p2p was really taking off and the network had never been built to expect that kind of traffic. To further mess things up, they removed all the pots telephone lines from the dorms and issued every student a cell phone. They got into a deal with Nextel that put a tower on campus, and created their own mini-cell network. Seemed like a good idea until everyone discovered push-to-talk. There were more phone's chirping than birds. And if you think Cell phones in the movies are bad, cell phones in the classroom are worse.
So anyway while it may seem like they are blazing forward, this is really just a much needed upgrade from an earlier deployment. Most of the students wanted these kinds of upgrades while I was still there. Really all they needed was more access points in the dorms, but I understand that there are only so many can be crammed together before they run all over each other.
It may sound like a rant against the school, but I really enjoyed my time there, Mainly because I commuted from (sorta) nearby Syracuse.
I was in Birmingham UK yesterday and saw a street warden with a head mounted camera. So this is already happening. Additionally, many police jurisdictions in the US already use similar technology like a camera with a wide angle lens attached to their vest.
With 2700+ servers they have a hard time handling more than 10k users? Less than 4 users per server is tough enough?
The problem is that the world is Zone Based, meaning each server is responsible for a equal size geographic portion of the world. The result is that processing power is spread evenly over the whole world. The problem is that people like to congregate causing some geographic areas to have more players, and other servers to have none. Where you have more players, you have more work for the server causing everything on that server to slow down. So the result is that the places players most want to be are also the places with the greatest lag. The unfortunate result is that many players have a negative experience right away.
Really, the whole server architecture needs to be reworked to behave more like a proper cluster, but that is too large of a change to ever consider implementing without starting over from scratch.
The add that was served up with this article; and I don't even need the money.
Paid Medical Research Get Paid to Participate in Medical Research and Mystery Shopping www.______.com Medical Research Trials Search & apply to medical research trials (paid & patient) listed here www.______.co.uk BioPharm Insight(TM) Drug Pipeline, Trials, Deals, Sales Companies and Project Managers www.______.com
Well if we consider that a dns nameat most can be a 255 byte string (not sure how/if unicode affects this.) the IP itself is another 4 bytes, but lets say 6 to support IPv6. Now we will also assume that to make the record searches fast we will also have a 32bit(4byte) hash on each name since matching strings is horribly slow. (you take the string you want to find, and create a hash which is a 32 bit number. You then match that number in the records, then you only compare strings on matching hashes in case two strings produce the same hash. If speed is your game, and there are many strings to match, you will do this.) so each element in the array of records is 265 bytes. So:
1024x1024x1024 = 1,073,741,824 = 1 GB 1,073,741,825 bytes / 265 bytes per record = 4,051,855 records per GB
This month, netcraft found 88,166,395 host names. so at 4M records per GB, you could cache every one of those host names in 22GB. So if we assume the system has 32GB of ram, and nearly all of that is devoted to RAM cached DNS entries, then you could store about 129,659,360 entries. Now at 3.25% growth in names per month. In 10 months you would be pushing the limit for caching all netcraft names.
So assuming that a server wouldn't need to cache every name in ram, 32GB is plenty of ram to cache the majority of names for the next 2 years By which time machines with 64 GB of ram will become commonplace.
Here is how the faster claim works. Say there is a 150ms round trip between you and your ISP's name server. You computer requests the IP for www.slashdot.org. If you are lucky then www.slashdot.org is in the name server's RAM cache, and you get a fast response in just a little over 150ms. If not (and for the majority of websites, its not) then the name server has to search its disk cache (this is where it is most likely to be. If its still not found, then your ISP's server has to look up slashdot.org with the root servers, and get the name server for that domain, and next it has query the dns server for slashdot.org to find the machine named www. each of these taking more time.
I presume what they do is have machines with loads of RAM (how many dns entries could you keep in say 4GB anyway?) and try to serve as many requests as possible from a RAM cache rather than disk cache. Thats my guess anyway.
What is wrong with telnet? Just because it isn't "real WWW" doesn't mean it is bad. If it works then why fix it?:)
Mainly, its an implemetation problem. z-modem over telnet inside SSL isn't widely supported as other protocols. Sure it is possible to get all sorts of librarys and components and tie them together, but you have to do all sorts of glue work, and it doesn't always work. You can litterly spend weeks trying to hammer a component onto another only to find out that the internal implementation of z-modem doesn't properly escape control characters so it will never reliably work over telnet, despite the fact that z-modem was designed to run on unreliable analog connections. You could also find out that the IRS says in their documentation that escaping control characters is required, but their own system is sending a z-modem handshake that instructs the connecting program to not escape control characters. This causes all sorts of baldness and sterility in your average programmer.
No, the new way is much better. The IRS publishes a WSDL, and I just load that into my favorite IDE, and ta-da I can send and recieve data with the IRS as what appears to be a native function call. It doesn't matter if I am using Java, Delphi, one of the Many.Net languages, Python or a host of other languages, it just works.
But of cause it doesn't really matter if the database is easily hack-able.
Your right on here. The point was that the IRS system was not built with security in mind. The new system, MEF, is being designed with security in mind. If they have to do a complete rebuild to secure it, they might as well do it in a way that reduces implemetation and maintenance costs for those who must use the system.
I work for a company that creates electronic filing software for the IRS, and I work with them on a regular basis. While Electronic filing has really only been popular the last few years its history goes back a very long time (in computer years). For example, currently to file a form 1040 electronically, it gets formatted in custom text format, attached to a whole bunch of other forms, gets all sorts of headers and summary information tacked on. It gets gzipped, then pushed through a z-modem connection over a telnet session, inside of an SSL connection. Why? Because it evolved that way. There was a time when electronic filing meant putting magnetic media in the mail. So the file formats go way back and are all fucked up because they are constantly updating the forms in respons to legislation. when they stopped with the magnetic media and started using modems, the whole thing was run like a BBS, so ta-da z-modem. When the bbs system was moved to the internet, it became telnet. Then they said oh shit its on the internet, we need encryption, so they moved that into an SSL connection.
Case in point the whole system is fucked up because its doing things it was never designed to do. So now we introduce Modernized E-File. MEF is basically the IRS rebuilding its entire system from the ground up. File formats are getting moved to XML, the network connections are moving to SOAP, and all sorts of other cool stuff.
Given the amound of stuff thats going on right now I would expect them to be scored poorly because basically the existing system is held together with duct tape while the new system is being built, and the new system probably wasn't considered in the score since its not completly up and running yet.
Slashdot Mirror
A high speed trader does not increase liquidity, because for a high speed trader to work it has to know that It can buy something now and sell it moments later. This means the initial seller and the final buyer already existed before the high speed trader got involved, they just hadn't found each other yet. The item being sold was already as liquid as it was going to get.
I hope you are being ironic but I can't tell. If you're serious then it requires rebuttal. If a museum doesn't have interesting artifacts, then they don't attract visitors. If they don't attract visitors they don't have admissions income (or in the case of free museums have a hard time justifying the public funding they receive). Without income, they can't acquire interesting artifacts. It is a catch 22. If museums had to be run as a business and pull themselves up by their bootstraps, we wouldn't have any museums. All the great museums owe their existence to gift or public grant: The Louvre, The British Museum, The Smithsonian, American Museum of Natural History.
If these items are currently NASA property then transferring an asset from one government body to another has zero cost and the museum should not have to pay to acquire them. If these are not NASA property then there are one of two possibilities. 1) They are stolen US Government property. 2) NASA was wrong to transfer them to private ownership in the first place.
Gee if only we had a government body charged with the preservation of important historical documents. Oh wait! We do! I don't understand why these items aren't going to the National Archives. Its not like they are gonna raise enough money for a rocket or anything. The Smithsonian Institution would be a better home than some private collection.
Item A is the only one that really matters. It doesn't matter if it is (C) made in the UK, by (D) EEA employees, if the end product is just an imitation of US culture. Item B is meant to prevent it from just being an non-diverse outsiders caricature of British culture, which is what a game about warm beer, bad cooking and drinking tea would be.
Shalshdotti ite Domum
Now, write it out a hundred times. Hail Taco ! And if it's not done by sunrise, I'll cut your balls off.
Your right that compressed air is a less energy efficient storage medium than Li-ion batteries, but only for the first couple of years. Li-ion battery storage capacity decreases at about 20% a year because of natural degradation. Consider the cost to frequently manufacture, replace and dispose of batteries compared to the wear cycles of a compressed air container which is probably measured in decades.
My point here is that the maintenance cost for compressed air energy storage is quite low compared to other options. You also have to consider the cost of making the storage devices. Steel tanks are mostly hollow and we are already really good at making them. We are good at recycling steel too. Air storage, unlike fuel cells or batteries options which consume lots of metals and require complex electronics to regulate, compressed air is extremely cheap and simple.
If our choice is cheap simple but supposedly inefficient storage of 50% via compressed air or storing 0% via other supposedly more efficient but unaffordable and unsustainable methods the choice is pretty simple.
Bad Science and Science that does not agree with the reader are not the same thing. Bad research is defined by inaccurate models, failure to consider important variables, or just plain bad math. Bad research is not defined by what its results show.
My wife's PhD thesis was a Meta-Analysis, and I helped her create some new tools for doing the math behind the analysis so I feel like I have a pretty good grasp on the topic. The process (greatly simplified) is this. Dig through hundreds of articles published in peer reviewed journals on the topic you are examining, and find as many as you can that test the specific theory you are studying. All the articles included in the meta-analysis must test the same theory. Next you need to reverse engineer the numbers reported in the article. This can be a bit tricky since each article may have reported their result using different statistical tests. Occasionally some articles don't have all the relevant numbers and you have to contact the author. Once you have all that data together the math is relatively straight forward.
Presuming that all the other articles that you feed into the process are based on high quality research, then a Meta-Analysis can give you an insight to the overall strength of the results of the theory being tested. As you might imagine this process can easily be a Garbage In Garbage Out sort of situation. The researcher performing the meta-analysis must have the ability to identify bad studies that overlooked key moderating variables, or were simply done poorly and remove these bad studies from their analysis. If you want to attack this meta-analysis, attack the articles it was based off of. A meta-analysis by itself is not 'conclusive' just because of the method it represents. The analysis itself must be performed on many many well done studies in order to have any credibility of its own.
If you are being cost aware, or just prefer to use public transportation, get an Oyster Card. An Oyster card is a prepaid fair card for the Underground and buses. It costs a £3 deposit, plus any fair you choose to put on it. The system automatically calculates the cheapest ticket combination for whatever travel you do. Therefore if you only use it for a single trip in day, it debits for that trip, however if you use it several times, it debits up to the cost of a day pass which is much cheaper than single tickets. When you leave you can return the card and get refunded the deposit and any fair on the card. If you are fly in and out of Heathrow Airport, there are kiosks in the airport for the cards. If you are flying into Gatwick, you can take the train from the airport to Victoria Station which will also have kiosks for the Oyster card.
Also don't bother using currency exchanges. No one ever seems to change enough, and if you change too much you can get burnt converting the extra back. You can use your bank's ATM card in the cash machines here. Don't take out too much cash, just a little pocket money for the occasional taxi or ice cream cone. For all other purchases I recommend using a credit card for purchases for the fraud protection (usual warnings about knowing your financial limits apply.) Some ATM, Debit, and Credit card providers block foreign transactions by default. You can usually contact them and explain you will be traveling to specific countries on specific days and will remove the block for you. This usually isn't a problem in the UK, but its worth checking anyway. Definitely check with your banks if you plan on making a side trip to France; everybody seems to block French transactions.
If you sum up the figures given in the article, it only accounts for 75.9% of the contributions. I am going to speculate that this missing quarter is contributed by many who contribute infrequently. IE, IT staff in companies that use Linux and find the occasional bug and submit a patch to correct it. If this speculation is correct, the largest group that contributes is 'Everyone Else'.
One of the best theories I've heard about declining bee populations is that humans have been selecting our crops for traits that we desire such as larger fruit, and may have inadvertently selected out traits that bees desire such as flower nectar. In this scenario pollinating our crops becomes a bigger job with ever smaller return on the work for the bees. I think a piece of information that might support this theory is to examine how wild bees near undeveloped areas have been affected. Presumably an undeveloped area would still contain wild flowering plants that would still have normal nectar levels making it easier for bees to survive in that area than near say a great big old corn field.
I would be interested to know just how much of each dollar spent goes to actual health care, and how much goes to all those other things.
When I first moved to the UK and found that all my choices for ISPs had a metered usage plan, I was against it at first. My major complaint was that I had no way to predict how much data transfer I was going to use, so I didn't know what tier I should sign up for. Now that I've been on such a system for a couple years now I really do say that its more fair. The provider I am with now (plus.net) has a pretty good system I set a fixed monthly cost. For each £ I prepay I get so many GB of transfer. If I go over limit, I can choose to have my speed capped at 128K (Still plenty for email and most surfing), or optionally choose to pay a per GB charge that is slightly more expensive than the prepaid rate. Additionally They make a distinction between peak and off peak hours. So only transfers during peak hours actually count to my monthly transfer. The result is that I've learned to schedule my large downloads into Off Peak Hours. I have a had a few months where my home transfer was nearly 100 GB. However 80+ GB of that was Off peak usage which I did not pay for directly. Whats the result of all this? My ISP gets to manage their network performance during peak hours so all users have a pleasant experience. I still get big downloads, and I pay whats fair for what I use. The limits on my account are clearly defined. There is no mysterious 'use too much and we'll cut you off' amount.
I am very happy with this system, but to be clear, the reason why I am happy with this system is my ISP has provided choices. If Time Warner fails to provide similar choice then it will be awful.
I had, er, still have but no longer use, a laptop that had a Stop Security Plate on it. The basic idea is that it has contact information on it that anyone can call and verify the ownership of the property. If the plate is removed, it leaves a permanent mark on the property indicating it is stolen property. The nice thing about this product is that it is not software based, so it will work on anything with a large enough flat spot to glue the plate to.
No, Not free. There was a $600 a semester line item on my bill over 4 semesters. Students are buying a laptop for $2400. Oh, and if you drop out after the third semester, you had to pay the last $600 or give it back. The school doesn't pay anything for the laptops. The cost goes right to the tuition.
On the otherhand, in fall 2002 when I was issued my Thinkpad T-30, there was no more powerful laptop on the market, and $2400 was slightly below market prices for that particular piece of equipment.
Some of the other commenters have mentioned that the school likes to be bleeding edge and its true. I went there for a two year stint from fall of 02 to spring of 04. They hit a lot of firsts. First school with a mandatory laptop program (you could not enroll in a CIS major without buying or providing a laptop.) First school with campus wide wireless. Yes you could get a signal on any part of school property (Even out in the equestrian program's barns.) The only trouble with the original wireless networks is that because they adopted so early, the existing network was 802.11a. As many of you may know, its getting harder and harder to find and support 802.11a hardware.
Additionally they removed all the copper Ethernet from the dorms so using the Internet from the dorms was horrible. There really was not enough bandwidth to go around, and lots of concrete and metal furniture didn't help either. This was also at the time when p2p was really taking off and the network had never been built to expect that kind of traffic. To further mess things up, they removed all the pots telephone lines from the dorms and issued every student a cell phone. They got into a deal with Nextel that put a tower on campus, and created their own mini-cell network. Seemed like a good idea until everyone discovered push-to-talk. There were more phone's chirping than birds. And if you think Cell phones in the movies are bad, cell phones in the classroom are worse.
So anyway while it may seem like they are blazing forward, this is really just a much needed upgrade from an earlier deployment. Most of the students wanted these kinds of upgrades while I was still there. Really all they needed was more access points in the dorms, but I understand that there are only so many can be crammed together before they run all over each other.
It may sound like a rant against the school, but I really enjoyed my time there, Mainly because I commuted from (sorta) nearby Syracuse.
I was in Birmingham UK yesterday and saw a street warden with a head mounted camera. So this is already happening. Additionally, many police jurisdictions in the US already use similar technology like a camera with a wide angle lens attached to their vest.
The problem is that the world is Zone Based, meaning each server is responsible for a equal size geographic portion of the world. The result is that processing power is spread evenly over the whole world. The problem is that people like to congregate causing some geographic areas to have more players, and other servers to have none. Where you have more players, you have more work for the server causing everything on that server to slow down. So the result is that the places players most want to be are also the places with the greatest lag. The unfortunate result is that many players have a negative experience right away.
Really, the whole server architecture needs to be reworked to behave more like a proper cluster, but that is too large of a change to ever consider implementing without starting over from scratch.
You do not talk about Usenet.
We have an Anime Babe Case Mod, a Bikini Babe Case Mod and a woman's torso case mod.
I tried to find a case modeled after the male form to be a bit-equal opportunity, but Google failed me.
Well if we consider that a dns nameat most can be a 255 byte string (not sure how/if unicode affects this.) the IP itself is another 4 bytes, but lets say 6 to support IPv6. Now we will also assume that to make the record searches fast we will also have a 32bit(4byte) hash on each name since matching strings is horribly slow. (you take the string you want to find, and create a hash which is a 32 bit number. You then match that number in the records, then you only compare strings on matching hashes in case two strings produce the same hash. If speed is your game, and there are many strings to match, you will do this.) so each element in the array of records is 265 bytes. So:
1024x1024x1024 = 1,073,741,824 = 1 GB
1,073,741,825 bytes / 265 bytes per record = 4,051,855 records per GB
This month, netcraft found 88,166,395 host names. so at 4M records per GB, you could cache every one of those host names in 22GB. So if we assume the system has 32GB of ram, and nearly all of that is devoted to RAM cached DNS entries, then you could store about 129,659,360 entries. Now at 3.25% growth in names per month. In 10 months you would be pushing the limit for caching all netcraft names.
So assuming that a server wouldn't need to cache every name in ram, 32GB is plenty of ram to cache the majority of names for the next 2 years By which time machines with 64 GB of ram will become commonplace.
Here is how the faster claim works. Say there is a 150ms round trip between you and your ISP's name server. You computer requests the IP for www.slashdot.org. If you are lucky then www.slashdot.org is in the name server's RAM cache, and you get a fast response in just a little over 150ms. If not (and for the majority of websites, its not) then the name server has to search its disk cache (this is where it is most likely to be. If its still not found, then your ISP's server has to look up slashdot.org with the root servers, and get the name server for that domain, and next it has query the dns server for slashdot.org to find the machine named www. each of these taking more time.
I presume what they do is have machines with loads of RAM (how many dns entries could you keep in say 4GB anyway?) and try to serve as many requests as possible from a RAM cache rather than disk cache. Thats my guess anyway.
Mainly, its an implemetation problem. z-modem over telnet inside SSL isn't widely supported as other protocols. Sure it is possible to get all sorts of librarys and components and tie them together, but you have to do all sorts of glue work, and it doesn't always work. You can litterly spend weeks trying to hammer a component onto another only to find out that the internal implementation of z-modem doesn't properly escape control characters so it will never reliably work over telnet, despite the fact that z-modem was designed to run on unreliable analog connections. You could also find out that the IRS says in their documentation that escaping control characters is required, but their own system is sending a z-modem handshake that instructs the connecting program to not escape control characters. This causes all sorts of baldness and sterility in your average programmer.
No, the new way is much better. The IRS publishes a WSDL, and I just load that into my favorite IDE, and ta-da I can send and recieve data with the IRS as what appears to be a native function call. It doesn't matter if I am using Java, Delphi, one of the Many
Your right on here. The point was that the IRS system was not built with security in mind. The new system, MEF, is being designed with security in mind. If they have to do a complete rebuild to secure it, they might as well do it in a way that reduces implemetation and maintenance costs for those who must use the system.
I work for a company that creates electronic filing software for the IRS, and I work with them on a regular basis. While Electronic filing has really only been popular the last few years its history goes back a very long time (in computer years). For example, currently to file a form 1040 electronically, it gets formatted in custom text format, attached to a whole bunch of other forms, gets all sorts of headers and summary information tacked on. It gets gzipped, then pushed through a z-modem connection over a telnet session, inside of an SSL connection. Why? Because it evolved that way. There was a time when electronic filing meant putting magnetic media in the mail. So the file formats go way back and are all fucked up because they are constantly updating the forms in respons to legislation. when they stopped with the magnetic media and started using modems, the whole thing was run like a BBS, so ta-da z-modem. When the bbs system was moved to the internet, it became telnet. Then they said oh shit its on the internet, we need encryption, so they moved that into an SSL connection.
Case in point the whole system is fucked up because its doing things it was never designed to do. So now we introduce Modernized E-File. MEF is basically the IRS rebuilding its entire system from the ground up. File formats are getting moved to XML, the network connections are moving to SOAP, and all sorts of other cool stuff.
Given the amound of stuff thats going on right now I would expect them to be scored poorly because basically the existing system is held together with duct tape while the new system is being built, and the new system probably wasn't considered in the score since its not completly up and running yet.