No branching (you can only add comments to the end of a conversation, not reply to comments in the middle).
Is there locking for posting in-depth, correct, messages, or if you spend lots of time presenting a well thought out post do you just get bounced when somebody else has already replied while you were typing? You can't restrict branching in a high traffic forum without some sort of syncronization, and you can't allow syncronization to be open to untrusted users without denial of service. Sounds like a broken decision to me.
All of these are part of the design of the system, not missing features.
When did these become mutually exclusive? Just beacuse something is intentionally left out doesn't mean it's not missing. Wether the features being missing is a good thing or not is the only thing that can be up for discussion.
90 degrees is the range, not the offset. That means you would be able to see it correctly from 45 degrees off center in either direction. 180 degrees would be the optimal viewing range.
At least that's how display manufacutrers talk about it.
The problem isn't with the logic, it's with the music. In recent years (the last 50 or so of them) the music industry has reduced music down into small 3 and a half minute chunks. Gone are the days of albums where songs lead into each other or corelate, and entire albums were an artistic unit like a full length symphony. There's no such thing as an album anymore, they're just discs with a bunch of bite-sized songs on them related to each other only by artist and sized as pop radio advertisements. Once music is sold only as individual three and a half minute pieces we will have lost the artistic concept of the album and potentially the existance of career bands.
Go listen to Abbey Road, Dark Side of the Moon, A Love Supreme, or (gasp) almost any symphony from what the modern industry calls the "classical" genre, and you'll see that an 'album' doesn't have to be a collection of bad songs with a few hits thrown on there.
I can't do it, but I've seen 50 WPM using T9. I know people who chat on their phones constantly, and they can type faster on the keypad than on a full sized keyboard.
A horrible estimate of trying to read the text in this post - I'd say a 160 degree viewing angle - still being able to read the text.
Is that 160 degrees without color distortion? Being able to read it is one thing. Having all the colors look right is another. An OLED display shows colors correctly for almost any angle, were even the new dell LCDs can't do that past 90 or so degrees.
Linear means if you graph out the number of processors by the performance of the system it goes diagonally up and to the right, linearly.
This is the only thing you said in your comment that is correct.
If you do this on a Linux box it begins to flatline after you reach 4 processors.
Linux is better now then whenever you got the idea that it sucked with more than four processors. It used to suck. It's way better now. This is beside the point.
Basically, this means that you get close to 100% performance out of every processor, as long as you're running multi-threaded applications like Oracle or Java.
You don't know what you're talking about. Linear doesn't mean you get close to 100% performance, it means you gain the same amount of performance for each additional processor you add. It could be 10% of a single processor's worth. It could be 25%, it could be 90%, but saying it's linear does not mean it's good. What's the slope of that linear curve?
[06:50pm]zaphod-primary:~>mount /dev/sda2 on / type ext3 (rw,errors=remount-ro) proc on/proc type proc (rw) devpts on/dev/pts type devpts (rw,gid=5,mode=620) /dev/sda3 on/var type ext3 (rw) automount(pid371) on/distro type autofs (rw,fd=5,pgrp=371,minproto=2,maxproto=4) automoun t(pid387) on/homes type autofs (rw,fd=5,pgrp=387,minproto=2,maxproto=4) automoun t(pid402) on/auto type autofs (rw,fd=5,pgrp=402,minproto=2,maxproto=4) /dev/sdb 6 on/opt/cluster/mounts/homes type xfs (rw) /dev/sdb5 on/opt/cluster/mounts/nfs type xfs (rw) [06:51pm]zaphod-primary:~>uname -a Linux zaphod-primary 2.4.20-xfs-1.1 #2 SMP Thu Feb 6 17:08:35 EST 2003 i686 unknown
That was just run on our production NFS server cluster.
Hmm, FS version 1.1 running is the kernel's stable tree version 2.4.20. We've been running XFS since version 1.0 for linux was out over 2 years ago. You anti-linux zealots need to get your facts straight and admit when you're wrong.
You can do all this stuff. See the scsiinfo utility for diag and inquiry, scsistop and scsi-spin for start and halt, etc. They're in the scsitools packare in debian and they're written by Eric Youngdale.
Solutions for all of your issues are explained in the HOWTO.
Solaris scales linearly all the way up to 106 processors on an Enterprise 15000.
That statement is not impressive by itself. Linear scaleability is not nesicarily a good thing. For example, every processor added after the first one could increase performance by 1%. That would suck. How much performance is gained for each processor added? Does the performance gain going from one to two of from two to four processors suffer so that scaleability from 4-106 processors will be linear?
The problem is they also have to charge all the regulated fees and taxes like universal access fees etc that come with voice service. Plus they have to take the entire line so you CAN'T get mixed service under this scheme.
This doesn't seem like a problem to me. DSL works better on a dedicated line anyway. If I didn't have to have a shared line for my connection, my ILEC would get zero dollars for it. So, they have to collect the USF tax. Small price to pay to keep your business alive.
Besides, covad isn't the only company around. I get my DSL from worldcom. They definatly offer voice.:)
So, all they have to do is get a (single) voice T1, and offer a really incredibly shitty overpriced voice service that nobody will ever sign up for because it costs too much and there'd hardly ever be a circuit available. Case closed.
If skills in one single industry were the only thing that defined the strength on an economy I would completely agree with you.
Also keep in mind that the size of an economy has to be devided by the population to determin agragate wealth. 1.1 billion people is a lot of people. India's economy would have to be more than three times the size of the US economy for each indian citizen to be as "wealthy" as each US citizen.
Also, wealth or leck of wealth isn't necissarily the cause of or solution to hunger. Geography and politics have a lot to do with it. Poor countries aren't the only places where people starve, and some poor people have plenty to eat.
Of course if you take a narrow enough veiw, you can "prove" anything is anybodies fault.
The bigger problem might be locking the tree waiting for the test build to complete. If you have a 12 hour work day, a one hour build, and 30 developers, less then half your team will be able to commit code on any given day, or you'd have a hard time commiting because you wouldn't know that somebody already has a pending commit that's waiting on the build.
This is a bad thing to programatically enforce during the commit on large projects using CVS.
Interesting little fact.. 2.2 million cards is.33% of outstanding cards in the US. Yes, you read that right.. one third of one percent.
I don't know why you find that so surprising. In a country with 300 million people you'd expect there to be a significantly larger number of existing credit card number out there than 2.2 million. Especially considering that they're used on things like gift cards issued by Novus and on check cards.
Actually, I think these storys are auto posted based on when the software update window pops up on his mac. Invariably when I'm at work and see the software update window pop up there's a corresponding slashdot story 20 minutes later.
Apparently most people don't understand Prior Art
on
NCR Patents the Internet
·
· Score: 4, Interesting
Prior art does not invalidate a patent or make an invention unpatentable unless the prior art covers every single part of the new patent down to the tiniest details. You can patent something that's already patented with only a slight addition. You'll have to have a licence to use the other patent to implement the invention, but you can still get a patent that grants you a limited monopoly to just the new parts. Since the time this article was posted until now (when I've seen dozens of people claiming that these patents are invalid due to prior art) there is NO FUCKING WAY that people who were unaware of these patents before could have figured out wether these should be valid patents or not.
In summary: prior art doesn't mean you can't have a patent. It just means you have to list it in the "prior art" section of the patent and licence the old patent to implement your "improvement" to the old invention (assuming the original patent is still valid).
ORGanizations TLDs. Chartered, for non-profit organizations.
From RFC 1591:
ORG - This domain is intended as the miscellaneous TLD for organizations that didn't fit anywhere else. Some non-government organizations may fit here.
It doesn't say anything about non-profit in there.
The uses for the TLDs were specifically defined. They mean what the RFC says they mean and not what you decided they should mean.
What? There are no separate TLD providers?
The TLDs are not all administered by the same entity. ICANN is authorised to award the contracts to manage any particular TLD, but the contracts are presently held by a variety of companies. In fact, the number of companies controlling TLDs is increasing, since Verisign is loosing controll of both the.net and.org TLDs by 2006.
Before you go blaming people for problems, make sure the problems exist.
Google was a poor choice when picking a company to base these predictions on. Computing power is growing exponentially, but google's core business is indexing web pages, and without an expontial human population explosion the amount of availble web content to index will grow linearly. That means it won't be too long before google can purchase a cluster of low-end machines that will do the job, and won't require upgrading for the forseeable future. Processor technolgy won't be driven by these pre-existing tasks with zero or linear growth simply because in an insignificant amount of time processor technology will make these tasks trivial. Processor technology will have to be driven by new tasks that we'll think up as the power to perform them becomes available. Google came into existance when the low-end PC became sufficient to accomplish the task. What new technology companies will spring up when the low-end is 2^4 times as powerful as when google started? How about 2^8 times? 2^32times? Perhaps eventually the tasks that will use the new processing will be designed by the machines that consume the power in the first place. Perhaps such machines can design the next generation processors. Then we won't need some company to drive the technology forward. Won't that be interesting.
Maybe it's "If they exceed 10, then they have to buy more OS licenses, or better yet a server" WTF????
Heh. This weekend I was upgrading a hard drive in an NT 4.0 server from 8GB to 36GB. I added the new drive to the SCSI bus and booted up with my newly acquired Partition Magic boot disk purchased from CompUSA just hours before. I proceded to copy the partitions onto the new drive. I then rebooted to the new drive to make sure it worked before growing the partitions to fill the drive. I was surprised to find that when I tried to run Partition Magic again it informed me that the product was not compatible with the "Server Edition" of windows, and that I needed to buy Server Magic instead. Server Magic does the same things as Partition Magic, but costs $400 more. Obviously Partition Magic works just fine with NT server, because I had just used it when the first disk in the system was blank, they had just put a check in there such that if you had deep enough pockets to buy NT server you'd have to pay $400 more for their partition utility. I would have been less pissed off about it had they stated the limitation clearly on the outside of the box, or if Server Magic were not so rediculously priced and was actually available at the local store, but it still seems wrong somehow.
Anyway, in short, it's not just Microsoft who charges you more for the same code based on what you're going to use it for.
--
BTW, I got around the stupid limitation by pulling a SCSI disk from one of the workstations and sticking it on the SCSI bus with a lower ID than the new drive. Then I just made my changes to the NT server disk and put the workstation drive back. Not a very good checking system apparently.
He has newer writings if you got to his website. He talks about how the ability to scan the brain to better understand how it's working without negatively affecting the person being scanned is increasing exponentially, and that at aproximately the same time as we acquire enough computing power to emulate the brain, we'll also have the ability to map every cell and interconnection, and record every firing of every neuron.
This is, of course, assuming that the progress of technology continues at the current rate. As we know, past numbers are no guarantee of future performance. Either way, we'll figure it out eventually, wether it be in 40 years or 400...
Unsubstantiated bullshit. And for every advance in smart OCR you come up with, I can come up with 10 obscuring transformations that leave it readable to humans but garbage to a computer.
Clearly you haven't read their paper.
"There is no way to prove that a program cannot pass a test which a human can pass, since there is a known program - the human brain - which passes the test."
The creaters for CAPTCHA acknowledge that you can programatically defeat their creations, but that it is hard to do with high accuracy. My assertions are as follows:
It is significantly easier to achieve a low degree of accuracy than a high degree of accuracy, and a low degree of accuracy is all that is necissary to successfully scrape web content.
As time goes by it will become easier do defeat any given CAPTCHA
Eventually you will run out of ways to obfuscate information such that a human bran will be the only way to decipher it, because eventually we will be able to outdo our own brains with software. In fact Ray Kurzwiel thinks we will do that in the next 40 years. I don't know if it will be that soon, but eventually you'll run ot of ways to fool the software.
I challange you do come up with a credible argument to refute any of those points.
Slashdot Mirror
No branching (you can only add comments to the end of a conversation, not reply to comments in the middle).
Is there locking for posting in-depth, correct, messages, or if you spend lots of time presenting a well thought out post do you just get bounced when somebody else has already replied while you were typing? You can't restrict branching in a high traffic forum without some sort of syncronization, and you can't allow syncronization to be open to untrusted users without denial of service. Sounds like a broken decision to me.
All of these are part of the design of the system, not missing features.
When did these become mutually exclusive? Just beacuse something is intentionally left out doesn't mean it's not missing. Wether the features being missing is a good thing or not is the only thing that can be up for discussion.
90 degrees is the range, not the offset. That means you would be able to see it correctly from 45 degrees off center in either direction. 180 degrees would be the optimal viewing range.
At least that's how display manufacutrers talk about it.
The problem isn't with the logic, it's with the music. In recent years (the last 50 or so of them) the music industry has reduced music down into small 3 and a half minute chunks. Gone are the days of albums where songs lead into each other or corelate, and entire albums were an artistic unit like a full length symphony. There's no such thing as an album anymore, they're just discs with a bunch of bite-sized songs on them related to each other only by artist and sized as pop radio advertisements. Once music is sold only as individual three and a half minute pieces we will have lost the artistic concept of the album and potentially the existance of career bands.
Go listen to Abbey Road, Dark Side of the Moon, A Love Supreme, or (gasp) almost any symphony from what the modern industry calls the "classical" genre, and you'll see that an 'album' doesn't have to be a collection of bad songs with a few hits thrown on there.
I can't do it, but I've seen 50 WPM using T9. I know people who chat on their phones constantly, and they can type faster on the keypad than on a full sized keyboard.
A horrible estimate of trying to read the text in this post - I'd say a 160 degree viewing angle - still being able to read the text.
Is that 160 degrees without color distortion? Being able to read it is one thing. Having all the colors look right is another. An OLED display shows colors correctly for almost any angle, were even the new dell LCDs can't do that past 90 or so degrees.
Linear means if you graph out the number of processors by the performance of the system it goes diagonally up and to the right, linearly.
This is the only thing you said in your comment that is correct.
If you do this on a Linux box it begins to flatline after you reach 4 processors.
Linux is better now then whenever you got the idea that it sucked with more than four processors. It used to suck. It's way better now. This is beside the point.
Basically, this means that you get close to 100% performance out of every processor, as long as you're running multi-threaded applications like Oracle or Java.
You don't know what you're talking about. Linear doesn't mean you get close to 100% performance, it means you gain the same amount of performance for each additional processor you add. It could be 10% of a single processor's worth. It could be 25%, it could be 90%, but saying it's linear does not mean it's good. What's the slope of that linear curve?
Hmm, FS version 1.1 running is the kernel's stable tree version 2.4.20. We've been running XFS since version 1.0 for linux was out over 2 years ago. You anti-linux zealots need to get your facts straight and admit when you're wrong.
You can do all this stuff. See the scsiinfo utility for diag and inquiry, scsistop and scsi-spin for start and halt, etc. They're in the scsitools packare in debian and they're written by Eric Youngdale.
Solutions for all of your issues are explained in the HOWTO.
Solaris scales linearly all the way up to 106 processors on an Enterprise 15000.
That statement is not impressive by itself. Linear scaleability is not nesicarily a good thing. For example, every processor added after the first one could increase performance by 1%. That would suck. How much performance is gained for each processor added? Does the performance gain going from one to two of from two to four processors suffer so that scaleability from 4-106 processors will be linear?
You've got a configuration problem, not a missing feature.
/var/lib/nfs when rebooting and your problem will go away.
Don't trash the contents of
The problem is they also have to charge all the regulated fees and taxes like universal access fees etc that come with voice service. Plus they have to take the entire line so you CAN'T get mixed service under this scheme.
:)
This doesn't seem like a problem to me. DSL works better on a dedicated line anyway. If I didn't have to have a shared line for my connection, my ILEC would get zero dollars for it. So, they have to collect the USF tax. Small price to pay to keep your business alive.
Besides, covad isn't the only company around. I get my DSL from worldcom. They definatly offer voice.
So, all they have to do is get a (single) voice T1, and offer a really incredibly shitty overpriced voice service that nobody will ever sign up for because it costs too much and there'd hardly ever be a circuit available. Case closed.
since Sony was supposed to lose money on consoles this would make perfect sense
Maybe the problem was that sony actually makes money on the consoles...
Either way, look what you learned from speculation last time? Why are you speculating again?
If skills in one single industry were the only thing that defined the strength on an economy I would completely agree with you.
Also keep in mind that the size of an economy has to be devided by the population to determin agragate wealth. 1.1 billion people is a lot of people. India's economy would have to be more than three times the size of the US economy for each indian citizen to be as "wealthy" as each US citizen.
Also, wealth or leck of wealth isn't necissarily the cause of or solution to hunger. Geography and politics have a lot to do with it. Poor countries aren't the only places where people starve, and some poor people have plenty to eat.
Of course if you take a narrow enough veiw, you can "prove" anything is anybodies fault.
The bigger problem might be locking the tree waiting for the test build to complete. If you have a 12 hour work day, a one hour build, and 30 developers, less then half your team will be able to commit code on any given day, or you'd have a hard time commiting because you wouldn't know that somebody already has a pending commit that's waiting on the build.
This is a bad thing to programatically enforce during the commit on large projects using CVS.
Interesting little fact.. 2.2 million cards is .33% of outstanding cards in the US. Yes, you read that right.. one third of one percent.
I don't know why you find that so surprising. In a country with 300 million people you'd expect there to be a significantly larger number of existing credit card number out there than 2.2 million. Especially considering that they're used on things like gift cards issued by Novus and on check cards.
Actually, I think these storys are auto posted based on when the software update window pops up on his mac. Invariably when I'm at work and see the software update window pop up there's a corresponding slashdot story 20 minutes later.
Prior art does not invalidate a patent or make an invention unpatentable unless the prior art covers every single part of the new patent down to the tiniest details. You can patent something that's already patented with only a slight addition. You'll have to have a licence to use the other patent to implement the invention, but you can still get a patent that grants you a limited monopoly to just the new parts. Since the time this article was posted until now (when I've seen dozens of people claiming that these patents are invalid due to prior art) there is NO FUCKING WAY that people who were unaware of these patents before could have figured out wether these should be valid patents or not.
In summary: prior art doesn't mean you can't have a patent. It just means you have to list it in the "prior art" section of the patent and licence the old patent to implement your "improvement" to the old invention (assuming the original patent is still valid).
From RFC 1591:
It doesn't say anything about non-profit in there.
The uses for the TLDs were specifically defined. They mean what the RFC says they mean and not what you decided they should mean.
What? There are no separate TLD providers?
The TLDs are not all administered by the same entity. ICANN is authorised to award the contracts to manage any particular TLD, but the contracts are presently held by a variety of companies. In fact, the number of companies controlling TLDs is increasing, since Verisign is loosing controll of both the
Before you go blaming people for problems, make sure the problems exist.
Google was a poor choice when picking a company to base these predictions on. Computing power is growing exponentially, but google's core business is indexing web pages, and without an expontial human population explosion the amount of availble web content to index will grow linearly. That means it won't be too long before google can purchase a cluster of low-end machines that will do the job, and won't require upgrading for the forseeable future. Processor technolgy won't be driven by these pre-existing tasks with zero or linear growth simply because in an insignificant amount of time processor technology will make these tasks trivial. Processor technology will have to be driven by new tasks that we'll think up as the power to perform them becomes available. Google came into existance when the low-end PC became sufficient to accomplish the task. What new technology companies will spring up when the low-end is 2^4 times as powerful as when google started? How about 2^8 times? 2^32times? Perhaps eventually the tasks that will use the new processing will be designed by the machines that consume the power in the first place. Perhaps such machines can design the next generation processors. Then we won't need some company to drive the technology forward. Won't that be interesting.
Maybe it's "If they exceed 10, then they have to buy more OS licenses, or better yet a server" WTF????
Heh. This weekend I was upgrading a hard drive in an NT 4.0 server from 8GB to 36GB. I added the new drive to the SCSI bus and booted up with my newly acquired Partition Magic boot disk purchased from CompUSA just hours before. I proceded to copy the partitions onto the new drive. I then rebooted to the new drive to make sure it worked before growing the partitions to fill the drive. I was surprised to find that when I tried to run Partition Magic again it informed me that the product was not compatible with the "Server Edition" of windows, and that I needed to buy Server Magic instead. Server Magic does the same things as Partition Magic, but costs $400 more. Obviously Partition Magic works just fine with NT server, because I had just used it when the first disk in the system was blank, they had just put a check in there such that if you had deep enough pockets to buy NT server you'd have to pay $400 more for their partition utility. I would have been less pissed off about it had they stated the limitation clearly on the outside of the box, or if Server Magic were not so rediculously priced and was actually available at the local store, but it still seems wrong somehow.
Anyway, in short, it's not just Microsoft who charges you more for the same code based on what you're going to use it for.
--
BTW, I got around the stupid limitation by pulling a SCSI disk from one of the workstations and sticking it on the SCSI bus with a lower ID than the new drive. Then I just made my changes to the NT server disk and put the workstation drive back. Not a very good checking system apparently.
He has newer writings if you got to his website. He talks about how the ability to scan the brain to better understand how it's working without negatively affecting the person being scanned is increasing exponentially, and that at aproximately the same time as we acquire enough computing power to emulate the brain, we'll also have the ability to map every cell and interconnection, and record every firing of every neuron.
This is, of course, assuming that the progress of technology continues at the current rate. As we know, past numbers are no guarantee of future performance. Either way, we'll figure it out eventually, wether it be in 40 years or 400...
Clearly you haven't read their paper.
"There is no way to prove that a program cannot pass a test which a human can pass, since there is a known program - the human brain - which passes the test."
The creaters for CAPTCHA acknowledge that you can programatically defeat their creations, but that it is hard to do with high accuracy. My assertions are as follows:
I challange you do come up with a credible argument to refute any of those points.
Whoops. Submit instead of preview. Sorry about the broken sentences. I was supposed to fix those before posting. You get the idea though.