The only companies who can afford to challenge it have a boatload of bogus patents themselves. So, they get together and agree not to challenge each others bogus patents, and sign a "cross-licensing" deal, trading use of each other's bogus patents.
The little guys get hit with patent infringment suits they cannot afford to defend against, so they abandon the technology.
The next release of ColdFusion (4.5) is being re-written in native code for the Unix/Linux platforms. All current versions were ported from MFC code using WindU. Expect performance improvements on non-Microsoft platforms with the next releases.
CF allows you to abstract database calls to allow you to use the same syntax in the templates for all database types. I can write an app using MySQL via ODBC, and switch it to Oracle via native drivers without editing a single bit of code (assuming I used standard SQL in my queries). In fact, I do things like this all the time. One of my projects was to be deployed on Oracle. I developed it on a system using SQL Server, and we just dropped it onto the production system, no problem. Database connections are managed via a central interface, and are referenced by name in the code. Re-directing a data source takes less than 10 seconds.
CF provides very powerful session management functions, allowing me to maintain state much more easily, without having to write every single variable to a cookie. Only one cookie is created for identification, and I don't even have to do THAT myself. All I have to do is "turn on" session management. The session variables themselves stay on the server.
CF provides some very nice security features if you're deploying it in a multi-developer environment (ISP). The security is granular, and can be adjusted differently for each site or developer.
CF is also a lot less tedious when integrating code with HTML.
Why do I use PHP then? Well, for other things, PHP is better.
PHP has some wicked powerful tools for messing with the file system, and for performing network communications.
PHP is tight, powerful, and fast.
In the case of doing network communications, CF can only do HTTP and FTP natively. Anything else requires writing custom language extensions. CF can do SMTP, but it's not real-time. PHP does these things far more easily.
It's a limitation on a "tag", which is the stuff between the dots.
Sure, you can keep adding tags (subdomains), up to 1024 characters (including the dots). But, the question had to do with the length of the second-level domain portion.
As to why NetSol limits to 26 characters? I cannot answer that. There's no limitation I can find in the protocol that would cause this. Several other registries also have this limitation. A limit of the root-server software? A limit of the registration database? International conspiracy?
The label length limit of 63 characters is a definite limitation of the protocol, since the length of a label is marked by a single byte, and the two high bits are used to signify a pointer to another location, leaving only six bits for defining the length.
Making the labels longer than 63 characters would require serious re-writing of the protocol.
Oh boy, I REALLY hate to tell you this, but the government DOES have a way to listen in to conversations in your house. If you have a regular, non-powered, non-digital phone, it can be used as a listening device. The phone company just "activates" your line from their end, and it's just like taking the reciever off the hook at your end, every phone in the house then acts as a microphone. This technique is taught to police officers in the academy for use in hostage situations. I would find it difficult to believe that agencies like the NSA could not do this from the comfort of their own offices. Shortly after learning about this in the police academy, I switched to using 900mhz channel-skipping phones.
They're running NT because the site is built in ColdFusion. I've talked to them before, and they have every intention of switching as soon as the Linux ColdFusion port is available (beta in a few weeks, release by the end of the year)
Number of addresses per cell for every human being on earth in 2010: ~494,380,890,485,000
Personally, I think we'd have been just fine with 64 bits. I had to write some DNS stuff a few months ago and had to write my own routines to deal with IPv6. Heck, there aren't even standard datatypes for dealing with 128-bit numbers.
64 bits would have resulted in: 2,680,044,177 addresses per human being
"They did have to do some kernel hacking to increase the multimedia performance, but would not release the code to me unless I purchaces one of their units"
Excuse my ignorance, but would this not be a violation of the GPL?
I've seen all of this "up close and personal", and it ain't Redmond doing it. On our local Linux User Group list, a company (primarily NT) put out a notice that they were hiring Linux admins. (I happen to know that they're considering switching) Instead of being welcomed with open arms, they were barbequed for being an NT shop who was obviously completely ignorant of the real world. Needless to say, the company was NOT impressed, and I highly suspect will be delaying if not outright abandoning their move to Linux. I regularly get ignored on the list because I run a 50/50 Linux/NT shop, even though I've got years more experience with Linux than most of the group.
The Linux crowd needs to grow up and realize that this is just an operating system, not a religion. Total domination is not necessary, but being taken seriously is of paramount importance.
Well, the allure of cracking vs. hacking can be summed up in two words...
Instant gratification.
To be a respected hacker is a lot of work. To be a famous hacker requires incredible talent. Inspiration, elegance and sophistication are all requirements to be a respected hacker. By definition, a hacker writes his own code, since the code IS the hack.
To be a "respected" cracker, requires little time. Most crackers do not have to write their own code. Elegance and sophistication may keep you from being caught, but are not requirements. Instead, to gain recognition as a cracker, you are judged by the prominence of those systems you break.
Me, I'm in it for the long term even if I never do become famous.
Actually, in one picture I can see the shadows. But it's kind of cheating since my uncle used to work at that facility. From ground level you would see immense "cages". Each of those circles contains an antenna that looks much like a round cage about 100 ft high. They're quite impressive. And they're not kidding about that area being the middle of nowhere. Uncle was a Navy Cryptotechnolgist-Interpreter.
My Pentium-II 350 running NT takes 15.5 hours to complete a block. Could be 256mb RAM helps? My Linux box isn't strong enough to make me bother. (486-33)
In 15 years, my wife has gone through several TI calculators, each one suffering some type of physical failure. (She takes very good care of them, too)
In the same time, I've had two HP calculators, and the first one (HP 11c) still has a revered position on my desk. I only bought the HP 48 because I wanted base-number conversions. All this, and I'm quite hard on them. They've been dropped, bounced around in the saddlebags of my motorcycle, spilled on, and all other manners of abuse. The only time they have failed to work is when I ran the batteries dead.
Slashdot Mirror
No, it's even worse.
The only companies who can afford to challenge it have a boatload of bogus patents themselves. So, they get together and agree not to challenge each others bogus patents, and sign a "cross-licensing" deal, trading use of each other's bogus patents.
The little guys get hit with patent infringment suits they cannot afford to defend against, so they abandon the technology.
Great system, eh?
The next release of ColdFusion (4.5) is being re-written in native code for the Unix/Linux platforms. All current versions were ported from MFC code using WindU. Expect performance improvements on non-Microsoft platforms with the next releases.
What makes CF better?
CF allows you to abstract database calls to allow you to use the same syntax in the templates for all database types. I can write an app using MySQL via ODBC, and switch it to Oracle via native drivers without editing a single bit of code (assuming I used standard SQL in my queries). In fact, I do things like this all the time. One of my projects was to be deployed on Oracle. I developed it on a system using SQL Server, and we just dropped it onto the production system, no problem. Database connections are managed via a central interface, and are referenced by name in the code. Re-directing a data source takes less than 10 seconds.
CF provides very powerful session management functions, allowing me to maintain state much more easily, without having to write every single variable to a cookie. Only one cookie is created for identification, and I don't even have to do THAT myself. All I have to do is "turn on" session management. The session variables themselves stay on the server.
CF provides some very nice security features if you're deploying it in a multi-developer environment (ISP). The security is granular, and can be adjusted differently for each site or developer.
CF is also a lot less tedious when integrating code with HTML.
Why do I use PHP then? Well, for other things, PHP is better.
PHP has some wicked powerful tools for messing with the file system, and for performing network communications.
PHP is tight, powerful, and fast.
In the case of doing network communications, CF can only do HTTP and FTP natively. Anything else requires writing custom language extensions. CF can do SMTP, but it's not real-time. PHP does these things far more easily.
It's a limitation on a "tag", which is the stuff between the dots.
Sure, you can keep adding tags (subdomains), up to 1024 characters (including the dots). But, the question had to do with the length of the second-level domain portion.
As to why NetSol limits to 26 characters? I cannot answer that. There's no limitation I can find in the protocol that would cause this. Several other registries also have this limitation. A limit of the root-server software? A limit of the registration database? International conspiracy?
The label length limit of 63 characters is a definite limitation of the protocol, since the length of a label is marked by a single byte, and the two high bits are used to signify a pointer to another location, leaving only six bits for defining the length.
Making the labels longer than 63 characters would require serious re-writing of the protocol.
Oh boy, I REALLY hate to tell you this, but the government DOES have a way to listen in to conversations in your house. If you have a regular, non-powered, non-digital phone, it can be used as a listening device. The phone company just "activates" your line from their end, and it's just like taking the reciever off the hook at your end, every phone in the house then acts as a microphone. This technique is taught to police officers in the academy for use in hostage situations. I would find it difficult to believe that agencies like the NSA could not do this from the comfort of their own offices. Shortly after learning about this in the police academy, I switched to using 900mhz channel-skipping phones.
Interesting. The case has an uncanny resemblance to one that Intel sells. Except that the SGI case has better colors and the window on the front.
The Intel case has an SCA hot-swap backplane for six or seven drives *without* the need for expensive caddys.
The Intel case also has rack-mount brackets.
My local reseller has one of these cases with a quad-Xeon board in it. Gets me drooling every time.
They're running NT because the site is built in ColdFusion. I've talked to them before, and they have every intention of switching as soon as the Linux ColdFusion port is available (beta in a few weeks, release by the end of the year)
I don't think we have to worry for a little while
0 ,000
Each ipv6 address is made up of 16 octets (2^128):
~340,282,366,921,000,000,000,000,000,000,000,00
World Population (2010 est.): 6,883,000,000
http://www.prb.org/prb/pubs/wpds99/w pds99b.htm
Number of addresses per person worldwide in 2010:
~49,438,089,048,500,000,000,000,000,000
estimated number of cells in the human body: 100,000,000,000,000
http://madsci.wust l.edu/posts/archives/mar98/889221957.An.r.html
Number of addresses per cell for every human being on earth in 2010:
~494,380,890,485,000
Personally, I think we'd have been just fine with 64 bits. I had to write some DNS stuff a few months ago and had to write my own routines to deal with IPv6. Heck, there aren't even standard datatypes for dealing with 128-bit numbers.
64 bits would have resulted in:
2,680,044,177 addresses per human being
"They did have to do some kernel hacking to increase the multimedia performance, but would not release the code to me unless I purchaces one of their units"
Excuse my ignorance, but would this not be a violation of the GPL?
I've seen all of this "up close and personal", and it ain't Redmond doing it. On our local Linux User Group list, a company (primarily NT) put out a notice that they were hiring Linux admins. (I happen to know that they're considering switching) Instead of being welcomed with open arms, they were barbequed for being an NT shop who was obviously completely ignorant of the real world. Needless to say, the company was NOT impressed, and I highly suspect will be delaying if not outright abandoning their move to Linux. I regularly get ignored on the list because I run a 50/50 Linux/NT shop, even though I've got years more experience with Linux than most of the group.
The Linux crowd needs to grow up and realize that this is just an operating system, not a religion. Total domination is not necessary, but being taken seriously is of paramount importance.
True.
:-)
Leeches are often used on skin grafts to promote good blood flow. My wife has had to do this many times working in recovery rooms.
She claims nothing works better or faster. I think she just likes to gross people out. Medical geeks, go figure.
Well, the allure of cracking vs. hacking can be summed up in two words...
Instant gratification.
To be a respected hacker is a lot of work. To be a famous hacker requires incredible talent. Inspiration, elegance and sophistication are all requirements to be a respected hacker. By definition, a hacker writes his own code, since the code IS the hack.
To be a "respected" cracker, requires little time. Most crackers do not have to write their own code. Elegance and sophistication may keep you from being caught, but are not requirements. Instead, to gain recognition as a cracker, you are judged by the prominence of those systems you break.
Me, I'm in it for the long term even if I never do become famous.
Actually, in one picture I can see the shadows. But it's kind of cheating since my uncle used to work at that facility. From ground level you would see immense "cages". Each of those circles contains an antenna that looks much like a round cage about 100 ft high. They're quite impressive. And they're not kidding about that area being the middle of nowhere. Uncle was a Navy Cryptotechnolgist-Interpreter.
My Pentium-II 350 running NT takes 15.5 hours to complete a block. Could be 256mb RAM helps? My Linux box isn't strong enough to make me bother. (486-33)
I buy HP calculators for the quality.
In 15 years, my wife has gone through several TI calculators, each one suffering some type of physical failure. (She takes very good care of them, too)
In the same time, I've had two HP calculators, and the first one (HP 11c) still has a revered position on my desk. I only bought the HP 48 because I wanted base-number conversions. All this, and I'm quite hard on them. They've been dropped, bounced around in the saddlebags of my motorcycle, spilled on, and all other manners of abuse. The only time they have failed to work is when I ran the batteries dead.
I'll take an HP over TI any day.