If you go with the current 255.255.255.255 scheme, that's a total of about 4.2 billion combinations. At the moment, isn't that about one IP per 1.5 people in the overall world population? I mean, only a subset of that is using the internet. However, consider individuals holding multiple IP's on one end. Is there a statistician in the house that can speculate how long until we've saturated the current system?
I know it's been discussed before, but I'm curious if anybody's considered what will break when they need to implement a new system.
Microsoft is great at adding new code, but it's the end of the world when it comes to maintaining any of it. It was easy enough "integrating" the OS with the browser, but now it's impossible to take existing code, and modify it.
Code maintenance is a costly pain. It's where the real debugging is at. Who can expect Microsoft to modify their code when they can't even debug it?
[ouch];)
But seriously, it's so easy sometimes to just start fresh. Contending with an old project, with all the bad memories of things gone wrong, can be overwhelming. That in itself should suggest something else. I think Microsoft is implying they don't want to touch that code with a 50-foot pole. It was probably developed under older development standards, with out-of-date documentation. Nobody that did the work in the first place can be found...
...remember that these are software engineers at a 8-to-5:30 job, not a little band of hackers that just dive in. And to make matters worse, the specifications are being handed to them from outside Microsoft! These same people probably go beserk when another division starts telling them how to do thing. So I can see why they've been kicking and screaming about the web integration, even if it's blown out of proportions.
Disregarding ZDNet's best efforts of attracting attention to themselves, this brings up a subtle point -- possibly alarming to those bonafide "computer scientists" out there. Unfortunately, this will be somewhat off-topic, but it reminds me of current programming method. It has been my experience that the popular languages follow a familiar syntax:
function foo {
do {
weeeeeeee();
}
}
This describes things like Perl, C, C++, Java, JavaScript,... you get the idea.
How many people do you know that write in any language as if it were C? There should be some nods out there. C is procedural, of course, and came from a bygone era where a console was the hip new thing. Unfortunately, people still seem stuck in these times. Object-oriented programming, is a technique, not a language. Some people have picked up on this and relish in object-oriented design, milking it like a cash cow. Others have taken it about this far:
function foo() {
do {
IhateOOP->weeeeeee();
}
}
Some people, no matter what one does, will take the newest thing and try their darndest to change the smallest amount while learning it. In the process, they disregard any new methodology these newer languages could provide.
At this point, I would like to make some people here scream by mentioning... LISP. Remember this thing? Did you have to program under one of its cousins in college, like Scheme? It's a nifty language, with some powerful stuff. In between all the brain damage of recursive, paranthetical expressions, there is some important lessons in programming in there.
I feel people forgot the big picture here. They can make these C-like languages do anything, if they hack at it enough. That includes GUI programming.
So what if it gets the job done. Does it have a level of elegance? There is plenty of disgusting GUI code out there. Programmers shrug and accept it as a way of life. But who is trying to innovate and create the language of the future?
We need to step back, and think about everything we really want to do. Then we need to ask ourselves if the languages we have right now do it best. I believe they don't. And worse yet, they don't try. The newest thing is just another extension on C. In some ways, Perl is C++++++, but it still has that C in there. It changes the words around, but the grammar is the same. And I'm betting the grammar is restricting the way we innovate in our software.
Slashdot Mirror
If you go with the current 255.255.255.255 scheme, that's a total of about 4.2 billion combinations. At the moment, isn't that about one IP per 1.5 people in the overall world population? I mean, only a subset of that is using the internet. However, consider individuals holding multiple IP's on one end. Is there a statistician in the house that can speculate how long until we've saturated the current system?
I know it's been discussed before, but I'm curious if anybody's considered what will break when they need to implement a new system.
Microsoft is great at adding new code, but it's the end of the world when it comes to maintaining any of it. It was easy enough "integrating" the OS with the browser, but now it's impossible to take existing code, and modify it.
;)
...remember that these are software engineers at a 8-to-5:30 job, not a little band of hackers that just dive in. And to make matters worse, the specifications are being handed to them from outside Microsoft! These same people probably go beserk when another division starts telling them how to do thing. So I can see why they've been kicking and screaming about the web integration, even if it's blown out of proportions.
Code maintenance is a costly pain. It's where the real debugging is at. Who can expect Microsoft to modify their code when they can't even debug it?
[ouch]
But seriously, it's so easy sometimes to just start fresh. Contending with an old project, with all the bad memories of things gone wrong, can be overwhelming. That in itself should suggest something else. I think Microsoft is implying they don't want to touch that code with a 50-foot pole. It was probably developed under older development standards, with out-of-date documentation. Nobody that did the work in the first place can be found...
Disregarding ZDNet's best efforts of attracting attention to themselves, this brings up a subtle point -- possibly alarming to those bonafide "computer scientists" out there. Unfortunately, this will be somewhat off-topic, but it reminds me of current programming method. It has been my experience that the popular languages follow a familiar syntax: function foo { do { weeeeeeee(); } } This describes things like Perl, C, C++, Java, JavaScript, ... you get the idea.
How many people do you know that write in any language as if it were C? There should be some nods out there. C is procedural, of course, and came from a bygone era where a console was the hip new thing. Unfortunately, people still seem stuck in these times. Object-oriented programming, is a technique, not a language. Some people have picked up on this and relish in object-oriented design, milking it like a cash cow. Others have taken it about this far:
function foo() {
do {
IhateOOP->weeeeeee();
}
}
Some people, no matter what one does, will take the newest thing and try their darndest to change the smallest amount while learning it. In the process, they disregard any new methodology these newer languages could provide.
At this point, I would like to make some people here scream by mentioning... LISP. Remember this thing? Did you have to program under one of its cousins in college, like Scheme? It's a nifty language, with some powerful stuff. In between all the brain damage of recursive, paranthetical expressions, there is some important lessons in programming in there.
I feel people forgot the big picture here. They can make these C-like languages do anything, if they hack at it enough. That includes GUI programming.
So what if it gets the job done. Does it have a level of elegance? There is plenty of disgusting GUI code out there. Programmers shrug and accept it as a way of life. But who is trying to innovate and create the language of the future?
We need to step back, and think about everything we really want to do. Then we need to ask ourselves if the languages we have right now do it best. I believe they don't. And worse yet, they don't try. The newest thing is just another extension on C. In some ways, Perl is C++++++, but it still has that C in there. It changes the words around, but the grammar is the same. And I'm betting the grammar is restricting the way we innovate in our software.