In the early 80's? It was more like $300, and that was like $600 in today's money.
The PC market will be saturated soon. The profit margins are on laptops, which are not quite as generic as PCs yet.
Who cares. The Lindows "business model" is a pain in the ass. Why on earth can't they just install all that software on the computer, like other Linux installs. The one thing that impresses people about Linux is the sheer volume of included apps. They're ELATED that it does so much, without requiring installers. They DON'T CARE about the latest downloads, because the free apps they get are "good enough".
Here's reality restated... People love their data. Most people love running familiar apps. People HATE installing software, and HATE losing data. Make a computer that satisfies those desires, and you'll have a popular machine.
Believe it or not... people don't really like to customize their systems much. They want to alter their desktops, icons, and sounds... They do not want to reconfigure all their software.
Seriously, though, they have solved two simple problems: config files, and network config. Their config files are in xml, and the generic property editor is like an xml editor. Netinfo is their networked config database, and that's been moved to LDAP.
They have some market, and some money, and the API had been around the block. If unix programmers clone it and adopt it, it'll be easier to write software that works in OS X, which matters if you're writing components that are going to be used in a desktop OS.
As for having specific key apps using these libs, well, don't sweat it. If it becomes popular in unix, people will start to adapt their software to use it. It might take a while, though.
The one missing element is dealing with "little languages" config files. These are config files that are parsed top to bottom, and support macros or variables. The most extreme example is the sendmail.cf, but even nc-ftp's config file qualifies.
The issue isn't just that 1/2 the people are underprivileged, but the degree to which that difference will be significant.
If 50,000 people are in a football stadium, and you seal it up and fill it full of water so that half are above the water, and half are below, then you have some people who are "underwater".
For those people underwater, the fact that they're there, and not several feet higher up is pretty f-cking significant. Likewise, the computer is that important -- if you don't have it, it's almost like being underwater. Here's why:
* fast becoming a dominant form of middle class communication
* access to educational material at a very low cost, and access to government services 24/7
* becoming a significant competitor to newspaper classified ads, which are used most often to trade, find jobs, and find housing
Just these three things can have a significant impact on the quality of your life. If you don't know how to use computers (maybe because you don't have a desk job) you may not think computers are important. So, you miss out, and don't even know it. Worse, your kids miss out, and after a while, discover that they've only put 500 hours in front of the screen, while the kid they're competing against has put 5000, or 10,000, or maybe even more.
Now, some of you are saying "not that much stuff is online". My answer is "not yet". But, given the pressure to keep expenses down, more and more important things will be put online, because it's just cost effective all-around.
I'm *not* *even* dealing with the fact that a computer is also an analytic, educational tool. If you learn to program a little, you'll end up learning a lot. You *can* get classes, but any programmer will tell you that, in order to really learn, you need your own computer. And, you need unlimited access to it for hours at a time.
When you talk about "poor people" you're lumping them all together like a pathological group of unfit people. That verges on something like racism.
Libraries in poor areas are used heavily. Believe me. The computers are so busy that you have to sign up hours ahead of time. Lots of kids use them.
The Internet isn't just entertainment and news (which are like the same thing, if you think about it). It's got lots of useful information about cities, laws, government forms, tax info, educational material, etc. By having it in your home, you get access to a lot of valuable information. These days, you *need* the internet if you're poor and want to help improve your situation.
Pretty soon, in the US at least, the net is going to be as important as owning a car. A car lets you find work in more places, take more kinds of jobs, and lets you work and go to school in the same day.
Sure, you could do all this without a car, but it's a lot harder.
Now, you could argue that they should just work harder. I don't agree with that, but, even if I did, you have to consider the kids. If they have loser parents who don't get "the big picture", the kids are going to suffer. They'll be "left behind".
Imagine how much harder school would be without a phone number or hot running water. Sure, you could manage, but it would be more effort, and you'd be at a significant disadvantage.
Nobody said life was easy, but if you could give some kid a leg up by getting them $500 worth of hardware and a book, you might as well try to do that.
LDAP stores contact info, but it's main purpose is to authenticate users. It's a network-wide replacement for/etc/passwd.
LDAP APIs are used to create plug-ins for different servers so that they can also be used to authenticate against the LDAP database.
The database differs from a relational SQL-style db, in that each user/record can be extended (have columns added) without modifying the entire database. In OO, I think this would be a composite object.
LDAP is the opposite of useless. It's vital. The problem, for Unix, is that LDAP exists, but the tools suck, and the default installs of unix still use/etc/passwd and don't come with LDAP plug ins for all the apps. Windows, by comparison, forces you to use their directory service, so everyone starts out being able to use a centralized directory server.
They have it right - they know that we're insensitive to blue.
If you read the description, you'll notice that pixels are defined as two green and two red dots.
The idea is that the green and red dots will define the edges of the pixel, and the blue dot just adds color information. That's pretty clever. I wonder if some graphics will get a "blue halo" effect as the blue dots are lit up around the perimeter of a shape.
It's probably not caught on, because it would require writing custom drivers for every OS out there. You'd have to write one thing to display the regular dots, and another thing to handle subpixel rendering for text.
I don't think this is GA, but it is an interesting application of data mining that will probably lead to GA-style programming.
It scans emails and finds viruses. It identifies them by finding substrings in the data from which a score is computed. Supposedly, it has a 95% accuracy rate identifying unknown viruses.
Only on Slashdot would supporters of the status quo come off as ignorant trolls.:-)
The market isn't the only way to develop software. It just happens to be the dominant model we use to distribute goods.
The problem with profit driven software development is that it tends to create monopolies. One reason is because creating a codebase for a significant product is expensive. It leads to "bloatware", where (useless) features are added to a product to help it retain market dominance.
Bloatware is at odds with good software design, which tends to favor small tools and interfaces, brought together into applications. These applications ideally are tailored for a specific audience, so that their use is simple and efficient.
I'm a db novice, but it seems pretty obvious that relational dbs are so popular because there's a lot of accounting data that fits the relational model.
When you have different needs, you need a different underlying model. A file system seems to be like a hierarchical database, no?
I'm guessing that if you need to access specific "paths to data" in a very very large database, the relational model will start to break down, and hierarchical databases (which I assume use pointers) could be faster.
I mean, a terabyte filesystem seems feasible and sounds somewhat manageable (and cheap), but a terabyte relational db for heterogenous file-like data seems rather complex and expensive.
Just don't ask me to do a "find" on the terabyte disk.:-)
Slashdot Mirror
In the early 80's? It was more like $300, and that was like $600 in today's money. The PC market will be saturated soon. The profit margins are on laptops, which are not quite as generic as PCs yet. Who cares. The Lindows "business model" is a pain in the ass. Why on earth can't they just install all that software on the computer, like other Linux installs. The one thing that impresses people about Linux is the sheer volume of included apps. They're ELATED that it does so much, without requiring installers. They DON'T CARE about the latest downloads, because the free apps they get are "good enough". Here's reality restated... People love their data. Most people love running familiar apps. People HATE installing software, and HATE losing data. Make a computer that satisfies those desires, and you'll have a popular machine. Believe it or not... people don't really like to customize their systems much. They want to alter their desktops, icons, and sounds... They do not want to reconfigure all their software.
I'd put a vote in for OSX, as an OSX user :-)
Seriously, though, they have solved two simple problems: config files, and network config. Their config files are in xml, and the generic property editor is like an xml editor. Netinfo is their networked config database, and that's been moved to LDAP.
They have some market, and some money, and the API had been around the block. If unix programmers clone it and adopt it, it'll be easier to write software that works in OS X, which matters if you're writing components that are going to be used in a desktop OS.
As for having specific key apps using these libs, well, don't sweat it. If it becomes popular in unix, people will start to adapt their software to use it. It might take a while, though.
The one missing element is dealing with "little languages" config files. These are config files that are parsed top to bottom, and support macros or variables. The most extreme example is the sendmail.cf, but even nc-ftp's config file qualifies.
I think that those should be left as-is.
The issue isn't just that 1/2 the people are underprivileged, but the degree to which that difference will be significant.
If 50,000 people are in a football stadium, and you seal it up and fill it full of water so that half are above the water, and half are below, then you have some people who are "underwater".
For those people underwater, the fact that they're there, and not several feet higher up is pretty f-cking significant. Likewise, the computer is that important -- if you don't have it, it's almost like being underwater. Here's why:
* fast becoming a dominant form of middle class communication
* access to educational material at a very low cost, and access to government services 24/7
* becoming a significant competitor to newspaper classified ads, which are used most often to trade, find jobs, and find housing
Just these three things can have a significant impact on the quality of your life. If you don't know how to use computers (maybe because you don't have a desk job) you may not think computers are important. So, you miss out, and don't even know it. Worse, your kids miss out, and after a while, discover that they've only put 500 hours in front of the screen, while the kid they're competing against has put 5000, or 10,000, or maybe even more.
Now, some of you are saying "not that much stuff is online". My answer is "not yet". But, given the pressure to keep expenses down, more and more important things will be put online, because it's just cost effective all-around.
I'm *not* *even* dealing with the fact that a computer is also an analytic, educational tool. If you learn to program a little, you'll end up learning a lot. You *can* get classes, but any programmer will tell you that, in order to really learn, you need your own computer. And, you need unlimited access to it for hours at a time.
Trolling eh?
When you talk about "poor people" you're lumping them all together like a pathological group of unfit people. That verges on something like racism.
Libraries in poor areas are used heavily. Believe me. The computers are so busy that you have to sign up hours ahead of time. Lots of kids use them.
The Internet isn't just entertainment and news (which are like the same thing, if you think about it). It's got lots of useful information about cities, laws, government forms, tax info, educational material, etc. By having it in your home, you get access to a lot of valuable information. These days, you *need* the internet if you're poor and want to help improve your situation.
Pretty soon, in the US at least, the net is going to be as important as owning a car. A car lets you find work in more places, take more kinds of jobs, and lets you work and go to school in the same day.
Sure, you could do all this without a car, but it's a lot harder.
Now, you could argue that they should just work harder. I don't agree with that, but, even if I did, you have to consider the kids. If they have loser parents who don't get "the big picture", the kids are going to suffer. They'll be "left behind".
Imagine how much harder school would be without a phone number or hot running water. Sure, you could manage, but it would be more effort, and you'd be at a significant disadvantage.
Nobody said life was easy, but if you could give some kid a leg up by getting them $500 worth of hardware and a book, you might as well try to do that.
That is pretty simplistic.
/etc/passwd.
/etc/passwd and don't come with LDAP plug ins for all the apps. Windows, by comparison, forces you to use their directory service, so everyone starts out being able to use a centralized directory server.
LDAP stores contact info, but it's main purpose is to authenticate users. It's a network-wide replacement for
LDAP APIs are used to create plug-ins for different servers so that they can also be used to authenticate against the LDAP database.
The database differs from a relational SQL-style db, in that each user/record can be extended (have columns added) without modifying the entire database. In OO, I think this would be a composite object.
LDAP is the opposite of useless. It's vital. The problem, for Unix, is that LDAP exists, but the tools suck, and the default installs of unix still use
They have it right - they know that we're insensitive to blue. If you read the description, you'll notice that pixels are defined as two green and two red dots. The idea is that the green and red dots will define the edges of the pixel, and the blue dot just adds color information. That's pretty clever. I wonder if some graphics will get a "blue halo" effect as the blue dots are lit up around the perimeter of a shape. It's probably not caught on, because it would require writing custom drivers for every OS out there. You'd have to write one thing to display the regular dots, and another thing to handle subpixel rendering for text.
I don't think this is GA, but it is an interesting application of data mining that will probably lead to GA-style programming.
It scans emails and finds viruses. It identifies them by finding substrings in the data from which a score is computed. Supposedly, it has a 95% accuracy rate identifying unknown viruses.
Malicious Email FilterOnly on Slashdot would supporters of the status quo come off as ignorant trolls. :-)
The market isn't the only way to develop software. It just happens to be the dominant model we use to distribute goods.
The problem with profit driven software development is that it tends to create monopolies. One reason is because creating a codebase for a significant product is expensive. It leads to "bloatware", where (useless) features are added to a product to help it retain market dominance.
Bloatware is at odds with good software design, which tends to favor small tools and interfaces, brought together into applications. These applications ideally are tailored for a specific audience, so that their use is simple and efficient.
I'm a db novice, but it seems pretty obvious that relational dbs are so popular because there's a lot of accounting data that fits the relational model.
:-)
When you have different needs, you need a different underlying model. A file system seems to be like a hierarchical database, no?
I'm guessing that if you need to access specific "paths to data" in a very very large database, the relational model will start to break down, and hierarchical databases (which I assume use pointers) could be faster.
I mean, a terabyte filesystem seems feasible and sounds somewhat manageable (and cheap), but a terabyte relational db for heterogenous file-like data seems rather complex and expensive.
Just don't ask me to do a "find" on the terabyte disk.