I started with basic too. It eventually frustrated me badly enough that I didn't program again for a few years. But I liked programing in it for some time.
When I started programming again it was in C, without a lot of help, and it was a tough transition initially.
I think that the questioner is right- the intricacies of C are likely to turn children off pretty quickly. But I certainly wouldn't start them off with any version of basic- better to use a language which is both a good teaching language and a fully featured programing language which runs on many platforms, with access to system services provided in reasonably uniform way.
There are a number of languages which meet these requirements to one degree or another. My opinion is that two of them stand head and shoulders above the rest- Python and Scheme.
For an older child who wants to learn about computer science, and good programming practice, and has a fair bit of maturity Scheme can't be beat. For a child who is more interested in getting quidck results Python is a nice clean language that can be used in both imperastive and OO modes. Both support Tk, so creating GUI programs is well within reach (as compared to C- it takes a long time to get competent enough at C or C++ to use MFC or Motif or Gtk or Qt or FLTK or etc.).
Well! I found your firewall extremely easy to breach. I was into your system in minutes. I think it wasn't very smart of you to use the same root password that I use. Gotta give you credit though- our taste in porn is very similar.
I'm not a lawyer. I do know a bit about IP law as I do most of my work in the publishing industry and have to deal with rights issues every day. What you say above is entirely wrong. Almost every post on this article is completely clueless. The ones that make the most sense are those that say: "Get a lawyer!" There is an alternative (sort of): go get a few law books out of your local universitie's law library. It won't qualify you to go to court on these issues, but at least you'll have _some_ clue. Of course if you live in a remote part of Wyoming you won't be able ro do this, but in any moderately urban area you should be able to find basic books on IP law. But for godsakes stop talking out your ass. Someone might take you seriously. The key word here is derivative.
It may not be enough to specify in a contract that something is or is not a work for hire. Recent court decisions have held that signing a contract agreeing that work performed is work for hire does not automatically make it work for hire.
Work for hire is work performed under a specific set of circumstances. In general factors that support a work being work for hire include the worker working on the premises of the employer, the worker working on equipment owned by the employer, the worker being paid by the hour, and the employer setting the hours which the employee works.
So if you hire me as a contractor to design a website for you, and the contract we sign says that my work is a work for hire, and I work at my home on my equipment at my own leisure for a fixed fee chances are that I still own the copyright. Conversely, if I work for you without contract, but do it on your site, on your equipment, at your beck and call for an hourly wage, chances are you get the copyright.
Of course pretty much anything done by an _actual_ employee (benefits, W4's, salary, etc.) is work for hire.
The bottom line is that if you want a really unambiguous contract it's a lot better to draft it in terms of assigned rights than to use the phrase "work for hire". Assigning all rights exclusively to the employer, for instance, is pretty unambiguous. You retain the copyright, but you can't resell or reuse the work, and the employer can do pretty much what they want with it.
Amazon controls the licenses and can withhold it from direct competitors (like B&N) while still allowing everyone else to use it.
Unethical business practices are _usually_ used against one's competitors. How is it a solution if Bezos et al can prevent a competitor from legally using technology that has been very clearly placed in the public domain. JB's letter is rubbish. All he says is "I'm gonna keep doin what I was doin, but I'll talk to my Sneator to make sure nobody else does it." Great.
You missed the point- Jeff Bezos said "I'm going to continue to do exactly what you criticised me for doing- that is active thievery and well pissing". Tim said "Oh, well as long as you put it that way."
Amazon has managed to enforce a patent on using cookies to streamline user interaction with websites. Tim O'Reilly has agreed that this is reasonable. End of story.
The problem with one-click wasn't with the patent office. Ultimate authority for decisions as to what is patentable and what is not rests with the courts. The injunction against B&N should not have been granted. It makes sense to talk about the lack of training for employees of the patent office. It does not make sense in reference to the judiciary. Judges are supposed to understand cases before ruling on them. I say we need some judges who are specialists in tech issues, and that any judge who isn't should be disqualified from hearing such a case.
>Which is that XML files are required to be valid. XML files are not required to be valid. They are required to be well formed. And making mistakes in configuration files can screw you whether or not you're using structured markup. I'm not sure I really see huge advantages to using xml for configuration files, and I like xml a lot and have found it quite useful in other situations. On the subject of error checking you could use a validating parser to ensure that the configuration file was at least a legal configuration file for the app that you're configuring, but that doesn't guarantee that it's a configuration file that does what you want, and then you are stuck with a dtd or some other schema for every app. Like I said above, I think xml is great in some contexts- it's saved me a fair bit of time in the last few months. But it seems like a lot of people want to use xml for the sake of using xml these days. I particularly agree with your last sentence. Retooling every working Unix tool out there to work with xml seems like way to high a cost for the payoff.
But the discussion was specifically about patents on software. The GPL does not rely on software patents. Some people are opposed to all IP law. I'm certainly not.
Using system memory to store instructions, linked data structures, interpreters, assemblers, compilers- all of these are far less obvious than 1-click. Let alone Quicksort. If these kinds of patents had been sought and granted there wouldn't have been any use in a GPL- you wouldn't be able to write software at all without paying licensing fees.
This is one of the dangers of abuses like Amazon's. Patent protection is important, and frivolous patents muddy the water, and could eventually lead to a weakening of legitimate IP law.
The point is the market. Wall Street likes it if you hold critical patents in your field. Right now (and for the forseeable future) Wall Street is all that's keeping Amazon alive. The point is also to slow down a potentially dangerous competitor. If Barnes and Noble had been smart they could probably have crushed Amazon. They really weren't too far behind them in getting into the online book selling business. Their original site was pretty bad though- among other things they committed what I think of as a cardinal sin- forcing the user to go through a lengthy registration process before browsing the site. But they've learned from their mistakes, and their site's not bad now. Amazon doesn't have to win the court case to have inflicted some real harm on them. Amazon's business model is built on branding- they have a nice site, good prices and good service. That's great, but a number of other sites have all those things as well. How much branding is going to matter on the net has yet to be seen. As Amazon continues to lose money with no end in sight this might occur to the shareholders and potential shareholders. A few critical patents might help to bolster confidence.
The GPL doesn't rely _at all_ on software patents. I've never heard of anyone in the Open source world using legal means to keep someone else from writing software with similar functionality to their own. There's a big difference between limiting the ways in which people can use the software you've written and limiting the kinds of software others can write.
What sources of information are there on methodologies for doing usability testing? I've read quite a bit of your site, and there is some information there, but I'd like to know more about the methods that you and others use. Some of the usability statistics you give are very specific- you might claim, for instance, a 128% increase in usability for a certain modification. Are there methods for estimating the error in these calculations?
I'd like to comment on Slashdot's usability- I really wish the pages weren't generated as one big table. It takes a really long time for a big page to load, and you can't see anything interesting until the whole thing does load.
>And then there's the majority of these websites >which are designed really badly,
This is a big problem. I've worked on some e-commerce and content sites (one of them was one of the biggest new sites of this kind last year) and my experience has been that most companies moving onto the web are unwilling to accept decent interface design. They tend to want all the bells and whistles and refuse to understand that people may be surfing the web with a different configuration than they have (i.e. not on a T1, Not on a 17" monitor, not using IE 5.0 for Windows, not at a screen resolution of 1280x1024). On top of this they don't want to test the sites for usability.
Computers are less comfortable than print even with great UI design. With lousy design they're just about unusable.
It looks like we're going to have "electronic paper" in the reasonably near future. Hopefully it will be as comfortable, or close to as comfortable, as real paper. It's not hard to imagine a time when you carry around one book, and can request any print publication you want, to be transmitted to you wirelessly. Of course you would be able to browse the web as well. Then you might _really_ see a blurring of the distinction between old and new media.
1) Like the old saying goes: "Liquor then beer, never fear; beer then liquor, never sicker." The drunker you are, the faster you drink. If you start on liquor and switch to beer you'll wind up drinking beer really quickly, which might get you kind of fucked up. On the other hand if you start drinking beer and then switch to liquor you'll wind up pounding shots, which might make you comatose.
This is an interesting point. Does this mean that you can go into a bar and absorb a certain amount of the beer floating around in the air? Some incentive to grow a beard I guess.
Sure, if you drink Bud. But the States also have a lot of really great microbreweries that make very good beer. I've lived in both Canada and the States and I think the best American beer is better than anything made in Canada.
GUI based admin tools aren't bad in and of themselves, but they tend to solve the wrong problem. One of the reasons that NT administrators are often pretty clueless is that the gui stuff gives you the illusion of being able to competently administer a system without understanding what's going on under the hood. That's dangerous. There's nothing wrong with them if they're there to make it more convenient to do something that you already know enough to do without them though.
Slashdot Mirror
I started with basic too. It eventually frustrated me badly enough that I didn't program again for a few years. But I liked programing in it for some time.
When I started programming again it was in C, without a lot of help, and it was a tough transition initially.
I think that the questioner is right- the intricacies of C are likely to turn children off pretty quickly. But I certainly wouldn't start them off with any version of basic- better to use a language which is both a good teaching language and a fully featured programing language which runs on many platforms, with access to system services provided in reasonably uniform way.
There are a number of languages which meet these requirements to one degree or another. My opinion is that two of them stand head and shoulders above the rest- Python and Scheme.
For an older child who wants to learn about computer science, and good programming practice, and has a fair bit of maturity Scheme can't be beat. For a child who is more interested in getting quidck results Python is a nice clean language that can be used in both imperastive and OO modes. Both support Tk, so creating GUI programs is well within reach (as compared to C- it takes a long time to get competent enough at C or C++ to use MFC or Motif or Gtk or Qt or FLTK or etc.).
Tcl is a rasonably close runner-up.
Well! I found your firewall extremely easy to breach. I was into your system in minutes. I think it wasn't very smart of you to use the same root password that I use. Gotta give you credit though- our taste in porn is very similar.
I'm not a lawyer. I do know a bit about IP law as I do most of my work in the publishing industry and have to deal with rights issues every day. What you say above is entirely wrong. Almost every post on this article is completely clueless. The ones that make the most sense are those that say: "Get a lawyer!" There is an alternative (sort of): go get a few law books out of your local universitie's law library. It won't qualify you to go to court on these issues, but at least you'll have _some_ clue. Of course if you live in a remote part of Wyoming you won't be able ro do this, but in any moderately urban area you should be able to find basic books on IP law. But for godsakes stop talking out your ass. Someone might take you seriously. The key word here is derivative.
It may not be enough to specify in a contract that something is or is not a work for hire. Recent court decisions have held that signing a contract agreeing that work performed is work for hire does not automatically make it work for hire.
Work for hire is work performed under a specific set of circumstances. In general factors that support a work being work for hire include the worker working on the premises of the employer, the worker working on equipment owned by the employer, the worker being paid by the hour, and the employer setting the hours which the employee works.
So if you hire me as a contractor to design a website for you, and the contract we sign says that my work is a work for hire, and I work at my home on my equipment at my own leisure for a fixed fee chances are that I still own the copyright. Conversely, if I work for you without contract, but do it on your site, on your equipment, at your beck and call for an hourly wage, chances are you get the copyright.
Of course pretty much anything done by an _actual_ employee (benefits, W4's, salary, etc.) is work for hire.
The bottom line is that if you want a really unambiguous contract it's a lot better to draft it in terms of assigned rights than to use the phrase "work for hire". Assigning all rights exclusively to the employer, for instance, is pretty unambiguous. You retain the copyright, but you can't resell or reuse the work, and the employer can do pretty much what they want with it.
Amazon controls the licenses and can withhold it from direct competitors (like B&N) while still
allowing everyone else to use it.
Unethical business practices are _usually_ used against one's competitors. How is it a solution if Bezos et al can prevent a competitor from legally using technology that has been very clearly placed in the public domain. JB's letter is rubbish. All he says is "I'm gonna keep doin what I was doin, but I'll talk to my Sneator to make sure nobody else does it." Great.
You missed the point- Jeff Bezos said "I'm going to continue to do exactly what you criticised me for doing- that is active thievery and well pissing". Tim said "Oh, well as long as you put it that way."
Amazon has managed to enforce a patent on using cookies to streamline user interaction with websites. Tim O'Reilly has agreed that this is reasonable. End of story.
The problem with one-click wasn't with the patent office. Ultimate authority for decisions as to what is patentable and what is not rests with the courts. The injunction against B&N should not have been granted. It makes sense to talk about the lack of training for employees of the patent office. It does not make sense in reference to the judiciary. Judges are supposed to understand cases before ruling on them. I say we need some judges who are specialists in tech issues, and that any judge who isn't should be disqualified from hearing such a case.
>Which is that XML files are required to be valid. XML files are not required to be valid. They are required to be well formed. And making mistakes in configuration files can screw you whether or not you're using structured markup. I'm not sure I really see huge advantages to using xml for configuration files, and I like xml a lot and have found it quite useful in other situations. On the subject of error checking you could use a validating parser to ensure that the configuration file was at least a legal configuration file for the app that you're configuring, but that doesn't guarantee that it's a configuration file that does what you want, and then you are stuck with a dtd or some other schema for every app. Like I said above, I think xml is great in some contexts- it's saved me a fair bit of time in the last few months. But it seems like a lot of people want to use xml for the sake of using xml these days. I particularly agree with your last sentence. Retooling every working Unix tool out there to work with xml seems like way to high a cost for the payoff.
But the discussion was specifically about patents on software. The GPL does not rely on software patents. Some people are opposed to all IP law. I'm certainly not.
Using system memory to store instructions, linked data structures, interpreters, assemblers, compilers- all of these are far less obvious than 1-click. Let alone Quicksort. If these kinds of patents had been sought and granted there wouldn't have been any use in a GPL- you wouldn't be able to write software at all without paying licensing fees.
This is one of the dangers of abuses like Amazon's. Patent protection is important, and frivolous patents muddy the water, and could eventually lead to a weakening of legitimate IP law.
The point is the market. Wall Street likes it if you hold critical patents in your field. Right now (and for the forseeable future) Wall Street is all that's keeping Amazon alive. The point is also to slow down a potentially dangerous competitor. If Barnes and Noble had been smart they could probably have crushed Amazon. They really weren't too far behind them in getting into the online book selling business. Their original site was pretty bad though- among other things they committed what I think of as a cardinal sin- forcing the user to go through a lengthy registration process before browsing the site. But they've learned from their mistakes, and their site's not bad now. Amazon doesn't have to win the court case to have inflicted some real harm on them. Amazon's business model is built on branding- they have a nice site, good prices and good service. That's great, but a number of other sites have all those things as well. How much branding is going to matter on the net has yet to be seen. As Amazon continues to lose money with no end in sight this might occur to the shareholders and potential shareholders. A few critical patents might help to bolster confidence.
The GPL doesn't rely _at all_ on software patents. I've never heard of anyone in the Open source world using legal means to keep someone else from writing software with similar functionality to their own. There's a big difference between limiting the ways in which people can use the software you've written and limiting the kinds of software others can write.
What sources of information are there on methodologies for doing usability testing? I've read quite a bit of your site, and there is some information there, but I'd like to know more about the methods that you and others use. Some of the usability statistics you give are very specific- you might claim, for instance, a 128% increase in usability for a certain modification. Are there methods for estimating the error in these calculations?
I'd like to comment on Slashdot's usability- I really wish the pages weren't generated as one big table. It takes a really long time for a big page to load, and you can't see anything interesting until the whole thing does load.
>And then there's the majority of these websites >which are designed really badly,
This is a big problem. I've worked on some e-commerce and content sites (one of them was one of the biggest new sites of this kind last year) and my experience has been that most companies moving onto the web are unwilling to accept decent interface design. They tend to want all the bells and whistles and refuse to understand that people may be surfing the web with a different configuration than they have (i.e. not on a T1, Not on a 17" monitor, not using IE 5.0 for Windows, not at a screen resolution of 1280x1024). On top of this they don't want to test the sites for usability.
Computers are less comfortable than print even with great UI design. With lousy design they're just about unusable.
It looks like we're going to have "electronic paper" in the reasonably near future. Hopefully it will be as comfortable, or close to as comfortable, as real paper. It's not hard to imagine a time when you carry around one book, and can request any print publication you want, to be transmitted to you wirelessly. Of course you would be able to browse the web as well. Then you might _really_ see a blurring of the distinction between old and new media.
1) Like the old saying goes: "Liquor then beer, never fear; beer then liquor, never sicker." The drunker you are, the faster you drink. If you start on liquor and switch to beer you'll wind up drinking beer really quickly, which might get you kind of fucked up. On the other hand if you start drinking beer and then switch to liquor you'll wind up pounding shots, which might make you comatose.
This is an interesting point. Does this mean that you can go into a bar and absorb a certain amount of the beer floating around in the air? Some incentive to grow a beard I guess.
Sure, if you drink Bud. But the States also have a lot of really great microbreweries that make very good beer. I've lived in both Canada and the States and I think the best American beer is better than anything made in Canada.
GUI based admin tools aren't bad in and of themselves, but they tend to solve the wrong problem. One of the reasons that NT administrators are often pretty clueless is that the gui stuff gives you the illusion of being able to competently administer a system without understanding what's going on under the hood. That's dangerous. There's nothing wrong with them if they're there to make it more convenient to do something that you already know enough to do without them though.
I Fear that the Uncertainty raised by this question cannot be resolved beyond a reasonable Doubt.