I'm going to say this isn't very likely. At least in the US, people are only executed for crimes where they cause direct physical harm to another person (generally murder and occasionally rape). For other offences you generally get a life sentence, or defacto life sentence (say 135 years in the clink).
Uh no - rape in of itself is a not a capital offense, the max penalty in most states is life imprisonment. You may be thinking of murder in the commission of a rape(or sexual assault in general), which is definitely a capital offense in most (if not all) U.S. states. I know in Virginia thats one of the cases that merits the possibility of capital punishment - others including murder of a police officer, murder in commission of a drug deal, and mass murder. There has to be specific criteria - not all first degree murder offenses are capital offenses.
We haven't fought a victorious full-scale battle on our own since the Civil War.
Spanish-American War, and then the resulting Phillipine insurrection, which we both won. All on our own.
And I can't think of any occasion where we have won a battle against a half-way decent foe.
D-day? Battle of Midway, June 1942? 'Battle' of the Atlantic 1941-1943?
We tend to run if they come at us hard. When was the last time you heard of a glorious last stand of US troops, outside a Hollywood film? We only fight when we think outnumber or out gun the enemy so much that the result is a certainty.
Battle of the Ardennes, December 1944. 101st Airborne held out for ~1 week against a numerically and technologically super ior German force before being relieved by Patton's 3rd Army. And no, the USAF(technically at the time the Army Air Corps) was not involved because there was bad weather during that week. (A big part of the reason the Germans launched the offensive then was the forceast called for bad weather). Also held out in Wake Island, Luzon and Corrigedor(sp?) for a while. Yes, the garrisons eventually surrendered, just like every large force of every combatant that has been cut off from supplies did for the last 200 years. If you have a counterexample, I would like to see it.
And when we find we made a mistake, like Vietnam, we collapse.
The US had beaten the North Vietnamese on the battlefield in every single major engagement when they were deployed. Even after the bulk of US ground forces left and all that was left was advisors and air support. The '72 NVA offensive failed. It was only after the US stopped funding the puppet South Vietnamese regime in '75 that they collapsed.
If you don't agree with my assements I would like to what other countries have been so much 'better' than the US.
Unfortunately there are not bike paths everywhere, and many places don't have reasonable shoulders. Combined with the typical American drivers attitude that they own the road, especially on highways, biking to work is dangerous proposition. I was looking into biking ot work sometimes, but I'd rather not get killed.
It was a good show, but it was still a soap opera. It doesn't need resolution an ending any more than Days of Our Lives or General Hospital needs one. The final episode convinced me of that. Not only did they leave you hanging as to Tony's fate, but none of the other major characters(that didn't die) seemed to have changed much at all over the course of ~8 years. A.J is still the same confused and spoiled brat that he was before. Meadow is engaged, but she was before(to Finch) and it looks like the same road as Jackie Aprile, Finch, etc. And she's still in denial about Tony's mob involvement. Carmela has become a bit more independent, but is still basically the same character she was about the beginning of the series. 'The Sopranos' had great dialogue, good use of symbolism, good directing and acting etc. But the series wasn't essentially different than 'Dallas'
Even if that photograph is authentic it doesn't necessarily mean its Oswald. There is a simlliarity but the photograph is of such poor quality that it could be just about anyone.
And where do the genetic changes that create new traits that are selected for or against come from? Ultimately from random mutations and crosses. The crosses aren't purely random, of course, but they can be viewed as a largely independent random variable (i.e. the factors that determine selection of a mate don't necessarily have anything to do with survival, except insofar as they help getting selected as a mate).
You seem to be thinking on an individual level - and evolution is not about individuals but groups. Mutation events for individuals might be 'random', but the selection on the phenotype that results from the mutation(assuming it does actually change the phenotype) is not, in a statistical sense. Mutations are either beneficial, harmful or neutral in terms of survival - and unless you are talking about sexual selection have little to do with selection of mate. I don't know of any biologist that would claim that evolution is a completely "random process".
The origin of the earliest forms of "life" (simple single "cells", pre-DNA, pre_RNA) is surely one of the easiet things to answer - this is nothing more than self-sustaining chemical reactions occuring in a lipid bubble (maybe naturally occuring - oily froth by the sea shore, or maybe the fatty polymers being a product of the chemical reactions that occured inside them). I
Ahh, no its not. There are some good guesses, and they've been able to discover things like short sequences of RNA that can catalyze their own reproduction, but (natural) abiogenesis is by no means a solved problem. The simplest organism that can reproduce on its own (not without a host organism like a virus) is a prokaryotic bacteria, but even there you still have millions and millions of base pairs of DNA, which could not come randomly together by chance.
The non-intuitive part comes with speciation - how do new species occur. A typical simplistic creationist objection goes something like "Well if humans evolved from apes then the 'first human' that was the offspring of a an ape would not have been able to find a mate because there were no other humans to mate with. Therefore humans could not have evolved from apes or any other species.' You have to explain to them that reproduction is not an all or nothing thing and that species boundaries are sometimes not clear cut. It is unfortunate that this is not made clear to more people when studying basic taxonomy.
"Do you believe that evolution is a result of purely random chance?"
To anyone who's studied even a overview of the the ToE the answer to this question should be "No." Evolution occurs due to the mechanisms of natural selection, sexual selection and genetic drift*. Only the last of those three can be considered to be "purely random".
*I know there are some others, but those are the major ones.
Having been a manager of sales people, the response to "If you want more money, sell more high commission items" is:
"I quit".
You must not have worked in the mortage origination industry or at car dealership then, where unless you have supervisory functions, 100% of your take-home pay is commissions on sales.
Every so often, someone wants to start using AJAX and automatic form reposting
and a bunch of stuff that makes their code unreadabale and unmaintainable by the rest of the
crew.
Well, there's you're problem. You shouold never look at the javascript code GWT generates anymore than you should look at the Java bytecode javac generates. Raw javascript might be OK for small applications, but for large, complex, apps it gets increasingly difficult to read, debug and maintain. And you can't build applications with scripting languages. At least, not without a massive headache.
Its getText-method is deprecated, so there goes your IS-A-relationship. For an added bonus, cosider what should happen when you call add(myJComponent) on a JTextField. Anything that forces designers to chop down trees like that is good, period.
Well just because one method (TextComponent.getText()) is overrridden/deprecated does not mean that the IS-A relationship is broken. Its just redefining the behavior. The basic idea "JTextComponents are GUI components that a user can enter text into" is still intact. For example, lets say I have a Car base class and a subclass Car. Lets say Car has a function called shift(gear) which sets the gear the car is currently in, and defines its behavior to be public. Lets say some subclasses of Cars have automatic transmissions, and they override the behavior of shift(). So are cars with automatic transmisssions not really Cars? No, they're still Cars, they just define the behavior differently. This would be a little cleaner with multiple inheritance (you could define the methods for ManualTransmissionCar and AutomaticTransmissionCar and subclass from them and inherit their behavior without having to define it for each class since all AutomaticTransmissionCars are pretty much the same)
You're exazmple of the add() to a JTextField(which subclasses from JComponent, which subclasses from Container??) is taken, but its because the design was bad, not that inheritance is bad. The Swing architechture is somewhat of a mess because Java does not allow multiple inheritance. They subclassed javax.swing.JCoomponent off java.awt.Container instead of coming up with their own base JContainer. (I don't know why they did that, except maybe for interoperability between Swing and AWT, but who combines AWT and Swing anyway? (With the exception of the the one required peer for the frame/dialog window?). Incidentally, JFrames and JDialogs are *not* JComponents, because the subclass from java.awt.Frame and java.awt.Dialog, respectively, so they can't sublcass from JComponent. Swing would look much nicer if it was designed from the get go (from 1.0) and there were very few "heavyweight" peers(they could just have one, for the native window, and draw all the borders/decorations in Java), or they allowed multiple inheritance.
If you have an alternate architechture for Swing that does *not* use inheritance and makes more sense, I would like to see it.
Have you never used composition and delegation in OOP?
Inheritance is not the only way to reuse code.
Delegation also does not have to be a burden if the language itself supports it as a feature vs. making the developer hand-code all of it.
Yes, I use composition all the time. Composition is no way to reuse code. It also breaks the IS-A/HAS-A paradigm. MyCoolComponent doesn't *have* a JComponent. It *is* a JComponent. And I should be able to add it to a Container class. Except the Container class will only take JComponents. Now you could say "well you could just redesign Container so that it will accept objects that implement JComponentInterface". Problem is now you have the same problem I mentioned before. Every object that implements JComponentInterface needs to *fully* define every method that JComponentInterface declares, or else it is an abstract class (not allowed in your system). Now you could solve that problem by adding some delegates mechanism - saying well MyCoolComponent delegates all methods of JComponentInteface to MyCoolComponents instance of JComponent. You can call it delegation if you want. But its really just multiple inhertance, without the benefit of MyCoolComponent being able to access JComponent's protected methods/variables.I can see maybe an argument that subclasses shouldn't be able to access their super classes' variables. But if you want to be that tight, just declare them private and noone but that class will be able to access them.
That said, to me, the best Gosling quote I ever read was in an Artima interview. Basically, Gosling said that he regretted not having gone "pure interface" in Java. Java could have been a much cleaner language if Gosling had gone this way: remove the "abstract" keyword/construct from the language (and hence remove the "protected" keyword too, which makes no sense once the "abstract" keywork is gone). And prevent concrete inheritance. So Java would have been a more OO language, where only inheritance of interface would have been possible. Some people (a tiny minority) program in Java in this way. Most people do not get that concept: that OO is all about ADT (Abstract Data Types) and that a concrete implementation should be a detail. To most people this is heresy, because "you've got to type more characters, you've got to create 'needless' interfaces,..." (in modern IDEs most of these problems are non-existent for you can delegate/compose using a single shortcut). Remove the abstract keyword today from the language, and you can millions of Java developers screaming and whining endlessly. It is sooo sad Gosling didn't do it before releasing Java to the public. Every single program out there today in Java would be more OO-ish and more testable.
What do you gain by doing this? Allowing sublcasses of an abstract base class allows all the subclasses to share a common code base. I believe that is one of the more powerful features of OOPL. If I have code that is identical across different implementations of a base class, when I sublcass the base class I get all this code for free. Without the ability to do that, I have to do a cut and paste of the code. And that is ugly and more difficult to maintain. For example, I don't have to implement for every single FooSubClass that extends Foo( because all Foo(x, y, z) does is copy x, y and z to foo.x, foo.y and foo.z Well, I guess you do have to write it. But then all you have to do is call super(x, y, z). A simple example, but there are lots of times when subclasses share a common code base. For example, I would really like to see what say, the Swing/AWT code base would look like without the ability to do subclass from abstract base classes. Or, for that matter, what client code would look like. What would you do if you can't subclass JComponent to provide your own paintComponent() method? Should you have to duplicate every method JComponent implements from JComponenetInterface(not a real interface, but I guess thats what you would puse something like that)?
On that note to that I would add that its a shame that Java doesn't have multiple inheritance. I know it opens up cans of worms and its easily misused, but there are some times when it really comes in handy, and it makes the code base alot more clean.
Whenever someone mods a game, you are doing exactly what George Lucas did, changing the "story" to what you thinks work better.
Actually most mods I've played are either epilogues or prologues, or involve completely different characters, or in a completely different setting. No one releases a mod that is a mere "tweak" of the main storyline. Or if they do, they aren't very popular. And by tweak I mean storyline tweak, not gameplay tweak.
Well if their transmitter or you're receiver dies there should be very good backup system that can detect cars without a radio signal. Both systems are extremely unlikely to fail at the same time without adequate warning. Also, there should probably always be a special lane for cars that don't have the system or in the cases when the. I think the system will actually start out by being a special lane(like we have now with HOV lnaes) and most of the old cars will be in the standard lanes. When other people see how much faster the cars moving in the special lane are going, they will want the system for their own cars and it will catch on so that more and more lanes will eventually be the automated lanes and the older lanes will be considered "special".
I agreee that different languages should be introduced to undergrads early, and that they should get the basics of algorithms down before they do a big project. Java, C++ and other OOP languages are more suited towards high-level software engineering type things, which is really something that should be done in the 3rd/4th year. With a language like Scheme/LISP, the student can just focus on the algorithms and not have to worry about object creation, data typing, garbage colleciton (C++), etc. Then, coming from the other end you should probably also teach some sort of machine-like language. Maybe not real assembly language, but soemthing close to it - perhaps something like this old 'RAM'(Random Access Machine) pure theoretical language, to get a feel for what things like are at that level, and/or standard C for systems programming (i.e. primarily OS's). Then you should move on to languages like Java. As for the transistor/gate level, I think the rudiments should be taught but I don't think anything beyond a semester should be required, as not everyone is going to want to go that route and most people are doing high-level stuff nowadays.
Slashdot Mirror
We get a Sagan once a generation, and to remain above the fray he had to go so far as refusing to denounce astrolog
Uhh, Sagan did denouce astrology. He devoted a part of his Cosmos series to it: http://video.google.com/videoplay?docid=2515801214674828336&q=sagan+astrology&total=15&start=0&num=10&so=0&type=search&plindex=0
Uh no - rape in of itself is a not a capital offense, the max penalty in most states is life imprisonment. You may be thinking of murder in the commission of a rape(or sexual assault in general), which is definitely a capital offense in most (if not all) U.S. states. I know in Virginia thats one of the cases that merits the possibility of capital punishment - others including murder of a police officer, murder in commission of a drug deal, and mass murder. There has to be specific criteria - not all first degree murder offenses are capital offenses.
Spanish-American War, and then the resulting Phillipine insurrection, which we both won. All on our own.
D-day? Battle of Midway, June 1942? 'Battle' of the Atlantic 1941-1943?
Battle of the Ardennes, December 1944. 101st Airborne held out for ~1 week against a numerically and technologically super ior German force before being relieved by Patton's 3rd Army. And no, the USAF(technically at the time the Army Air Corps) was not involved because there was bad weather during that week. (A big part of the reason the Germans launched the offensive then was the forceast called for bad weather). Also held out in Wake Island, Luzon and Corrigedor(sp?) for a while. Yes, the garrisons eventually surrendered, just like every large force of every combatant that has been cut off from supplies did for the last 200 years. If you have a counterexample, I would like to see it.
The US had beaten the North Vietnamese on the battlefield in every single major engagement when they were deployed. Even after the bulk of US ground forces left and all that was left was advisors and air support. The '72 NVA offensive failed. It was only after the US stopped funding the puppet South Vietnamese regime in '75 that they collapsed.
If you don't agree with my assements I would like to what other countries have been so much 'better' than the US.
Use solar power?
If not, you're probably better off running off the grid.
Unfortunately there are not bike paths everywhere, and many places don't have reasonable shoulders. Combined with the typical American drivers attitude that they own the road, especially on highways, biking to work is dangerous proposition. I was looking into biking ot work sometimes, but I'd rather not get killed.
It was a good show, but it was still a soap opera. It doesn't need resolution an ending any more than Days of Our Lives or General Hospital needs one. The final episode convinced me of that. Not only did they leave you hanging as to Tony's fate, but none of the other major characters(that didn't die) seemed to have changed much at all over the course of ~8 years. A.J is still the same confused and spoiled brat that he was before. Meadow is engaged, but she was before(to Finch) and it looks like the same road as Jackie Aprile, Finch, etc. And she's still in denial about Tony's mob involvement. Carmela has become a bit more independent, but is still basically the same character she was about the beginning of the series. 'The Sopranos' had great dialogue, good use of symbolism, good directing and acting etc. But the series wasn't essentially different than 'Dallas'
They have solar-cell manufacturing plants that run nearly entirely on solar power
Even if that photograph is authentic it doesn't necessarily mean its Oswald. There is a simlliarity but the photograph is of such poor quality that it could be just about anyone.
Heh, you must have never taken Biochemistry :-)
Heh, you must have taken Biochemistry :-)
You seem to be thinking on an individual level - and evolution is not about individuals but groups. Mutation events for individuals might be 'random', but the selection on the phenotype that results from the mutation(assuming it does actually change the phenotype) is not, in a statistical sense. Mutations are either beneficial, harmful or neutral in terms of survival - and unless you are talking about sexual selection have little to do with selection of mate. I don't know of any biologist that would claim that evolution is a completely "random process".
Ahh, no its not. There are some good guesses, and they've been able to discover things like short sequences of RNA that can catalyze their own reproduction, but (natural) abiogenesis is by no means a solved problem. The simplest organism that can reproduce on its own (not without a host organism like a virus) is a prokaryotic bacteria, but even there you still have millions and millions of base pairs of DNA, which could not come randomly together by chance.
The non-intuitive part comes with speciation - how do new species occur. A typical simplistic creationist objection goes something like "Well if humans evolved from apes then the 'first human' that was the offspring of a an ape would not have been able to find a mate because there were no other humans to mate with. Therefore humans could not have evolved from apes or any other species.' You have to explain to them that reproduction is not an all or nothing thing and that species boundaries are sometimes not clear cut. It is unfortunate that this is not made clear to more people when studying basic taxonomy.
To anyone who's studied even a overview of the the ToE the answer to this question should be "No." Evolution occurs due to the mechanisms of natural selection, sexual selection and genetic drift*. Only the last of those three can be considered to be "purely random".
*I know there are some others, but those are the major ones.
You must not have worked in the mortage origination industry or at car dealership then, where unless you have supervisory functions, 100% of your take-home pay is commissions on sales.
Every so often, someone wants to start using AJAX and automatic form reposting and a bunch of stuff that makes their code unreadabale and unmaintainable by the rest of the crew.
Well, there's you're problem. You shouold never look at the javascript code GWT generates anymore than you should look at the Java bytecode javac generates. Raw javascript might be OK for small applications, but for large, complex, apps it gets increasingly difficult to read, debug and maintain. And you can't build applications with scripting languages. At least, not without a massive headache.
Its getText-method is deprecated, so there goes your IS-A-relationship. For an added bonus, cosider what should happen when you call add(myJComponent) on a JTextField. Anything that forces designers to chop down trees like that is good, period.
Well just because one method (TextComponent.getText()) is overrridden/deprecated does not mean that the IS-A relationship is broken. Its just redefining the behavior. The basic idea "JTextComponents are GUI components that a user can enter text into" is still intact. For example, lets say I have a Car base class and a subclass Car. Lets say Car has a function called shift(gear) which sets the gear the car is currently in, and defines its behavior to be public. Lets say some subclasses of Cars have automatic transmissions, and they override the behavior of shift(). So are cars with automatic transmisssions not really Cars? No, they're still Cars, they just define the behavior differently. This would be a little cleaner with multiple inheritance (you could define the methods for ManualTransmissionCar and AutomaticTransmissionCar and subclass from them and inherit their behavior without having to define it for each class since all AutomaticTransmissionCars are pretty much the same)
You're exazmple of the add() to a JTextField(which subclasses from JComponent, which subclasses from Container??) is taken, but its because the design was bad, not that inheritance is bad. The Swing architechture is somewhat of a mess because Java does not allow multiple inheritance. They subclassed javax.swing.JCoomponent off java.awt.Container instead of coming up with their own base JContainer. (I don't know why they did that, except maybe for interoperability between Swing and AWT, but who combines AWT and Swing anyway? (With the exception of the the one required peer for the frame/dialog window?). Incidentally, JFrames and JDialogs are *not* JComponents, because the subclass from java.awt.Frame and java.awt.Dialog, respectively, so they can't sublcass from JComponent. Swing would look much nicer if it was designed from the get go (from 1.0) and there were very few "heavyweight" peers(they could just have one, for the native window, and draw all the borders/decorations in Java), or they allowed multiple inheritance.
If you have an alternate architechture for Swing that does *not* use inheritance and makes more sense, I would like to see it.
Have you never used composition and delegation in OOP? Inheritance is not the only way to reuse code. Delegation also does not have to be a burden if the language itself supports it as a feature vs. making the developer hand-code all of it.
Yes, I use composition all the time. Composition is no way to reuse code. It also breaks the IS-A/HAS-A paradigm. MyCoolComponent doesn't *have* a JComponent. It *is* a JComponent. And I should be able to add it to a Container class. Except the Container class will only take JComponents. Now you could say "well you could just redesign Container so that it will accept objects that implement JComponentInterface". Problem is now you have the same problem I mentioned before. Every object that implements JComponentInterface needs to *fully* define every method that JComponentInterface declares, or else it is an abstract class (not allowed in your system). Now you could solve that problem by adding some delegates mechanism - saying well MyCoolComponent delegates all methods of JComponentInteface to MyCoolComponents instance of JComponent. You can call it delegation if you want. But its really just multiple inhertance, without the benefit of MyCoolComponent being able to access JComponent's protected methods/variables.I can see maybe an argument that subclasses shouldn't be able to access their super classes' variables. But if you want to be that tight, just declare them private and noone but that class will be able to access them.
That said, to me, the best Gosling quote I ever read was in an Artima interview. Basically, Gosling said that he regretted not having gone "pure interface" in Java. Java could have been a much cleaner language if Gosling had gone this way: remove the "abstract" keyword/construct from the language (and hence remove the "protected" keyword too, which makes no sense once the "abstract" keywork is gone). And prevent concrete inheritance. So Java would have been a more OO language, where only inheritance of interface would have been possible. Some people (a tiny minority) program in Java in this way. Most people do not get that concept: that OO is all about ADT (Abstract Data Types) and that a concrete implementation should be a detail. To most people this is heresy, because "you've got to type more characters, you've got to create 'needless' interfaces,..." (in modern IDEs most of these problems are non-existent for you can delegate/compose using a single shortcut). Remove the abstract keyword today from the language, and you can millions of Java developers screaming and whining endlessly. It is sooo sad Gosling didn't do it before releasing Java to the public. Every single program out there today in Java would be more OO-ish and more testable.
What do you gain by doing this? Allowing sublcasses of an abstract base class allows all the subclasses to share a common code base. I believe that is one of the more powerful features of OOPL. If I have code that is identical across different implementations of a base class, when I sublcass the base class I get all this code for free. Without the ability to do that, I have to do a cut and paste of the code. And that is ugly and more difficult to maintain. For example, I don't have to implement for every single FooSubClass that extends Foo( because all Foo(x, y, z) does is copy x, y and z to foo.x, foo.y and foo.z Well, I guess you do have to write it. But then all you have to do is call super(x, y, z). A simple example, but there are lots of times when subclasses share a common code base. For example, I would really like to see what say, the Swing/AWT code base would look like without the ability to do subclass from abstract base classes. Or, for that matter, what client code would look like. What would you do if you can't subclass JComponent to provide your own paintComponent() method? Should you have to duplicate every method JComponent implements from JComponenetInterface(not a real interface, but I guess thats what you would puse something like that)?
On that note to that I would add that its a shame that Java doesn't have multiple inheritance. I know it opens up cans of worms and its easily misused, but there are some times when it really comes in handy, and it makes the code base alot more clean.
Whenever someone mods a game, you are doing exactly what George Lucas did, changing the "story" to what you thinks work better.
Actually most mods I've played are either epilogues or prologues, or involve completely different characters, or in a completely different setting. No one releases a mod that is a mere "tweak" of the main storyline. Or if they do, they aren't very popular. And by tweak I mean storyline tweak, not gameplay tweak.
Well if their transmitter or you're receiver dies there should be very good backup system that can detect cars without a radio signal. Both systems are extremely unlikely to fail at the same time without adequate warning. Also, there should probably always be a special lane for cars that don't have the system or in the cases when the. I think the system will actually start out by being a special lane(like we have now with HOV lnaes) and most of the old cars will be in the standard lanes. When other people see how much faster the cars moving in the special lane are going, they will want the system for their own cars and it will catch on so that more and more lanes will eventually be the automated lanes and the older lanes will be considered "special".
Scheme is also very weakly typed, right? I can imagine that adding to considerable inefficiency.
I agreee that different languages should be introduced to undergrads early, and that they should get the basics of algorithms down before they do a big project. Java, C++ and other OOP languages are more suited towards high-level software engineering type things, which is really something that should be done in the 3rd/4th year. With a language like Scheme/LISP, the student can just focus on the algorithms and not have to worry about object creation, data typing, garbage colleciton (C++), etc. Then, coming from the other end you should probably also teach some sort of machine-like language. Maybe not real assembly language, but soemthing close to it - perhaps something like this old 'RAM'(Random Access Machine) pure theoretical language, to get a feel for what things like are at that level, and/or standard C for systems programming (i.e. primarily OS's). Then you should move on to languages like Java. As for the transistor/gate level, I think the rudiments should be taught but I don't think anything beyond a semester should be required, as not everyone is going to want to go that route and most people are doing high-level stuff nowadays.
They do use contraception, at least in the first world. Its just kept hush-hush, and they are encouraged to have large families regardless.