Not yet here in the US. At least not in practice, since the bill often includes surcharges that are not advertised. There are some proposed laws to fix this.
Interesting, I think HR departments are fundamentally unable to vet out the good candidates in the initial cut from the pile of resumes. I get very little response from my resume, but can easily answer your two questions. I typically interview well, but get few interviews.
First question is trivial. The latter I can answer with very little knowledge of Java. In Python, I've always used a global flag to signal a thread to exit itself at a safe point in the thread main loop. I'm kinda shocked that people would actually think it is safe to use a generic stop/kill function without a custom cleanup, and that the Java language was designed that way.
I realize this is way off-topic, but I can see it both ways. I agree that freedom of economic decisions is extremely important. On the flip-side, we can take:
"It's my money, if I want to buy a TV fridge then I will buy a TV fridge." and look at it differently:
'It's my economic leverage, if I want to consume more resources than is required for the utility of keeping my food from rotting, then I will exercise my economic leverage to do so.'
This is the human condition. Most people cannot seem to voluntarily restrain consumption.
"Congress shall have power . . . To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."
It's not for the sake of creating a monopoly. The purpose is "To promote the progress of science and useful arts.." Exclusive rights to their discoveries is all that is promised, that's not quite the same thing as a guaranteeing a monopoly. Competitors still need to exist, which is why patents need to have specific claims that are not overly broad and obvious.
Using the same packaging to return an item is pretty obvious in many industries. Has been for decades. Parts service exchanges, advance unit replacements, evaluation units, etc. are all sent out with return waybills included inside the box so that the box can be re-used for the return trip. And I'm not buying this argument that the netflix mailer is so unique, they are just applying incremental changes as the obvious need arises. There is NOTHING novel about using a tried and true 'business method' by simply applying it to a new market. The equivalent is being able to re-patent the cash register everytime a retailer sells an emerging product line.
Extraordinary claims require extraordinary proof. What I see here is the run-of-the mill smoke and mirrors, not a professional white-paper quality description. But you are correct that amazing things do happen from time-to-time, and they usually get named after the person that discovered it. e.g. 'Edison Effect' But you cannot attack the laws of thermodynamics with a single starry-eyed press release, which they did by calling it free-energy. Your point #2 by definition wouldn't be classified as 'free-energy' (a.k.a. over-unity, greater that 100% efficiency, etc.), it would be 'cheap energy', or 'unexpected energy.' But they arent even making those hedges, they are pretty blatent about it:
"Steorn is making three claims for its technology:
1. The technology has a coefficient of performance greater than 100%.
2. The operation of the technology (i.e. the creation of energy) is not derived from the degradation of its component parts.
3. There is no identifiable environmental source of the energy (as might be witnessed by a cooling of ambient air temperature).
The sum of these claims is that our technology creates free energy."
But like I said, if they really think they have tapped into a new source of energy then show us hard evidence that meets professional technical standards.
The reason why everyone should be hostile to the claims of 'free-energy' is that it is physically impossible. This is known by anyone who has a reasonable education in a scientific field. And I'm not talking about something like maybe it could be done if people are clever enough... it's just impossible by the laws of thermodynamics. I'm kinda surprised that 99% of the posters on Slashdot aren't saying the same thing. It is amazing to me that in this current day that charlatans can keep using the same pitch over and over. It's like an Amway pitch, you know it's BS, but you just can't believe that people keep falling for it.
If someone makes claims of free-energy and also invites the scientific community to verify his results, then he is just looking for credibility. In reality, there will be excuses and foot-dragging. Prove me wrong, but that will be the end-result. There is nothing to see here.
it's both. In a nutshell, you make a tree of the most common sequences then index those values using sequences of bits (possibly shorter than a byte.) The shortest bit sequences get assigned to the most commonly-occuring characters and strings.
true. It's not the best solution, just one possible way to do it. Other posters described it better than I did - it's best to use a dedicated money-type if you can get access to it. But the topic at hand is that floating-point isn't the automatic answer just because you have a non-integer value to represent. It is a complex representation problem with non-trivial complications that include issues such as overflow, rounding, storage efficiency, cents, fractions of cents in some cases but not in others, accumulated error, internal representation, expressed representation, internationalization, etc.
Decimal type, (not as commonly available as floating point type.) You could also store money as an integer by counting cents. Or you could use an integer for dollars and another for cents.
Agreed. And yet that still proves nothing. If I spent a million pounds making a traditional control system redundant by duplicating everything 5 times then I'm sure I could take some of the boards out of it without too much problem. oh.. wait.. that's already how aerospace companies design airplane control systems.
I'm pretty un-impressed by the so called advances in AI. I've seen fuzzy logic, neural nets, genetic algorithms, etc. all make the same claims over time of redundancy and other magical possibilites, yet it's still just as easy to achieve these things with straight-up microcontrollers running some simple math. Granted there are some interesting properties of each one, but still nobody is making the really big breakthroughs into making systems cognitive.
Yeah, that always bothered me too. Neil Armstrong testifies that he did say it with the 'a' as would be logically expected. It was lost somewhere in transmission. It's pretty hard to overcome the incompetence of bad reporting though, especially when it gets repeated thousands of times the wrong way. Unfortunately, Armstrong is reclusive and doesn't get the word out enough about his actual quote. The official records and transcripts were amended upon Armstrongs request, for anyone who cares to verify it.
they won't pay off. Unless it's firmly in your contract that they have to pay with real equity at some point....by the way, i get your sarcasm... Your bosses (the business type) know this and they would never agree to such a thing. That's why they drive fancy cars, have big houses and hot wives.
exactly. most people just call this the 'hit by a bus factor.' Unfortunately, it's very hard to overcome when your co-workers don't want to know what you are working on. They don't understand when you try to get them to cross-train. I have a pretty healthy curiosity for what is going on outside my immediate scope, but many people just aren't willing to do the same. I get very nervous when I get closer to a deployment schedule that something bad could happen knowing I'm the only one that can make it work.
My beginnings were on QuickBasic, which looking back.. it kinda sucked, I only wish I had Python back then. But it's a fact that if I had to learn on C, I would have quit right then and taken up a sport instead of programming.:) It's not worth getting religious about, but i think *Basic, JavaScript, and Python are all reasonable choices for beginners. They can all be used in real applications as well. C and/or assembler is not a good idea until people are ready for it. I don't think there is all that much benefit to learning memory allocation, and all the complexities of declarations, pointers, page violations, header files, and compiler/linker headaches until you really need it.
To be a good programmer, you will eventually have to study these things and more, including digital logic, system hardware architecture, and assembly. But I don't agree with the 'sink or swim' school of learning, most CS students I knew in college just sank unless they had some experience already. Learning more complex concepts in due time will help you with problem solving and analysis in the future, even if you don't really ever write in C, or design a CPU.
Back when Basic was the only thing available to beginners there were some major elements missing. For instance it was nearly impossible to do any kind of dynamic data structures unless you somehow implemented your own heap in a giant array, only I didn't know what a heap was at the time. In that case, it would be pretty important to be able to write C functions instead. Now with Python, it becomes second nature to create dynamic data structures... you just append to a list. It was so easy, in fact, that I had to erase years of Basic and C habits out of my head before I could grasp how powerful and easy Python is.
Good point, it is a never ending loop if the firewall is no more secure than what you are protecting. Nothing is perfect, patch your routers when you can. I'm generally referring to your average Linksys/Belkin/Netgear home cable/DSL router.
Generally though, there is significantly more design effort put into routers and firewalls to actually make them secure. Simply having packets non-routable to the internal network by default, i.e. NAT, gives a significant level of security with very little effort. Of course NAT isn't going to magically filter your traffic if you forward a port or put a computer in the DMZ.
General purpose OS's are made to just barely work with as many features as possible. The average computer has many services running that create the exposure risk. If anyone doubts me, type 'netstat -an' in your command shell and look at how many TCP ports are in the 'listen' state, and how many UDP sockets are bound. Shutting down services is a great idea, but not always possible. Many of these you have no control over short of cutting out entire parts of the OS's core services.
That is exactly what happens these days If you plug in an unpatched XP box or a year-old version of Red Hat directly to the internet without a hardware firewall or NAT router. Start your stopwatch and see how long before you get rooted. You don't even have to be doing anything for this to happen.
Alternately, look at packet traffic incoming on an unsecured system (Using a software firewall that can be set to pop up with suspicious traffic.) And count off the number of hits in any given minute. These are real attempts to hack you, and they happen all the time 24/7 on any given IP address in the IPv4 space. Check logs on a web server sometime, they are full of requests for things kind of like: scripts/..%5c../winnt/system32/cmd.exe These are real hacks being attempted all the time.
It's that bad. Don't even think about connecting directly to the internet on an exposed IP address. Always at least have a NAT router to block uninvited incoming traffic.
With most TCP-based applications, it is possible to implement a sane firewall strategy, but H.323 (Netmeeting) makes it pretty much impossible to do so. The protocol has a standard port for the control connection, but it sets up any port it feels like for incoming UDP voice/video traffic. The protocol expects you to leave the server AND CLIENTS in the DMZ, with all the problems that brings; limits other hosts in a NAT network, and obvious over-exposure to security attacks. When I started working with H.323, I realized very quickly that this alone is a show stopper that will/has limited the adoption of practical video/voice conferencing.
The main issue is that NAT routers and firewalls work well with outgoing TCP connections, because it is easy to contruct the return route with that information alone. UDP and unsolicited TCP connections are nearly impossible to deal with without some protocol specific knowledge. Most video and voice solutions are stuck in the mindset that they have to use UDP for its unreliable, but timely data transmission. One successful exception are systems like Yahoo chat, which I know for a fact works like a charm behind a NAT firewall. It seems they fall back to a server-based TCP connection if UDP fails to route. Netmeeting seems to just go blissfully along in silence as long as the control connection can be established. It won't even do a sanity check to see if the data traffic is getting through at all.
Only semi-usable solution I know of is to have a extra-fancy router or firewall that does packet inspection and is specifically aware of H.323
It all works great on a LAN or possibly a VLAN or VPN though.
Good point, but economically speaking there isn't a demand. Economists don't count a buyer's desires, only what they are really willing to put on the table.
Sure, they bought one today. The next one tomorrow. The next one the day after... get the point?
Anti-virus software is probably the one product that needs to be commerical and subscription based, due to the true need to have updates made on a regular basis. Therefore, to say that there are many free alternatives to choose from isn't exactly true.
Use a *real* language like Python and you will get:
>>> int("spam") Traceback (most recent call last):
File "", line 1, in ? ValueError: invalid literal for int(): spam >>>
This is the correct behavior for invalid input. Not C/PHP's idea of - 'Oh we can't represent _that_ result, so maybe the user would just prefer silent errors instead!'
Slashdot Mirror
Not yet here in the US. At least not in practice, since the bill often includes surcharges that are not advertised. There are some proposed laws to fix this.
Interesting, I think HR departments are fundamentally unable to vet out the good candidates in the initial cut from the pile of resumes. I get very little response from my resume, but can easily answer your two questions. I typically interview well, but get few interviews.
First question is trivial. The latter I can answer with very little knowledge of Java. In Python, I've always used a global flag to signal a thread to exit itself at a safe point in the thread main loop. I'm kinda shocked that people would actually think it is safe to use a generic stop/kill function without a custom cleanup, and that the Java language was designed that way.
I realize this is way off-topic, but I can see it both ways. I agree that freedom of economic decisions is extremely important. On the flip-side, we can take:
"It's my money, if I want to buy a TV fridge then I will buy a TV fridge."
and look at it differently:
'It's my economic leverage, if I want to consume more resources than is required for the utility of keeping my food from rotting, then I will exercise my economic leverage to do so.'
This is the human condition. Most people cannot seem to voluntarily restrain consumption.
Agreed.
"Congress shall have power . . . To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."
It's not for the sake of creating a monopoly. The purpose is "To promote the progress of science and useful arts.." Exclusive rights to their discoveries is all that is promised, that's not quite the same thing as a guaranteeing a monopoly. Competitors still need to exist, which is why patents need to have specific claims that are not overly broad and obvious.
Using the same packaging to return an item is pretty obvious in many industries. Has been for decades. Parts service exchanges, advance unit replacements, evaluation units, etc. are all sent out with return waybills included inside the box so that the box can be re-used for the return trip. And I'm not buying this argument that the netflix mailer is so unique, they are just applying incremental changes as the obvious need arises. There is NOTHING novel about using a tried and true 'business method' by simply applying it to a new market. The equivalent is being able to re-patent the cash register everytime a retailer sells an emerging product line.
Extraordinary claims require extraordinary proof. What I see here is the run-of-the mill smoke and mirrors, not a professional white-paper quality description. But you are correct that amazing things do happen from time-to-time, and they usually get named after the person that discovered it. e.g. 'Edison Effect' But you cannot attack the laws of thermodynamics with a single starry-eyed press release, which they did by calling it free-energy. Your point #2 by definition wouldn't be classified as 'free-energy' (a.k.a. over-unity, greater that 100% efficiency, etc.), it would be 'cheap energy', or 'unexpected energy.' But they arent even making those hedges, they are pretty blatent about it:
"Steorn is making three claims for its technology:
1. The technology has a coefficient of performance greater than 100%.
2. The operation of the technology (i.e. the creation of energy) is not derived from the degradation of its component parts.
3. There is no identifiable environmental source of the energy (as might be witnessed by a cooling of ambient air temperature).
The sum of these claims is that our technology creates free energy."
But like I said, if they really think they have tapped into a new source of energy then show us hard evidence that meets professional technical standards.
The reason why everyone should be hostile to the claims of 'free-energy' is that it is physically impossible.
This is known by anyone who has a reasonable education in a scientific field. And I'm not talking about something like maybe it could be done if people
are clever enough... it's just impossible by the laws of thermodynamics. I'm kinda surprised that 99% of the posters
on Slashdot aren't saying the same thing.
It is amazing to me that in this current day that charlatans can keep using the same pitch over and over. It's like an Amway pitch,
you know it's BS, but you just can't believe that people keep falling for it.
If someone makes claims of free-energy and also invites the scientific community to verify his results, then he is just looking for credibility. In
reality, there will be excuses and foot-dragging. Prove me wrong, but that will be the end-result. There is nothing to see here.
it's both. In a nutshell, you make a tree of the most common sequences then index those values using sequences of bits (possibly shorter than a byte.) The shortest bit sequences get assigned to the most commonly-occuring characters and strings.
true. It's not the best solution, just one possible way to do it. Other posters described it better than I did - it's best to use a dedicated money-type if you can get access to it. But the topic at hand is that floating-point isn't the automatic answer just because you have a non-integer value to represent. It is a complex representation problem with non-trivial complications that include issues such as overflow, rounding, storage efficiency, cents, fractions of cents in some cases but not in others, accumulated error, internal representation, expressed representation, internationalization, etc.
yes, sounds like Huffman encoding.
Decimal type, (not as commonly available as floating point type.) You could also store money as an integer by counting cents. Or you could use an integer for dollars and another for cents.
Agreed. And yet that still proves nothing. If I spent a million pounds making a traditional control system redundant by duplicating everything 5 times then I'm sure I could take some of the boards out of it without too much problem. oh.. wait.. that's already how aerospace companies design airplane control systems.
I'm pretty un-impressed by the so called advances in AI. I've seen fuzzy logic, neural nets, genetic algorithms, etc. all make the same claims over time of redundancy and other magical possibilites, yet it's still just as easy to achieve these things with straight-up microcontrollers running some simple math. Granted there are some interesting properties of each one, but still nobody is making the really big breakthroughs into making systems cognitive.
Yeah, that always bothered me too. Neil Armstrong testifies that he did say it with the 'a' as would be logically expected. It was lost somewhere in transmission. It's pretty hard to overcome the incompetence of bad reporting though, especially when it gets repeated thousands of times the wrong way. Unfortunately, Armstrong is reclusive and doesn't get the word out enough about his actual quote. The official records and transcripts were amended upon Armstrongs request, for anyone who cares to verify it.
m l
http://www.cbc.ca/news/indepth/words/misquotes.ht
P=(V^2)/R
My bill is in the mail.
they won't pay off. Unless it's firmly in your contract that they have to pay with real equity at some point. ...by the way, i get your sarcasm... Your bosses (the business type) know this and they would never agree to such a thing. That's why they drive fancy cars, have big houses and hot wives.
exactly. most people just call this the 'hit by a bus factor.' Unfortunately, it's very hard to overcome when your co-workers don't want to know what you are working on. They don't understand when you try to get them to cross-train. I have a pretty healthy curiosity for what is going on outside my immediate scope, but many people just aren't willing to do the same. I get very nervous when I get closer to a deployment schedule that something bad could happen knowing I'm the only one that can make it work.
To be a good programmer, you will eventually have to study these things and more, including digital logic, system hardware architecture, and assembly. But I don't agree with the 'sink or swim' school of learning, most CS students I knew in college just sank unless they had some experience already. Learning more complex concepts in due time will help you with problem solving and analysis in the future, even if you don't really ever write in C, or design a CPU.
Back when Basic was the only thing available to beginners there were some major elements missing. For instance it was nearly impossible to do any kind of dynamic data structures unless you somehow implemented your own heap in a giant array, only I didn't know what a heap was at the time. In that case, it would be pretty important to be able to write C functions instead. Now with Python, it becomes second nature to create dynamic data structures... you just append to a list. It was so easy, in fact, that I had to erase years of Basic and C habits out of my head before I could grasp how powerful and easy Python is.
I've always found Python to be a lot like creating an outline for a term paper. Most people new to programming really get that concept pretty well.
Although you need to have a good editor to make Python effective. Textpad is the best IMO.
Good point, it is a never ending loop if the firewall is no more secure than what you are protecting. Nothing is perfect, patch your routers when you can. I'm generally referring to your average Linksys/Belkin/Netgear home cable/DSL router.
Generally though, there is significantly more design effort put into routers and firewalls to actually make them secure. Simply having packets non-routable to the internal network by default, i.e. NAT, gives a significant level of security with very little effort. Of course NAT isn't going to magically filter your traffic if you forward a port or put a computer in the DMZ.
General purpose OS's are made to just barely work with as many features as possible. The average computer has many services running that create the exposure risk. If anyone doubts me, type 'netstat -an' in your command shell and look at how many TCP ports are in the 'listen' state, and how many UDP sockets are bound. Shutting down services is a great idea, but not always possible. Many of these you have no control over short of cutting out entire parts of the OS's core services.
That is exactly what happens these days If you plug in an unpatched XP box or a year-old version of Red Hat directly to the internet without a hardware firewall or NAT router. Start your stopwatch and see how long before you get rooted. You don't even have to be doing anything for this to happen.
Alternately, look at packet traffic incoming on an unsecured system (Using a software firewall that can be set to pop up with suspicious traffic.) And count off the number of hits in any given minute. These are real attempts to hack you, and they happen all the time 24/7 on any given IP address in the IPv4 space. Check logs on a web server sometime, they are full of requests for things kind of like: scripts/..%5c../winnt/system32/cmd.exe These are real hacks being attempted all the time.
It's that bad. Don't even think about connecting directly to the internet on an exposed IP address. Always at least have a NAT router to block uninvited incoming traffic.
With most TCP-based applications, it is possible to implement a sane firewall strategy, but H.323 (Netmeeting) makes it pretty much impossible to do so. The protocol has a standard port for the control connection, but it sets up any port it feels like for incoming UDP voice/video traffic. The protocol expects you to leave the server AND CLIENTS in the DMZ, with all the problems that brings; limits other hosts in a NAT network, and obvious over-exposure to security attacks. When I started working with H.323, I realized very quickly that this alone is a show stopper that will/has limited the adoption of practical video/voice conferencing.
The main issue is that NAT routers and firewalls work well with outgoing TCP connections, because it is easy to contruct the return route with that information alone. UDP and unsolicited TCP connections are nearly impossible to deal with without some protocol specific knowledge. Most video and voice solutions are stuck in the mindset that they have to use UDP for its unreliable, but timely data transmission. One successful exception are systems like Yahoo chat, which I know for a fact works like a charm behind a NAT firewall. It seems they fall back to a server-based TCP connection if UDP fails to route. Netmeeting seems to just go blissfully along in silence as long as the control connection can be established. It won't even do a sanity check to see if the data traffic is getting through at all.
Only semi-usable solution I know of is to have a extra-fancy router or firewall that does packet inspection and is specifically aware of H.323
It all works great on a LAN or possibly a VLAN or VPN though.
Good point, but economically speaking there isn't a demand. Economists don't count a buyer's desires, only what they are really willing to put on the table.
It's underline markup language. Your browser must not support HTML+ULML extensions. :p
Sure, they bought one today. The next one tomorrow. The next one the day after... get the point?
Anti-virus software is probably the one product that needs to be commerical and subscription based, due to the true need to have updates made on a regular basis. Therefore, to say that there are many free alternatives to choose from isn't exactly true.
Use a *real* language like Python and you will get:
>>> int("spam")
Traceback (most recent call last):
File "", line 1, in ?
ValueError: invalid literal for int(): spam
>>>
This is the correct behavior for invalid input. Not C/PHP's idea of - 'Oh we can't represent _that_ result, so maybe the user would just prefer silent errors instead!'