That era of the human race is over. We are now moving into post-industrialism. Many people have not noticed or are in denial, but there is no stopping it.
Indeed. With that many unemployed and on minimal or no benefits, society begins to crumble fast. The economy does too, because people need to be able to buy things to keep commerce going.
That is not an accurate model of what is happening now. We never ever had machinery before on the capability level that weak AI is currently reaching. At the same time, a lot of manufacturing technology is in a final state, and there is no next step, except eliminate the humans still in it.
Indeed. This is the first time in history that machinery is not just shifting the labor to higher qualified labor, it is replacing the labor. Well, there _is_ a shift, but before it was something like 1 person shifted for 1 person working. Now it is 1 person shifted (engineers, IT experts,...) for 10 person working before. The problem is that even dumb AI ("weak AI", i.e. automation, no intelligence) can replace an incredible number of workers. And while there are too few of the highly qualified people that are needed to keep this going, most people cannot get to that level of qualification. Not possible. And while there still is a need for, say, plumbers, electricians, etc., how many hours per month do you need one? Right. That is about the amount of people needed for those jobs in relation to the whole.
Fact is, around 10% of the population is more than enough to keep everything going. That is even already true today, but at this time the system is hugely inefficient. Automation does not only replace workers, it removes much of that inefficiency. And there is no way to prevent that from happening.
The whole thing comes with a huge socael problem though: We will have 3 classes of people in the future: 1. Those with essential work 2. Those with "nice to have" work and 3. Those without work. How to make sure the 3rd class is not going bonkers will be a huge challenge and giving them money to live is a rather small part of the solution. Most people cannot just watch TV all day, every day.
The thing you UBI-detractors do not understand is that the UBI is Universal. It has minimal overhead, quite unlikely the system in place now. That is the whole point. It just distributes mostly the same money as is distributed now, but with almost no bureaucracy and nobody falling through the cracks. It eliminates a lot of value-destruction the current system does and fosters.
However, there are certainly problems with an UBI. For example, many people will find it really difficult to live without work and not because of financial aspects. That is the real killer here. With an UBI, all the make-work things currently done will fall away and that is going to hit hard. Just look at how many people run into massive issues when they retire or how many dies soon afterwards. People need something to do and many cannot create that by and for themselves.
That said, an UBI will happen, there is no way around that. The numbers just do not add up any other way, unless the whole world agrees to go back to a non-tech model of society. That is not going to happen.
Crushing poverty is an absolute disgrace for a modern nation. There is no need at all for it economically either, because it is _expensive_ to have people live in poverty. The only thing it serves is to keep the middle-class in fear and timid. No other use.
Oh, it is. If you hold back automation domestically, it will just mean things get moved abroad where it is not held back. That is much, much worse. You can create a short-term straw-fire with attempts to hold back automation though (with a really nasty price to be paid later) and that seems to be what the current US administration is all about.
For your definition of "very easy" (which is utterly demented), that is true for any language and any coder. Your statement is a null-statement and worthless.
Likely both. Dumb people always think they know and understand everything. And when they are proven to be wrong in a non-ignorable fashion, then that is just a fluke to them. Ever tried to somebody really dumb that they do not get it? It is completely hopeless.
I don't think he has the mental capacity for it. We, as in "the human race", select the most dumb loudmouths to lead us. I do not see that changing, unfortunately.
Will be interesting to see what happens in the first mass-hack because of backdoored encryption. May take a while, but even the NSA has had their secrets stolen. If they cannot keep backdoor information secure, then nobody can. I can assure you however that no large financial institution will ever use backdoored encryption. They may be willing to hand over a disk with all the transaction willingly, but they will not allow themselves to be hacked that way by just anybody that tries. The risk that it would kill them is just far, far too big.
Unfortunately, for most of humanity, including basically all politicians and their fans, that is completely true. Only a small part of the human race qualifies as rational and these people do not seek power.
It is very difficult to write and maintain secure code in _any_ language. If you do not know that, then you have been living under a rock. This is not a language issue. This is an issue of the coders not understanding what they are doing.
Seriously, there is no silver bullet [Brooks, 1986] and Rust is not something that really matters to software security. At best, it will be a small incremental improvement, but only when it is a finished language with a stable feature-set and a clear understanding what it can _not_ do.
No. Most vulnerabilities in C are cases where the coder did not understand what they were doing. In C that often gives a memory safety issue, in Rust it will simply be something else that breaks. Your misunderstanding is that you think if the language prevents one type of error, than this error will go away and will not be replaced by something else. That is a wrong model and there is ample evidence that this model is wrong. This has been known for a long time. But, as usual, those that do not know or understand the history of computing are doomed to repeat it.
As long as we allow bad coders to write mission-critical code, the problem is not going to go away. Sure, there are other problems as well, as the one you describe, but the core problem is that people that have no business doing so are writing code. And this "code boot-camp" madness is making things worse.
For myself, I cant complain. While I usually do security consulting, I recently had a larger project coding some Apache modules in C (no other choice as there were both tight memory and speed constraints) and I did it at the full consulting rate. That boils down to around 2.3 times what that customer pays for a "coder" (Java monkey, really). Also, all the bad coders around assure my primary job. But the sheer stupidity of what is going on does offend me, in particular all those morons (this time "Rust morons") charging off screaming "Hurrah!" in the _known_ to be wrong direction.
Well, possibly. At the very least, academic coder education has gone down the drain with that utterly stupid focus on Java (probably not the worst language ever, but surely a close contender) that one encounters everywhere. But that does not change the fact that the problem are the coders and not the language. I am absolutely not opposed to restrict C coding to the places where it is necessary, but for these you need somebody that really knows what they are doing.
Bullshit from a proponent of yet another failed language, that was advertised with grande claims it could not deliver on. Kind of like Rust.
I will not even go into the details of how wrong the statements made are, there obviously is no point educating somebody that has gotten it this massively wrong.
So your argument is we should move to something new because it has not been proven that it is worse than the old thing? Are you suffering from dementia?
Rust has one of the most aggressive fanboy cultures since the Ruby hey days.
And that is just it. They have not understood the most fundamental thing about coding: "There is no silver bullet" [Brooks, 1986]. No language or tool will ever turn a bad coder into a good coder. No language will ever compensate for lack of insight on the part of the one using it. If you cannot hack it, then you are not a coder and that cannot be fixed, no matter how hard you cheer for a language that thinks will turn you into a good coder.
Rust is by far not the first failed attempt in that direction (albeit the one with the nastiest proponents so far). The problem is a fundamental one, which smart people have understood right from the beginning. But since there is an endless supply of non-smart people in the human race, it will not remain the last failed attempt. And when in, say 20 years, nobody cares about Rust anymore, C will still be around, because in the hands of an expert, it does its job exceptionally well. It just does not cater to incompetents, and that is why the Rust people do not like it: It forces them to realize what they are.
And that is exactly it. Sure, C is not that great a language, but it is simple, easy to understand in full (if you have what it takes) and it does not stand in your way. It also does not eat your cycles, unlike many, "modern" languages, but gives you native speed for what you are doing. As such, it works pretty well for anybody really competent. Insecure C code is not a language issue, it is a coder issue.
And that is just it. A C coder needs to be competent, or the code will be horrible. But the same happens in other languages, just that it is less obvious and that is a problem in itself. The problem are not the languages used, it is that there are far too many bad coders.
Formal verification has failed. It is one of many coding hypes that has. "There is no silver bullet" still applies and it does not look like it will go away, possibly not ever. Sure, formal verification has some niche uses and it is a nice tool to teach code semantics, but that is it.
Indeed. Use C for core algorithms, containers with tight space requirements, hardware access, etc. Use something modern for glue code, "business logic", UI, etc. like Python.
The problem here is not C, it is incompetent C coders. I actually thing using C is an advantage, because it makes incompetent coders far more obvious than, for example, Java, were even the most incompetent ones can hack something together. The second problem is however incompetent "managers" that do not fire incompetent C coders, because competent C coders are quite a bit more expensive. That they are worth it overall several times over is not something the bean-counters can grasp.
Slashdot Mirror
That era of the human race is over. We are now moving into post-industrialism. Many people have not noticed or are in denial, but there is no stopping it.
Indeed. With that many unemployed and on minimal or no benefits, society begins to crumble fast. The economy does too, because people need to be able to buy things to keep commerce going.
That is not an accurate model of what is happening now. We never ever had machinery before on the capability level that weak AI is currently reaching. At the same time, a lot of manufacturing technology is in a final state, and there is no next step, except eliminate the humans still in it.
Indeed. This is the first time in history that machinery is not just shifting the labor to higher qualified labor, it is replacing the labor. Well, there _is_ a shift, but before it was something like 1 person shifted for 1 person working. Now it is 1 person shifted (engineers, IT experts, ...) for 10 person working before. The problem is that even dumb AI ("weak AI", i.e. automation, no intelligence) can replace an incredible number of workers. And while there are too few of the highly qualified people that are needed to keep this going, most people cannot get to that level of qualification. Not possible. And while there still is a need for, say, plumbers, electricians, etc., how many hours per month do you need one? Right. That is about the amount of people needed for those jobs in relation to the whole.
Fact is, around 10% of the population is more than enough to keep everything going. That is even already true today, but at this time the system is hugely inefficient. Automation does not only replace workers, it removes much of that inefficiency. And there is no way to prevent that from happening.
The whole thing comes with a huge socael problem though: We will have 3 classes of people in the future: 1. Those with essential work 2. Those with "nice to have" work and 3. Those without work. How to make sure the 3rd class is not going bonkers will be a huge challenge and giving them money to live is a rather small part of the solution. Most people cannot just watch TV all day, every day.
The thing you UBI-detractors do not understand is that the UBI is Universal. It has minimal overhead, quite unlikely the system in place now. That is the whole point. It just distributes mostly the same money as is distributed now, but with almost no bureaucracy and nobody falling through the cracks. It eliminates a lot of value-destruction the current system does and fosters.
However, there are certainly problems with an UBI. For example, many people will find it really difficult to live without work and not because of financial aspects. That is the real killer here. With an UBI, all the make-work things currently done will fall away and that is going to hit hard. Just look at how many people run into massive issues when they retire or how many dies soon afterwards. People need something to do and many cannot create that by and for themselves.
That said, an UBI will happen, there is no way around that. The numbers just do not add up any other way, unless the whole world agrees to go back to a non-tech model of society. That is not going to happen.
Crushing poverty is an absolute disgrace for a modern nation. There is no need at all for it economically either, because it is _expensive_ to have people live in poverty. The only thing it serves is to keep the middle-class in fear and timid. No other use.
Oh, it is. If you hold back automation domestically, it will just mean things get moved abroad where it is not held back. That is much, much worse. You can create a short-term straw-fire with attempts to hold back automation though (with a really nasty price to be paid later) and that seems to be what the current US administration is all about.
For your definition of "very easy" (which is utterly demented), that is true for any language and any coder. Your statement is a null-statement and worthless.
I think this moron just has absolutely no idea what "mathematical law" means and that there is no way around it. Means he failed his basic education.
Likely both. Dumb people always think they know and understand everything. And when they are proven to be wrong in a non-ignorable fashion, then that is just a fluke to them. Ever tried to somebody really dumb that they do not get it? It is completely hopeless.
I don't think he has the mental capacity for it. We, as in "the human race", select the most dumb loudmouths to lead us. I do not see that changing, unfortunately.
Will be interesting to see what happens in the first mass-hack because of backdoored encryption. May take a while, but even the NSA has had their secrets stolen. If they cannot keep backdoor information secure, then nobody can. I can assure you however that no large financial institution will ever use backdoored encryption. They may be willing to hand over a disk with all the transaction willingly, but they will not allow themselves to be hacked that way by just anybody that tries. The risk that it would kill them is just far, far too big.
Unfortunately, for most of humanity, including basically all politicians and their fans, that is completely true. Only a small part of the human race qualifies as rational and these people do not seek power.
It is very difficult to write and maintain secure code in _any_ language. If you do not know that, then you have been living under a rock. This is not a language issue. This is an issue of the coders not understanding what they are doing.
Seriously, there is no silver bullet [Brooks, 1986] and Rust is not something that really matters to software security. At best, it will be a small incremental improvement, but only when it is a finished language with a stable feature-set and a clear understanding what it can _not_ do.
No. Most vulnerabilities in C are cases where the coder did not understand what they were doing. In C that often gives a memory safety issue, in Rust it will simply be something else that breaks. Your misunderstanding is that you think if the language prevents one type of error, than this error will go away and will not be replaced by something else. That is a wrong model and there is ample evidence that this model is wrong. This has been known for a long time. But, as usual, those that do not know or understand the history of computing are doomed to repeat it.
As long as we allow bad coders to write mission-critical code, the problem is not going to go away. Sure, there are other problems as well, as the one you describe, but the core problem is that people that have no business doing so are writing code. And this "code boot-camp" madness is making things worse.
For myself, I cant complain. While I usually do security consulting, I recently had a larger project coding some Apache modules in C (no other choice as there were both tight memory and speed constraints) and I did it at the full consulting rate. That boils down to around 2.3 times what that customer pays for a "coder" (Java monkey, really). Also, all the bad coders around assure my primary job. But the sheer stupidity of what is going on does offend me, in particular all those morons (this time "Rust morons") charging off screaming "Hurrah!" in the _known_ to be wrong direction.
Well, possibly. At the very least, academic coder education has gone down the drain with that utterly stupid focus on Java (probably not the worst language ever, but surely a close contender) that one encounters everywhere. But that does not change the fact that the problem are the coders and not the language. I am absolutely not opposed to restrict C coding to the places where it is necessary, but for these you need somebody that really knows what they are doing.
Bullshit from a proponent of yet another failed language, that was advertised with grande claims it could not deliver on. Kind of like Rust.
I will not even go into the details of how wrong the statements made are, there obviously is no point educating somebody that has gotten it this massively wrong.
So your argument is we should move to something new because it has not been proven that it is worse than the old thing? Are you suffering from dementia?
Indeed. Well said.
Rust has one of the most aggressive fanboy cultures since the Ruby hey days.
And that is just it. They have not understood the most fundamental thing about coding: "There is no silver bullet" [Brooks, 1986]. No language or tool will ever turn a bad coder into a good coder. No language will ever compensate for lack of insight on the part of the one using it. If you cannot hack it, then you are not a coder and that cannot be fixed, no matter how hard you cheer for a language that thinks will turn you into a good coder.
Rust is by far not the first failed attempt in that direction (albeit the one with the nastiest proponents so far). The problem is a fundamental one, which smart people have understood right from the beginning. But since there is an endless supply of non-smart people in the human race, it will not remain the last failed attempt. And when in, say 20 years, nobody cares about Rust anymore, C will still be around, because in the hands of an expert, it does its job exceptionally well. It just does not cater to incompetents, and that is why the Rust people do not like it: It forces them to realize what they are.
And that is exactly it. Sure, C is not that great a language, but it is simple, easy to understand in full (if you have what it takes) and it does not stand in your way. It also does not eat your cycles, unlike many, "modern" languages, but gives you native speed for what you are doing. As such, it works pretty well for anybody really competent. Insecure C code is not a language issue, it is a coder issue.
And that is just it. A C coder needs to be competent, or the code will be horrible. But the same happens in other languages, just that it is less obvious and that is a problem in itself. The problem are not the languages used, it is that there are far too many bad coders.
Formal verification has failed. It is one of many coding hypes that has. "There is no silver bullet" still applies and it does not look like it will go away, possibly not ever. Sure, formal verification has some niche uses and it is a nice tool to teach code semantics, but that is it.
Indeed. Use C for core algorithms, containers with tight space requirements, hardware access, etc. Use something modern for glue code, "business logic", UI, etc. like Python.
The problem here is not C, it is incompetent C coders. I actually thing using C is an advantage, because it makes incompetent coders far more obvious than, for example, Java, were even the most incompetent ones can hack something together. The second problem is however incompetent "managers" that do not fire incompetent C coders, because competent C coders are quite a bit more expensive. That they are worth it overall several times over is not something the bean-counters can grasp.