I second that. SJWs are annoying and may pose real danger, but they are not interesting at all, as they just rehash millennia-old arrogance, greed and stupidity.
And then, when all competent contenders are out of the picture, prices raise and quality drops. This is _not_ a problem where a capitalist competitive approach is a good idea, as this is not about standardized products that a lot of people can produce.
Um... WHY? There's been a lot of studies showing that emotions dampen critical thought and vice versa. Give an engineer a problem then leave them alone until they come back with an answer.
Exactly, and while people will need different amounts of time, once you do this several times with the same people to eliminate flukes, you will identify two classes: One that comes back with a working solution most of the time and one that does not. The former are the good engineers and the latter are the bad ones.
Seriously, whether somebody takes a week or a month to solve a problem difficult enough that you cannot look it up does not matter much. The kicker is whether they can or cannot solve it. Coding is not speed-thinking, unless you do really low-quality stuff.
Not long ago, Max Howell, the author of Homebrew (software that basically every engineer with a Mac uses), famously quipped about being rejected from Google after being unable to invert a binary tree.
What does that even mean? You can traverse it left-first or right-first, but "inverted"?
I went through a Google interview, and I thought the questions were really stupid. I think I gave them quite a few answers that likely were wrong in their eyes, because I had too much experience with the subjects. For example, when they asked me how to do a hash-function, clearly expecting one of the standard (pretty bad) constructs. I told them to use the functions by Bob Jenkins, or, if there was time, a full-blown crypto hash. Now, I have filled hash-tables with 100 million elements and got collision chains up to 200 elements long with the STL hash function, but only 30 with SpookyHash by Jenkins. And if there is a spinning disk access in there, the 10us or so a crypto-hash costs you is not a problem either, and the randomization will be excellent under all conditions. But my impression was that they though I was evading the question because I did not really know how this works. That s a pretty bad fail on their side. There were several more. I think the real problem was that I had actual hands-on experience with almost everything they asked me, while they expected me to work though the questions from the data they gave me.
The problem hence is that these questions prefer people with some, but not too deep knowledge or actual experience. As soon as you know more, your chances of failing increase. That is really stupid.
Incidentally, I know a few ex-Googlers now and I am pretty glad they did not hire me. Many people there are not nearly as smart as they think they are and the 20% time is more of a way to press even more working hours out of employees. They kept pestering me for a few years to re-interview, until I told them, sure, no problem, my daily fee is $1600 and I will be happy to do more interviews if you pay for my time. That finally go the message across.
Incidentally, this may just cut down on the part of the problem created by common criminals. Their "business opportunity" just vanished. Now we mainly have to worry about state-sponsored and employed terrorists, like certain employees of the NSA, GCHQ, Chinese and Russian intelligence, etc.
No. Expect this to maybe have an impact during the next 100 years or so. Mathematicians are not like the stupid masses that want everything to pay off _now_. They see the value in doing things with a very long-term perspective.
The assumption P!=NP is a shortcut that simplifies things. But if I have, say, some computation that is O(n) with key and Omega(n^3) without the key and scales, then I can still do public-key crypto with it, just slower. Now, if it turns out that P=NP (unlikely), some things will need to be changed as a whole class of computations would not be ensured to be exponential anymore, but it does not break things fundamentally. Same if some problems used for public-key crypto turn out to be in P, not NP. The idea is still sound, it just makes things less convenient.
Well said. Otherwise you could just lock the president away in a box or not even power up the computer. Prefect security, perfectly useless. In fact a major part of being a security expert is explaining this to one sort of people. The other sort you have to explain to that some level of security (always based on what the risks in the concrete situation are) is pretty necessary.
Both black and white are entirely wrong and useless in the security space. It is all about finding the most useful shade of gray.
Incompetent and unaware of it. This person qualifies. For these idiots, it is always others that make the mistakes, never they themselves, and hence they never produce anything good because they do not learn.
As a security expert, I fully agree. Security is something that you need to think about from the beginning, but you only ever need enough that your residual risks are acceptable.
These "critics" often do not get how to do professional risk management (Linus does) and, quite often, I get the impression they do not have any significant coding experience, as they seem to think the changes they would like are easy to implement. I run into these black vs. white people in security quite frequently. These are the amateurs that do not understand that actually building things that work is already very, very hard and if you keep changing things all the time you just end with a dysfunctional, insecure mess. Also, you want a stable product, you incorporate research results only after they have been tested out in practice for a few years and only if they bring you a significant gain.
The Linux kernel has an excellent security track record in its core. Some drivers are not that good, but that is why if you need high security, you only compile those that you really need.
Some people working as software architects, designers, implementers actually have what it takes to be called "engineers". That means a certain type of education, skills, experience, etc. Most do not. But that is because most coders are really, really bad at their job and should have gone for a different type of job altogether.
The argument about bridges and buildings is bogus though: The problem why software is often unreliable and insecure is due to budget constraints, project management issues, and, yes, the wrong people working on it. But other clear engineering areas have the same: Many electronic gadgets are not much better, yet there clearly were some EEs involved in creating them and these people are in fact and also formally engineers.
It is completely immaterial who Snowden is. What is important is what the NSA and the GCHQ have been doing. And nobody sane is claiming the documents on that are faked.
If fall-of-the-wall was in there, then it was the Stasi (hard to keep track of all that fascist scum, I know).
One important factor of fascism is to make sure the population is mostly in their side. As people are stupid and external enemies can be easily created (just look at the US today, or northern Korea, same principle), this is pretty easy.
Or not much of it. Otherwise somebody would have used that ineffectiveness by now. The people that established and are supporting the continued existence of the TSA must know that. My guess is this is a field study on how much loss of freedom US citizens are willing to accept when that loss is justified with bogus arguments. Incidentally, the same thing is being done in Britain, and the British population seems to not mind living in a police- and surveillance-state.
So ignoring the law and becoming a criminal organization is fine as long as there is some abstract goal behind it? You would the also be perfectly fine with what the KGB, the Stasi and the GeStaPo were doing?
Slashdot Mirror
But that is a very capitalist idea! Just outsource it!
I second that. SJWs are annoying and may pose real danger, but they are not interesting at all, as they just rehash millennia-old arrogance, greed and stupidity.
And then, when all competent contenders are out of the picture, prices raise and quality drops. This is _not_ a problem where a capitalist competitive approach is a good idea, as this is not about standardized products that a lot of people can produce.
Nice quote!
Not only that, when you have good data models, good interfaces, well-structured code etc. then a fix will be easy to do.
Um... WHY? There's been a lot of studies showing that emotions dampen critical thought and vice versa. Give an engineer a problem then leave them alone until they come back with an answer.
Exactly, and while people will need different amounts of time, once you do this several times with the same people to eliminate flukes, you will identify two classes: One that comes back with a working solution most of the time and one that does not. The former are the good engineers and the latter are the bad ones.
Seriously, whether somebody takes a week or a month to solve a problem difficult enough that you cannot look it up does not matter much. The kicker is whether they can or cannot solve it. Coding is not speed-thinking, unless you do really low-quality stuff.
Not long ago, Max Howell, the author of Homebrew (software that basically every engineer with a Mac uses), famously quipped about being rejected from Google after being unable to invert a binary tree.
What does that even mean? You can traverse it left-first or right-first, but "inverted"?
I went through a Google interview, and I thought the questions were really stupid. I think I gave them quite a few answers that likely were wrong in their eyes, because I had too much experience with the subjects. For example, when they asked me how to do a hash-function, clearly expecting one of the standard (pretty bad) constructs. I told them to use the functions by Bob Jenkins, or, if there was time, a full-blown crypto hash. Now, I have filled hash-tables with 100 million elements and got collision chains up to 200 elements long with the STL hash function, but only 30 with SpookyHash by Jenkins. And if there is a spinning disk access in there, the 10us or so a crypto-hash costs you is not a problem either, and the randomization will be excellent under all conditions. But my impression was that they though I was evading the question because I did not really know how this works. That s a pretty bad fail on their side. There were several more. I think the real problem was that I had actual hands-on experience with almost everything they asked me, while they expected me to work though the questions from the data they gave me.
The problem hence is that these questions prefer people with some, but not too deep knowledge or actual experience. As soon as you know more, your chances of failing increase. That is really stupid.
Incidentally, I know a few ex-Googlers now and I am pretty glad they did not hire me. Many people there are not nearly as smart as they think they are and the 20% time is more of a way to press even more working hours out of employees. They kept pestering me for a few years to re-interview, until I told them, sure, no problem, my daily fee is $1600 and I will be happy to do more interviews if you pay for my time. That finally go the message across.
Incidentally, this may just cut down on the part of the problem created by common criminals. Their "business opportunity" just vanished. Now we mainly have to worry about state-sponsored and employed terrorists, like certain employees of the NSA, GCHQ, Chinese and Russian intelligence, etc.
No. Expect this to maybe have an impact during the next 100 years or so. Mathematicians are not like the stupid masses that want everything to pay off _now_. They see the value in doing things with a very long-term perspective.
The assumption P!=NP is a shortcut that simplifies things. But if I have, say, some computation that is O(n) with key and Omega(n^3) without the key and scales, then I can still do public-key crypto with it, just slower. Now, if it turns out that P=NP (unlikely), some things will need to be changed as a whole class of computations would not be ensured to be exponential anymore, but it does not break things fundamentally. Same if some problems used for public-key crypto turn out to be in P, not NP. The idea is still sound, it just makes things less convenient.
That said, this is cool research!
Well said. Otherwise you could just lock the president away in a box or not even power up the computer. Prefect security, perfectly useless. In fact a major part of being a security expert is explaining this to one sort of people. The other sort you have to explain to that some level of security (always based on what the risks in the concrete situation are) is pretty necessary.
Both black and white are entirely wrong and useless in the security space. It is all about finding the most useful shade of gray.
Incompetent and unaware of it. This person qualifies. For these idiots, it is always others that make the mistakes, never they themselves, and hence they never produce anything good because they do not learn.
That is complete nonsense. Any UNIX-like OS comes with a lot of security considerations right from the beginning.
Aaaaaand fail. This discussion is about security, not safety. These are two entirely different things.
As a security expert, I fully agree. Security is something that you need to think about from the beginning, but you only ever need enough that your residual risks are acceptable.
These "critics" often do not get how to do professional risk management (Linus does) and, quite often, I get the impression they do not have any significant coding experience, as they seem to think the changes they would like are easy to implement. I run into these black vs. white people in security quite frequently. These are the amateurs that do not understand that actually building things that work is already very, very hard and if you keep changing things all the time you just end with a dysfunctional, insecure mess. Also, you want a stable product, you incorporate research results only after they have been tested out in practice for a few years and only if they bring you a significant gain.
The Linux kernel has an excellent security track record in its core. Some drivers are not that good, but that is why if you need high security, you only compile those that you really need.
Some people working as software architects, designers, implementers actually have what it takes to be called "engineers". That means a certain type of education, skills, experience, etc. Most do not. But that is because most coders are really, really bad at their job and should have gone for a different type of job altogether.
The argument about bridges and buildings is bogus though: The problem why software is often unreliable and insecure is due to budget constraints, project management issues, and, yes, the wrong people working on it. But other clear engineering areas have the same: Many electronic gadgets are not much better, yet there clearly were some EEs involved in creating them and these people are in fact and also formally engineers.
You are an idiot. Or a paid shill.
It is completely immaterial who Snowden is. What is important is what the NSA and the GCHQ have been doing. And nobody sane is claiming the documents on that are faked.
If fall-of-the-wall was in there, then it was the Stasi (hard to keep track of all that fascist scum, I know).
One important factor of fascism is to make sure the population is mostly in their side. As people are stupid and external enemies can be easily created (just look at the US today, or northern Korea, same principle), this is pretty easy.
Or not much of it. Otherwise somebody would have used that ineffectiveness by now. The people that established and are supporting the continued existence of the TSA must know that. My guess is this is a field study on how much loss of freedom US citizens are willing to accept when that loss is justified with bogus arguments. Incidentally, the same thing is being done in Britain, and the British population seems to not mind living in a police- and surveillance-state.
So ignoring the law and becoming a criminal organization is fine as long as there is some abstract goal behind it? You would the also be perfectly fine with what the KGB, the Stasi and the GeStaPo were doing?
Obviously Biology is sexist. Time to end that science and all its subjects!
Labeling is not bad. What people associate with labels is.
What are you talking about, Opera 12.17 being actively blocked? I have not run into that and I still use it for most browsing.
For really incompetent values of "web designers", i.e. the average one.