I meant "nigger". It does have a lot of baggage. Which is why I am curious if your "it's just words" statement still holds, or whether there are some exceptions to this rule. "Fuck" is generally deemed inappropriate for work use as well.
We should even use header files. We should just have all your code in a single file. All this fancy modularity bullshit just complicates things. I should just be able to compile a single source file into a single executable.
I find it weird that Linus is so against the concept of idiot-proofing code, considering he is apparently surrounded by idiots. In C++ land we can use integer wrapper classes that provide checks against overflows in arithmetic operations. And the code just looks the same. For example:
SafeInt c = a + b;
It's a bit slower, so maybe it doesn't belong in some very nested loop structure. I understand Linus wants the code to be fast, but in places where you are doing that overflow check anyway, it seems like having a class or function do it for you rather than writing your own check is probably better from a consistency standpoint.
I was intrigued by your post, and tried to find out if there were any good uses for optional that I could understand. I found one that it provides a way to distinguish types returned from hash.get() style methods (null that indicates key doesn't exist, or null that means the key maps to a null value).
I'm not really a java guru, and probably would not have arrived at this solution myself, but do you have any thoughts on this usage of Optional?
I have faith that once I finish typing this it will appear on the Slashdot website (do I know or hypothesize?).
I wouldn't say this is faith. You have no doubt posted comments on slashdot in the past, and so you have formed an expectation of what will occur in the future based on this evidence. This is different than someone who has never used a computer having faith in how a particular website will work.
Someone who has never been to heaven having faith that they will go to heaven is different than having faith that the sun will rise tomorrow.
Of course in religion these become more complicated, and I am willing to admit that I don't know if there is life after death but I act as if there is (or I act as if there is not). It does not even matter what I believe, it's how I act - some people say they believe in life after death but act as if that was not true, others say they are agnostics but act as if there was no life after death, without a provision that it may be untrue (if there's some probability of getting to hell shouldn't I go for a more careful approach if I really don't know, or do I have actual faith that there is nothing after death?).
I agree that how you behave is probably a more true representation of what you believe than what you say you believe (e.g. the preacher caught doing drugs with a hooker paid for by his parishioners).
I would say that you should absolutely take into account the >0 probability that heaven/hell exist. That said, acting as if heaven and hell exist is potentially costly in this life. Think of all the more fulfilling things you could be doing with your limited time on earth than studying some book of lies and wasting X hours every week listening to another person tell you about his interpretation of this book. This is what is known as opportunity cost to economists.
Secondly, you should also take into account that every religious claim from every religion has some >0 probability of being true, including any religious claims that you might imagine from yet invented religions. The possibilities are infinite.
The argument then becomes... Should you place more weight (i.e. have more faith) in one religion over another? And if so, what criteria should you use to base this decision on. Should you base this decision on evidence? (e.g. which religion has the most adherents, or which religion made the most correct predictions according to someone, etc), or are you going to base this decision on faith (e.g. having faith that the religion you were brought up in just happens to be the correct one, etc)?
I think there are objectively good reasons to believe things to some level of certainty and bad reasons to believe things to some level of certainty. If you don't have any good reasons to believe something to some level of certainty, but want to believe it anyway, you might be tempted to simply have faith.
For example, a gambler who has lost at roulette many times in a row may feel his chances are better now that his good luck is now "due". This is bad reasoning for estimating your chances to win the next game. You could show this gambler a history of all his previous bets and there would likely be good evidence to show that the law of independent assortment is a reality. He might decide your evidence is no good, and cite his own flawed evidence that he finds more convincing, or he may discount the value of logic and evidence altogether. The latter is what I consider to be faith.
When I am comparing military "sizes", I am comparing strength/capability, and using money spent as a yardstick. Obviously the literal physical volume or mass of a military is not important.
"Size" does not always refer to literal physical size (e.g. hard drive sizes).
And yes, spending almost as much money on our military as the entire rest of the world, and being the world's policeman, is as big government as it gets. I would also like to take this moment to say that traditionally America's conservative were non-interventionists, and some time in the late 20th century, the conservatives became pro-war.
There is nothing unreasonable about faith. Faith is the understanding that you will not be able to investigate the underlying mechanism of the thing you are dealing with.
False. That's not faith. That's simply acknowledgement of agnosticism.
Faith is taking one more step and believing something without evidence. It's taking something you don't know and simply switching it to the category of something you do know without actually learning anything new.
Most people have to exercise at least a little of that every day, when even when discussing scientific topics.
Science involves a lot of guessing. I don't think any part of science involves believing something is true without evidence. There's a difference. Science is very clear about when things count as hypotheses/conjectures, and when they count as theories/facts/etc (hint: it's after the experiment).
but faith is not the opposite of reason.
No faith is not the opposite of reason, for the same reason that dogs are not the opposite of cats. That doesn't mean that they are not mutually exclusive.
Science and the scientific method was the outgrowth of quite a few people, particularly minor clerics, who eventually developed the concepts of Reason and the scientific method.
I find this claim to be rather dubious.
Preachers and such may certainly be irrational, but their job description does not make it automatically so, nor does faith mean that they disable their "reason" to accept it.
I think their job description does imply some irrationality. Maybe they are not being irrational, but their job certainly involves inculcating others to be irrational. I would not assume that every preacher practices what they preach. It's quite common for them to exposed as conmen, and conmen are certainly acting rationally.
It is eminently reasonable to accept that there is something out there that you don't understand and can't investigate which, nevertheless, may be true.
That is reasonable. To then go that next step from "Something can be true even if we can't prove it" to "Something *is* true even if we can't prove it" is not reasonable.
Certainly, the concept of things like atoms and smaller particles were an object of speculation without the ability to investigate for thousands of years before we could design experiments for them. I wouldn't have called those ancient philosophers "irrational".
It would have been irrational to say "Atoms must exist even though we have no evidence", even if it turned out to be true.
I can say "I know this next die roll will be a 6". That is irrational even if the next die roll is a 6.
But ancient philosophers were not basing their speculation purely on faith. They were using logic to come to those conclusions. They were evaluating the reasonableness of substances being infinitely divisible vs. there being some indivisible smallest unit.
Not all evidence is empirical. Other forms of theoretical evidence like logical and mathematical consistency, are also valid sources of evidence.
I would also like to point out that since discovering "the atom" we have discovered subatomic particles like protons and neutrons, and even smaller particles like quarks which compose them. We also potentially have yet another level deeper to go with string/M theory.
As logical as it may sound that there be a smallest indivisible particle, it may in fact be the case (as pointed out by Feynman and others) that there may not even be a bottom. It could be smaller and smaller sub-particles forever in an infinite regression.
My point is this: There is a big difference between saying
1. I don't have any evidence, but I think maybe X is true (agnosticism + hypothesis)
2. I don't have any evidence, but I know X is true (faith)
One of those is compatible with science and one is not.
Most uses of templates are provided by the STL and there is not much need for them in normal programming. That is not to say there are not instances for their use, but there are often better design patterns to be following.
Maybe I'm not exactly sure what "normal programming" is, but I have become a lot better at templates in the last year or so, and I find more and more uses for them, where I probably would have used other language features.
I get the impression that because templates are somewhat hard to use, people avoid them for all but the most obvious cases.
I really can't think of any cases where I saw a template and thought "They really should have done X instead".
If you can think of any, I'd love to see an example.
Imagine the wasted money, time, and human effort that could be saved from not rewriting the same books every few years. Imagine all the good that could be achieved not just in California, or the US, but the entire world, if textbooks where open source. The only losers would be the textbook publishers and those receiving their kickbacks.
I have a plan for that. You pick a name that's toxic to Republicans. My plan is for non-shitty libertarians to call themselves "liberals". Republicans shit on this label so much that democrats don't want it anymore. Think of all the bumper stickers republicans would have to change if they tried to co-opt "liberal".
I still think "liberal" is a great label. It references to the lineage of libertarians to classical liberalism. And it's vacant. Nobody wants this label anymore. It's pejorative. I say we co-opt it back from "progressives". I will wear a Republican pejorative label as a badge of honor.
Just because some people in a particular group happen to be shitty people doesn't mean all people in that particular group are shitty people.
Obviously... Myself being a libertarian, I wouldn't hold the position that *all* libertarians are shitty. But I also find myself in a unique position of being surrounded by shitheads who don't give a fuck about liberty calling themselves the same label as me.
They might care about freedoms for themselves or their own demographics, but they don't really care about freedom for others and may actually support restrictions on the freedoms of other groups and individuals.
There are plenty of "libertarians" who want to lower taxes, but don't particularly care about a persons right to put whatever they want into their own bodies, or the equal rights for non-straight people.
I can guarantee that this bothers me more than most everybody. I was there when republicans co-opted the libertarian party. I remember when Bob Fucking Bar and Wayne Allen Root won the nomination for president from the libertarian party (Both of whom have now gone back to being republicans).
I remember when the tea party started out as a grass roots campaign that was opposed to GW Bush and Fox News. I remember when they formed and angry mob and chased Sean Hannity down the street (not that I endorse that). Now I get newsletter emails from Sarah Palin, Anne Coulter, Michelle Malkin, Charles Krauthammer, and all these fuck faces that are the new face of libertarianism.
No it's not all shitty people. I'm still here, explaining to people how even though I am a libertarian, I am not friends with Sarah Palin and Anne Coulter. But I honestly don't know if there is any point to it, and I might just give up at some point.
Depending on how you use the templates, they can actually lead to less bloat on embedded systems. You can use templates to do zero overhead static polymorphism. This is actually quite useful in safety critical code where the non-deterministic behavior of standard vtable lookups is a problem.
In fact, writing deterministic safety critical code (usually on embedded systems), often forces you into using templates for polymorphism rather than traditional inheritance.
It's hard to get speed, correctness, and proper abstraction all in one language. C++ does probably the best job at coming up with a good compromise of all 3, and I don't think that would be true without templates.
Vtables really do slow things down quite a bit. Luckily C++ gives you the option of templates as a fast (but somewhat limited) alternative to vtables.
I find templates to be one of the most useful features of c++. Although it may not be it's most obvious usage, through templates, you can get zero overhead versions of object oriented programming minus the ability to dynamic cast (static polymorphism).
I like the exception pattern when used locally (i.e. within a function), but I think the overhead of actually excepting makes usage of this mechanism undesirable. Although I think compilers may one day (or may already) be able to optimize out the overhead of exceptions when used in a limited scope.
How do you even do something like this:
template <typename T> T min(T x, T y) { return x < y ? x : y; }
I agree that's what it should be used for. But when you compare it to other tools that fit in the same niche, it's just not as good. I'd probably still be using perl if python didn't exist. It's not like python is a perfect language either, but for like 9 times out of 10 python's way of doing something is better than all 20 ways of doing it in perl.
And I'm not even one of these language evangelists. I normally can't stand when people pretend their personal preferences for languages are objective facts (e.g. java vs. c++ vs. c#), but I've found that most people who use perl have never used python.
The military is one of the few constitutionally mandated missions of the central government. Welfare is not. Do you see the difference?
Where in he constitution does it say that the military needs to be almost the size of all the other militaries in the world combined?
The rest of the issues you talk of are not federal issues, but local ones.
So republicans are for a small federal government, but a big state government with lots of regulations? Republicans are for 50 individual nanny states? It's ok if the government takes away your freedoms as long as it's a state government?
Nobody (at least not anyone I know) is arguing that C++ is a perfect language. But it's pretty good at getting things done in a way that is fairly pleasing to a lot of people which is why it's so popular. There's lots of nasty language features, but you don't have to use them. It's mostly backwards compatible with C. It's fast. It's versatile. And it has a lot of community support and 3rd party libraries and frameworks that remove a lot of the nastiness.
Nearly everyone who writes good C++ is using a subset of the full C++ language. And some of those subsets are pretty good languages.
Because republicans are not actually for smaller government. They are for making the parts they don't like smaller and the parts they do like bigger (just like the democrats). How many republicans are for shrinking the military? How many republicans are for ending the war on drugs? How many republicans support legalizing prostitution? How many republicans support removing marriage not just from federal jurisdiction, but from all government jurisdiction?
It's not bad for writing quick and dirty scripts. It's just that there are better alternatives. A car with no doors is not bad at doing it's job of getting you from point A to point B, but there are options that are better.
The things that make perl good are not unique, and the things that make perl unique are not good.
left vs. right are also artificial labels. Yes, "liberals" are associated with "the left", so what? Is it the case that only "the left" is opposed to freedom of speech? No it's just that the left and the right oppose freedom of speech in different ways that suit them.
Why bother with ruby when perl has served us so well for so long?
Ruby is already a dead language and perl has always been a terrible language and should be murdered. I say this as a person who actually liked perl back when it was the first and only scripting language I knew.
What is true, though, is they sometimes don't understand the appeal of the new stuff, nor why anyone would consider using it. After all, when it comes down to it most new approaches don't really accomplish anything that the "old way" cannot... at least from the perspective of an IT professional.
I am a software developer. I do a lot of cutting edge stuff in my spare time (cloud, nosql, etc), and a lot of low level c++ 2003 at work. There are a lot of people at my work who work in C and don't even see the point of migrating to C++ as they can do everything they need to do in C, without all this bullshit object oriented crap. They would claim to understand C++, but not the appeal, as it seems to them to be far easier to do everything in C. The fact that they don't see the appeal is evidenced that they don't understand that c++ actually makes doing things correctly (i.e. maintainable, scalable, modular, etc) easier, maybe because they don't even see the appeal of those features.
And this is the resistance to getting people to go from a 4 decade old technology to a 3 decade old technology
Slashdot Mirror
I meant "nigger". It does have a lot of baggage. Which is why I am curious if your "it's just words" statement still holds, or whether there are some exceptions to this rule. "Fuck" is generally deemed inappropriate for work use as well.
What about a competent person who said the n-word all the time?
We should even use header files. We should just have all your code in a single file. All this fancy modularity bullshit just complicates things. I should just be able to compile a single source file into a single executable.
agreed
kernel development has a different set of values and priorities.
I find it weird that Linus is so against the concept of idiot-proofing code, considering he is apparently surrounded by idiots. In C++ land we can use integer wrapper classes that provide checks against overflows in arithmetic operations. And the code just looks the same. For example:
SafeInt c = a + b;
It's a bit slower, so maybe it doesn't belong in some very nested loop structure. I understand Linus wants the code to be fast, but in places where you are doing that overflow check anyway, it seems like having a class or function do it for you rather than writing your own check is probably better from a consistency standpoint.
I was intrigued by your post, and tried to find out if there were any good uses for optional that I could understand. I found one that it provides a way to distinguish types returned from hash.get() style methods (null that indicates key doesn't exist, or null that means the key maps to a null value).
I'm not really a java guru, and probably would not have arrived at this solution myself, but do you have any thoughts on this usage of Optional?
I have faith that once I finish typing this it will appear on the Slashdot website (do I know or hypothesize?).
I wouldn't say this is faith. You have no doubt posted comments on slashdot in the past, and so you have formed an expectation of what will occur in the future based on this evidence. This is different than someone who has never used a computer having faith in how a particular website will work.
Someone who has never been to heaven having faith that they will go to heaven is different than having faith that the sun will rise tomorrow.
Of course in religion these become more complicated, and I am willing to admit that I don't know if there is life after death but I act as if there is (or I act as if there is not). It does not even matter what I believe, it's how I act - some people say they believe in life after death but act as if that was not true, others say they are agnostics but act as if there was no life after death, without a provision that it may be untrue (if there's some probability of getting to hell shouldn't I go for a more careful approach if I really don't know, or do I have actual faith that there is nothing after death?).
I agree that how you behave is probably a more true representation of what you believe than what you say you believe (e.g. the preacher caught doing drugs with a hooker paid for by his parishioners).
I would say that you should absolutely take into account the >0 probability that heaven/hell exist. That said, acting as if heaven and hell exist is potentially costly in this life. Think of all the more fulfilling things you could be doing with your limited time on earth than studying some book of lies and wasting X hours every week listening to another person tell you about his interpretation of this book. This is what is known as opportunity cost to economists.
Secondly, you should also take into account that every religious claim from every religion has some >0 probability of being true, including any religious claims that you might imagine from yet invented religions. The possibilities are infinite.
The argument then becomes... Should you place more weight (i.e. have more faith) in one religion over another? And if so, what criteria should you use to base this decision on. Should you base this decision on evidence? (e.g. which religion has the most adherents, or which religion made the most correct predictions according to someone, etc), or are you going to base this decision on faith (e.g. having faith that the religion you were brought up in just happens to be the correct one, etc)?
I think there are objectively good reasons to believe things to some level of certainty and bad reasons to believe things to some level of certainty. If you don't have any good reasons to believe something to some level of certainty, but want to believe it anyway, you might be tempted to simply have faith.
For example, a gambler who has lost at roulette many times in a row may feel his chances are better now that his good luck is now "due". This is bad reasoning for estimating your chances to win the next game. You could show this gambler a history of all his previous bets and there would likely be good evidence to show that the law of independent assortment is a reality. He might decide your evidence is no good, and cite his own flawed evidence that he finds more convincing, or he may discount the value of logic and evidence altogether. The latter is what I consider to be faith.
When I am comparing military "sizes", I am comparing strength/capability, and using money spent as a yardstick. Obviously the literal physical volume or mass of a military is not important.
"Size" does not always refer to literal physical size (e.g. hard drive sizes).
And yes, spending almost as much money on our military as the entire rest of the world, and being the world's policeman, is as big government as it gets. I would also like to take this moment to say that traditionally America's conservative were non-interventionists, and some time in the late 20th century, the conservatives became pro-war.
There is nothing unreasonable about faith. Faith is the understanding that you will not be able to investigate the underlying mechanism of the thing you are dealing with.
False. That's not faith. That's simply acknowledgement of agnosticism.
Faith is taking one more step and believing something without evidence. It's taking something you don't know and simply switching it to the category of something you do know without actually learning anything new.
Most people have to exercise at least a little of that every day, when even when discussing scientific topics.
Science involves a lot of guessing. I don't think any part of science involves believing something is true without evidence. There's a difference. Science is very clear about when things count as hypotheses/conjectures, and when they count as theories/facts/etc (hint: it's after the experiment).
but faith is not the opposite of reason.
No faith is not the opposite of reason, for the same reason that dogs are not the opposite of cats. That doesn't mean that they are not mutually exclusive.
Science and the scientific method was the outgrowth of quite a few people, particularly minor clerics, who eventually developed the concepts of Reason and the scientific method.
I find this claim to be rather dubious.
Preachers and such may certainly be irrational, but their job description does not make it automatically so, nor does faith mean that they disable their "reason" to accept it.
I think their job description does imply some irrationality. Maybe they are not being irrational, but their job certainly involves inculcating others to be irrational. I would not assume that every preacher practices what they preach. It's quite common for them to exposed as conmen, and conmen are certainly acting rationally.
It is eminently reasonable to accept that there is something out there that you don't understand and can't investigate which, nevertheless, may be true.
That is reasonable. To then go that next step from "Something can be true even if we can't prove it" to "Something *is* true even if we can't prove it" is not reasonable.
Certainly, the concept of things like atoms and smaller particles were an object of speculation without the ability to investigate for thousands of years before we could design experiments for them. I wouldn't have called those ancient philosophers "irrational".
It would have been irrational to say "Atoms must exist even though we have no evidence", even if it turned out to be true.
I can say "I know this next die roll will be a 6". That is irrational even if the next die roll is a 6.
But ancient philosophers were not basing their speculation purely on faith. They were using logic to come to those conclusions. They were evaluating the reasonableness of substances being infinitely divisible vs. there being some indivisible smallest unit.
Not all evidence is empirical. Other forms of theoretical evidence like logical and mathematical consistency, are also valid sources of evidence.
I would also like to point out that since discovering "the atom" we have discovered subatomic particles like protons and neutrons, and even smaller particles like quarks which compose them. We also potentially have yet another level deeper to go with string/M theory.
As logical as it may sound that there be a smallest indivisible particle, it may in fact be the case (as pointed out by Feynman and others) that there may not even be a bottom. It could be smaller and smaller sub-particles forever in an infinite regression.
My point is this: There is a big difference between saying
1. I don't have any evidence, but I think maybe X is true (agnosticism + hypothesis)
2. I don't have any evidence, but I know X is true (faith)
One of those is compatible with science and one is not.
Many clergy have very reasonable approaches to convincing the faithful to give them their money. Cults included.
Most uses of templates are provided by the STL and there is not much need for them in normal programming. That is not to say there are not instances for their use, but there are often better design patterns to be following.
Maybe I'm not exactly sure what "normal programming" is, but I have become a lot better at templates in the last year or so, and I find more and more uses for them, where I probably would have used other language features.
I get the impression that because templates are somewhat hard to use, people avoid them for all but the most obvious cases.
I really can't think of any cases where I saw a template and thought "They really should have done X instead".
If you can think of any, I'd love to see an example.
Imagine the wasted money, time, and human effort that could be saved from not rewriting the same books every few years. Imagine all the good that could be achieved not just in California, or the US, but the entire world, if textbooks where open source. The only losers would be the textbook publishers and those receiving their kickbacks.
I have a plan for that. You pick a name that's toxic to Republicans. My plan is for non-shitty libertarians to call themselves "liberals". Republicans shit on this label so much that democrats don't want it anymore. Think of all the bumper stickers republicans would have to change if they tried to co-opt "liberal".
I still think "liberal" is a great label. It references to the lineage of libertarians to classical liberalism. And it's vacant. Nobody wants this label anymore. It's pejorative. I say we co-opt it back from "progressives". I will wear a Republican pejorative label as a badge of honor.
Just because some people in a particular group happen to be shitty people doesn't mean all people in that particular group are shitty people.
Obviously... Myself being a libertarian, I wouldn't hold the position that *all* libertarians are shitty. But I also find myself in a unique position of being surrounded by shitheads who don't give a fuck about liberty calling themselves the same label as me.
They might care about freedoms for themselves or their own demographics, but they don't really care about freedom for others and may actually support restrictions on the freedoms of other groups and individuals.
There are plenty of "libertarians" who want to lower taxes, but don't particularly care about a persons right to put whatever they want into their own bodies, or the equal rights for non-straight people.
I can guarantee that this bothers me more than most everybody. I was there when republicans co-opted the libertarian party. I remember when Bob Fucking Bar and Wayne Allen Root won the nomination for president from the libertarian party (Both of whom have now gone back to being republicans).
I remember when the tea party started out as a grass roots campaign that was opposed to GW Bush and Fox News. I remember when they formed and angry mob and chased Sean Hannity down the street (not that I endorse that). Now I get newsletter emails from Sarah Palin, Anne Coulter, Michelle Malkin, Charles Krauthammer, and all these fuck faces that are the new face of libertarianism.
No it's not all shitty people. I'm still here, explaining to people how even though I am a libertarian, I am not friends with Sarah Palin and Anne Coulter. But I honestly don't know if there is any point to it, and I might just give up at some point.
Depending on how you use the templates, they can actually lead to less bloat on embedded systems. You can use templates to do zero overhead static polymorphism. This is actually quite useful in safety critical code where the non-deterministic behavior of standard vtable lookups is a problem.
In fact, writing deterministic safety critical code (usually on embedded systems), often forces you into using templates for polymorphism rather than traditional inheritance.
It's hard to get speed, correctness, and proper abstraction all in one language. C++ does probably the best job at coming up with a good compromise of all 3, and I don't think that would be true without templates.
Vtables really do slow things down quite a bit. Luckily C++ gives you the option of templates as a fast (but somewhat limited) alternative to vtables.
I find templates to be one of the most useful features of c++. Although it may not be it's most obvious usage, through templates, you can get zero overhead versions of object oriented programming minus the ability to dynamic cast (static polymorphism).
I like the exception pattern when used locally (i.e. within a function), but I think the overhead of actually excepting makes usage of this mechanism undesirable. Although I think compilers may one day (or may already) be able to optimize out the overhead of exceptions when used in a limited scope.
How do you even do something like this:
template <typename T> T min(T x, T y) { return x < y ? x : y; }
without templates? macros?
I agree that's what it should be used for. But when you compare it to other tools that fit in the same niche, it's just not as good. I'd probably still be using perl if python didn't exist. It's not like python is a perfect language either, but for like 9 times out of 10 python's way of doing something is better than all 20 ways of doing it in perl.
And I'm not even one of these language evangelists. I normally can't stand when people pretend their personal preferences for languages are objective facts (e.g. java vs. c++ vs. c#), but I've found that most people who use perl have never used python.
The military is one of the few constitutionally mandated missions of the central government. Welfare is not. Do you see the difference?
Where in he constitution does it say that the military needs to be almost the size of all the other militaries in the world combined?
The rest of the issues you talk of are not federal issues, but local ones.
So republicans are for a small federal government, but a big state government with lots of regulations? Republicans are for 50 individual nanny states? It's ok if the government takes away your freedoms as long as it's a state government?
Nobody (at least not anyone I know) is arguing that C++ is a perfect language. But it's pretty good at getting things done in a way that is fairly pleasing to a lot of people which is why it's so popular. There's lots of nasty language features, but you don't have to use them. It's mostly backwards compatible with C. It's fast. It's versatile. And it has a lot of community support and 3rd party libraries and frameworks that remove a lot of the nastiness.
Nearly everyone who writes good C++ is using a subset of the full C++ language. And some of those subsets are pretty good languages.
Because republicans are not actually for smaller government. They are for making the parts they don't like smaller and the parts they do like bigger (just like the democrats). How many republicans are for shrinking the military? How many republicans are for ending the war on drugs? How many republicans support legalizing prostitution? How many republicans support removing marriage not just from federal jurisdiction, but from all government jurisdiction?
It's not bad for writing quick and dirty scripts. It's just that there are better alternatives. A car with no doors is not bad at doing it's job of getting you from point A to point B, but there are options that are better.
The things that make perl good are not unique, and the things that make perl unique are not good.
left vs. right are also artificial labels. Yes, "liberals" are associated with "the left", so what? Is it the case that only "the left" is opposed to freedom of speech? No it's just that the left and the right oppose freedom of speech in different ways that suit them.
Why bother with ruby when perl has served us so well for so long?
Ruby is already a dead language and perl has always been a terrible language and should be murdered. I say this as a person who actually liked perl back when it was the first and only scripting language I knew.
What is true, though, is they sometimes don't understand the appeal of the new stuff, nor why anyone would consider using it. After all, when it comes down to it most new approaches don't really accomplish anything that the "old way" cannot... at least from the perspective of an IT professional.
I am a software developer. I do a lot of cutting edge stuff in my spare time (cloud, nosql, etc), and a lot of low level c++ 2003 at work. There are a lot of people at my work who work in C and don't even see the point of migrating to C++ as they can do everything they need to do in C, without all this bullshit object oriented crap. They would claim to understand C++, but not the appeal, as it seems to them to be far easier to do everything in C. The fact that they don't see the appeal is evidenced that they don't understand that c++ actually makes doing things correctly (i.e. maintainable, scalable, modular, etc) easier, maybe because they don't even see the appeal of those features.
And this is the resistance to getting people to go from a 4 decade old technology to a 3 decade old technology
Well stop it, you're making everyone else look bad.