I concur. Just ask the wine developers about how complete and accurate MS documentation is. They are reverse engineering a whole lot just because the documentation sucks in a way.
Walmart may have a market cap of 50% yearly revenue because they have razor-thin margins when you compare them with Apple. They have got much less of a cushion should things turn bad, compared to Apple.
Yep. See, for example, javascript memory leaks in Firefox. And those leaks are not any sort of an underlying C/C++ problem, just that it's very simple to hold on to a reference to a DOM that has a 100mb compartment on top of it. Plug-ins seem to do it left and right, and then Firefox slows to a crawl -- no wonder.
It's not about what's best -- there's no such thing, it's all a matter of opinion. It's about not following some coherent policy/convention! In PHP, unless you have it all memorized, you won't ever know how a particular function you may not have used recently would be called. You have to look it up. This is in sharp contrast to, say, Qt, or many CLR/.Net libraries, where you can often correctly guess method and class names. The problem with PHP is that this lack of attention to detail pervades the API. What's the point of a high level programming language if a lot of its API is pretty much a 1:1 (but slightly broken) mapping onto the underlying C libraries?! I mean, come on, can't anybody really come up with anything better than the str* and printf* family?
I've done a few projects in PHP and it felt like writing something that would be input to cweave/ctangle - pretty much C with garbage collection that happens to be autoextracted from an html file. Good grief.
Not only is PHPs API byzantine, the functionality it provides is often incomplete or promotes unsafe or poorly performing practices. PHPs functionality is nothing special at all. You get that and much more with any half-decent Python installation, or any CLR server.
IDEs normally don't do a whole lot of that. It's been only with VS2010 and Qt Creator (and maybe KDevelop) that this is being done online. Maybe Eclipse does that too, but such full checking requires integrating essentially an entire compiler front-end into the IDE -- for example Visual Studio didn't have that until VS2010 where they integrated the EDG front end.
So, pray tell, what IDEs that do "automatic syntax correction, type verification, argument checking" did you have in mind?
C++ has a rather flat learning curve if you're doing it on your own. Plan on spending a decade learning it. If you want to accelerate it, you need to delve into hardcore books (no "teach yourself in x days" type stuff!) -- stuff by Scott Myers, Alexandrescu, Strostrup, the GOF.
I vilify PHP simply for having seemingly absolutely no direction in its development. Just look at the function names, they are what they are with no rhyme nor reason. All of its API just reeks of being designed by a bunch of people in an insane asylum. I value myself enough to be able to afford working with technologies that don't leave a sour aftertaste day after day. I can afford to work with stuff that has a certain amount of beauty to it.
PHP's high adoption rate is partly because people just accept the byzantine without winking an eye. There's no expectation, it seems, at least not among the novice developers, towards stuff that works well and doesn't reek.
I concur. RAII is not a java idiom, and you don't do things that way in java. Depending on finalization in java is about the most misguided thing ever. If you're the parent is not trolling, then I'm very sorry for his customers...
Obviously, in your mind, people who pay for, say, RedHat EL, must not exist. Heck, even I pay for two RHEL self-support subscriptions, it's well worth the price in a business.
One would hope that a particular library/framework that is being used would be familiar to the people who maintain the code. If that's the case, then I don't see how operator overloading would be considered a drawback. Alas, IMHO C++ operator overloading is, to an extent, a crutch needed because the language is not very expressive. For example, in C++ the most concise way of setting up a constant type-safe matrix might be like this (this syntax is from eigen): Matrix m; m 1, 2, 3,
4, 5, 6,
7, 8, 9;
This required overloading both and , just to allow for minimum verbosity. Compare this to LISP, where the matrix constant constructor could be a macro, and you wouldn't need to overload anything at all -- the matrix constant would use competely standard syntax, and it could look somewhat like: (setq m (matrix:rows 3:cols 3
1 2 3
4 5 6
7 8 9)) Note that if matrix were a macro, it'd run some code at compile time to take the numbers in the argument list and produce a matrix constant.
By extension, think of the children. Sigh, this kind of thinking just rubs me all the wrong way. Are there no ways of disabling someone without killing them? You'd think it's a solved problem by now.
IIRC: it has to do with William the Conqueror (from Normandy) who invaded and took over England in mid 1000s. They brought French as their language, and set it up as the language of the royal court and of business. English was the language of illiterate peasants, back in those days if you knew how to write you would not speak English in dealing with people at or above your level in the society, neither would you write it. The problem with English was that it couldn't very well express the more complex abstract notions (it would be impossible to translate quite a few ancient works into it, for example): it just didn't work too well for much else besides a rather simple, uneducated life. This lasted for ~400 years. Those Normans were pretty much what shaped the development of written middle English. See this and that.
Why would one need such an epistemological model? Is it practical? I thought that the links to related entries in wikipedia articles are more than sufficient in practice.
I don't think it's a dialect, "of" sounds awfully close to "have" when uttered in "should have". It's one of many mistakes that are made because the french have fucked up English writing system so badly...
I have a bit of a problem with just killing anyone who tried to break into your house. I mean, come on, burglary is not a capital crime in most places. Why do you think it's OK in your home, but not in court?
When looking at a particular encyclopedia, one may be able to figure out the quality of the school curriculum just by seeing how useful that encyclopedia was in the schoolwork. My hypothesis is: the more useful it is, the worse the school curriculum is. I went to public schools that had, in hindsight, excellent curricula, and did set me up for life; Britannica (from early 90s) was pretty much a waste of shelf space -- almost useless as homework and test prep reference.
I really cannot think of any occasion where the two-paragraph overview from a printed encyclopedia ever helped me accomplish anything. If I needed to study something specific, I went to the library and borrowed a few books on the topic.
I couldn't agree more: the major problem with the idea of a classically edited encyclopedia is that it's mostly useless, at least for schoolwork of any sort, and I don't even mean college. Many wikipedia articles are genuinely useful, with worked out examples of calculations where applicable, instantly available links, etc. I've tried using Britannica as a reference in grades 8 through 12. Not a single time could I find answers to my questions, and it wasn't for lack of trying. It had a lot of facts but nothing specific enough to be useful in the homework problems and assignments that I had. I remember that I mostly tried to use it for mathematics, geography and biology. In mathematics it presented facts that were useless in isolation, even if true -- it could have served as a reference for a career mathematician perhaps, but nothing more. In biology, it provided vastly insufficient detail, it seemed -- I mean, come on, stuff edited by experts in the field (or so they claimed) and it was less detailed that we had learned in damn high school. In geography it provided stuff that was maybe useful for filling in crosswords or for a trivia contest, but, again, utterly useless for schoolwork. Whatever maps were presented were missing out on some pretty basic things (important river tributaries, for one).
Sorry, should I have meant that the paper you linked to is devoid of calculations. Mathis's site has a bunch of stuff where calculations are present, but still, he doesn't show any coherent theory. He just rehashes what's known, but it gives no new insights at all (contrary to what he's claiming). It was educational, but still a waste of time...
Using neutrinos will be even more expensive, even if you discount the size of the detector needed to receive information with any decent bandwidth. Alas, such discounting is a pipe dream: my guesstimate is that if you want to receive with bandwidth of a modern subsea cable (terabits/s), you'd need a detector whose volume is no less in order of magnitude than the volume of all the buildings in the world (yes, every one of them, even shacks in shantytowns).
It's a yet another site full of text but devoid of any calculations. I don't care how "nice" his theory is, I care that he shows that not only his theory works at predicting how nature works up to what we currently know, but can correctly predict some other things that current theories don't (and that, perhaps, we didn't even measure just yet). That's the big issue that somehow otherwise intelligent people don't get. It's absolutely irrelevant if their theories make "more sense": they have to show that they are indeed scientific theories: that they have predictive power. Mathis has not shown that he can calculate anything. His crap is just as useless as all the other crap out there. Yes. Shut up and calculate, indeed.
Slashdot Mirror
Borland Pascal documentation was done the same way!
I concur. Just ask the wine developers about how complete and accurate MS documentation is. They are reverse engineering a whole lot just because the documentation sucks in a way.
Walmart may have a market cap of 50% yearly revenue because they have razor-thin margins when you compare them with Apple. They have got much less of a cushion should things turn bad, compared to Apple.
Yep. See, for example, javascript memory leaks in Firefox. And those leaks are not any sort of an underlying C/C++ problem, just that it's very simple to hold on to a reference to a DOM that has a 100mb compartment on top of it. Plug-ins seem to do it left and right, and then Firefox slows to a crawl -- no wonder.
But then you'll be writing what is essentially object-oriented Pascal, not C++.
It's not about what's best -- there's no such thing, it's all a matter of opinion. It's about not following some coherent policy/convention! In PHP, unless you have it all memorized, you won't ever know how a particular function you may not have used recently would be called. You have to look it up. This is in sharp contrast to, say, Qt, or many CLR/.Net libraries, where you can often correctly guess method and class names. The problem with PHP is that this lack of attention to detail pervades the API. What's the point of a high level programming language if a lot of its API is pretty much a 1:1 (but slightly broken) mapping onto the underlying C libraries?! I mean, come on, can't anybody really come up with anything better than the str* and printf* family?
I've done a few projects in PHP and it felt like writing something that would be input to cweave/ctangle - pretty much C with garbage collection that happens to be autoextracted from an html file. Good grief.
Not only is PHPs API byzantine, the functionality it provides is often incomplete or promotes unsafe or poorly performing practices. PHPs functionality is nothing special at all. You get that and much more with any half-decent Python installation, or any CLR server.
IDEs normally don't do a whole lot of that. It's been only with VS2010 and Qt Creator (and maybe KDevelop) that this is being done online. Maybe Eclipse does that too, but such full checking requires integrating essentially an entire compiler front-end into the IDE -- for example Visual Studio didn't have that until VS2010 where they integrated the EDG front end.
So, pray tell, what IDEs that do "automatic syntax correction, type verification, argument checking" did you have in mind?
C++ has a rather flat learning curve if you're doing it on your own. Plan on spending a decade learning it. If you want to accelerate it, you need to delve into hardcore books (no "teach yourself in x days" type stuff!) -- stuff by Scott Myers, Alexandrescu, Strostrup, the GOF.
Huh? People who made the eric python IDE would like to disagree.
I vilify PHP simply for having seemingly absolutely no direction in its development. Just look at the function names, they are what they are with no rhyme nor reason. All of its API just reeks of being designed by a bunch of people in an insane asylum. I value myself enough to be able to afford working with technologies that don't leave a sour aftertaste day after day. I can afford to work with stuff that has a certain amount of beauty to it.
PHP's high adoption rate is partly because people just accept the byzantine without winking an eye. There's no expectation, it seems, at least not among the novice developers, towards stuff that works well and doesn't reek.
I concur. RAII is not a java idiom, and you don't do things that way in java. Depending on finalization in java is about the most misguided thing ever. If you're the parent is not trolling, then I'm very sorry for his customers...
Obviously, in your mind, people who pay for, say, RedHat EL, must not exist. Heck, even I pay for two RHEL self-support subscriptions, it's well worth the price in a business.
One would hope that a particular library/framework that is being used would be familiar to the people who maintain the code. If that's the case, then I don't see how operator overloading would be considered a drawback. Alas, IMHO C++ operator overloading is, to an extent, a crutch needed because the language is not very expressive. For example, in C++ the most concise way of setting up a constant type-safe matrix might be like this (this syntax is from eigen):
Matrix m;
m 1, 2, 3,
4, 5, 6,
7, 8, 9;
This required overloading both and , just to allow for minimum verbosity. Compare this to LISP, where the matrix constant constructor could be a macro, and you wouldn't need to overload anything at all -- the matrix constant would use competely standard syntax, and it could look somewhat like: :rows 3 :cols 3
(setq m (matrix
1 2 3
4 5 6
7 8 9))
Note that if matrix were a macro, it'd run some code at compile time to take the numbers in the argument list and produce a matrix constant.
By extension, think of the children. Sigh, this kind of thinking just rubs me all the wrong way. Are there no ways of disabling someone without killing them? You'd think it's a solved problem by now.
IIRC: it has to do with William the Conqueror (from Normandy) who invaded and took over England in mid 1000s. They brought French as their language, and set it up as the language of the royal court and of business. English was the language of illiterate peasants, back in those days if you knew how to write you would not speak English in dealing with people at or above your level in the society, neither would you write it. The problem with English was that it couldn't very well express the more complex abstract notions (it would be impossible to translate quite a few ancient works into it, for example): it just didn't work too well for much else besides a rather simple, uneducated life. This lasted for ~400 years. Those Normans were pretty much what shaped the development of written middle English. See this and that.
Why would one need such an epistemological model? Is it practical? I thought that the links to related entries in wikipedia articles are more than sufficient in practice.
I don't think it's a dialect, "of" sounds awfully close to "have" when uttered in "should have". It's one of many mistakes that are made because the french have fucked up English writing system so badly...
1 in 10 people going into a store are going to steal stuff. If it isn't locked down, it will disappear.
And I thought it was Poles who where the thiefs... Alas, got anything to back this statistic? I think it's pulled from some stinky place :(
I have a bit of a problem with just killing anyone who tried to break into your house. I mean, come on, burglary is not a capital crime in most places. Why do you think it's OK in your home, but not in court?
Their digital cameras were crap, and the consumer film ones were barely any better :(
When looking at a particular encyclopedia, one may be able to figure out the quality of the school curriculum just by seeing how useful that encyclopedia was in the schoolwork. My hypothesis is: the more useful it is, the worse the school curriculum is. I went to public schools that had, in hindsight, excellent curricula, and did set me up for life; Britannica (from early 90s) was pretty much a waste of shelf space -- almost useless as homework and test prep reference.
I really cannot think of any occasion where the two-paragraph overview from a printed encyclopedia ever helped me accomplish anything. If I needed to study something specific, I went to the library and borrowed a few books on the topic.
I couldn't agree more: the major problem with the idea of a classically edited encyclopedia is that it's mostly useless, at least for schoolwork of any sort, and I don't even mean college. Many wikipedia articles are genuinely useful, with worked out examples of calculations where applicable, instantly available links, etc. I've tried using Britannica as a reference in grades 8 through 12. Not a single time could I find answers to my questions, and it wasn't for lack of trying. It had a lot of facts but nothing specific enough to be useful in the homework problems and assignments that I had. I remember that I mostly tried to use it for mathematics, geography and biology. In mathematics it presented facts that were useless in isolation, even if true -- it could have served as a reference for a career mathematician perhaps, but nothing more. In biology, it provided vastly insufficient detail, it seemed -- I mean, come on, stuff edited by experts in the field (or so they claimed) and it was less detailed that we had learned in damn high school. In geography it provided stuff that was maybe useful for filling in crosswords or for a trivia contest, but, again, utterly useless for schoolwork. Whatever maps were presented were missing out on some pretty basic things (important river tributaries, for one).
Sorry, should I have meant that the paper you linked to is devoid of calculations. Mathis's site has a bunch of stuff where calculations are present, but still, he doesn't show any coherent theory. He just rehashes what's known, but it gives no new insights at all (contrary to what he's claiming). It was educational, but still a waste of time...
Using neutrinos will be even more expensive, even if you discount the size of the detector needed to receive information with any decent bandwidth. Alas, such discounting is a pipe dream: my guesstimate is that if you want to receive with bandwidth of a modern subsea cable (terabits/s), you'd need a detector whose volume is no less in order of magnitude than the volume of all the buildings in the world (yes, every one of them, even shacks in shantytowns).
It's a yet another site full of text but devoid of any calculations. I don't care how "nice" his theory is, I care that he shows that not only his theory works at predicting how nature works up to what we currently know, but can correctly predict some other things that current theories don't (and that, perhaps, we didn't even measure just yet). That's the big issue that somehow otherwise intelligent people don't get. It's absolutely irrelevant if their theories make "more sense": they have to show that they are indeed scientific theories: that they have predictive power. Mathis has not shown that he can calculate anything. His crap is just as useless as all the other crap out there. Yes. Shut up and calculate, indeed.