Where is x_base stored? What if I pass an array as a parameter? Do you pass x_base on the stack with it it? What it if I pass a pointer to an array? What if I serialize and deserialize it?
What if I recursively process an array by passing references to sub-arrays? Do you create x_base on the fly at every level of recusion for every pointer that gets passed?
That doesn't change the observation that zero based is the natural order of things for a computer implementation. Your x_base solution is just trying to mitigate the performance hit, (and is only partially successful). It doesn't counter the argument.
It is quite true though that the 0-based thing is entirely an artifact of C (and of course languages that cribbed its syntax).
Its not an "artifact of C" its an artifact of reality.
If you allocate 10 bytes at address X, then:
the first one is going to be at address X, the second one is address X+1, etc.
The x[y] syntax then reduces to address x + y
And so x[0] is the byte at x, x[1] is the byte at x+1, etc.
Now you could 1 base the indexes, where x[1] resolves to address of x, but the implementation of that is:
x + 1 - 1, and x[y] in general becomes x + y - 1
That's an extra subtraction instruction on every single array access.
For larger datatypes it becomes:
x + y*sizeof(type) for zero based and x + (y-1)*sizeof(type) for one based arrays
That's going to be true for any language, and the question is posed to every language designer: either the programmer counts from zero, or every array access in the language has a subtraction added to it.
For C, which is designed to be a bare bones close to the hardware language the decision was a nobrainer, but even for higher level languages its always a choice between a performance hit, or using zero based indexing.
Another wrinkle is that if you use 1 based indexing, then x[0] is undefined.
Another wrinkle is that if you use 1 based indexing then the maximum length of an array was reduced by 1.
No. Ada begins iterating wherever you tell it to. You can index your arrays from -100 to 0 if you like.
I don't deny that it's useful, but Ada does that by effectively adjusting all the indexes to zero based behind the scenes.
"Thinking that's a feature of programming is a sure sign of a inexperienced programmer."
No, only an inexperienced programmer would not understand that zero based indexing is the 'natural' indexing, and that anything else requires extra processing.
school could do a better job preparing you for work.
It could, but the question is should it.
Your complaint about C++ is like a CGA not knowing how to use Simply Accounting. You got a degree in computing science. Not C++ programming. Yes, C++ programming is useful, and yes, you might have wanted to take a course in that, but its not related to the degree.
Sure, it's great by their standards, but their standards isn't the one people care about at the end of the day.
True, but industry wants the other extreme, and nothing will satisfy them. Bottom line is if they want an employee who can do X. They can train that employee. That's how it works in other industries.
. I knew lots other really cool, advanced CS stuff that their company could use, but unfortunately that's not a substitute for C++ experience (coursework would have been acceptable).
In the upper level compsci courses i took the languages were treated as a tool, not a subject. We were given a 1 day primer and the manuals, but it was a course on "advanced computing topic" not "language". You could have taken it on yourself to learn C++; if you can get a compsci degree you can learn C++. Take a correspondance course, write and release a freeware program. Done.
I'm surprised it took you until the interviews to catch this out. This should have come up earlier -- surely the career counselors would have caught it, or the work experience programs and internship placement stuff, or even just talking with other students and graduates and profs.
My university had a couple C++ electives I could take, including one by correspondence. Usually worth 1 credit. Or you could take C++ separately from university. I took a C++ windows programming course that looked at MFC and COM and the windows event model from an academic point of view. You didn't need them to graduate but everyone was encouraged to take them as they were "relevant" in the job market. The profs were all well aware of C++ in the world, and just thought it was a terrible teaching language for most subjects.
That's no surprise, my post wasn't a hypothetical, but drew on those real scheduler controversies.
The current CFS scheduler was inspired by Kolivas' work, and largely addressed desktop interactivity.
The CFS is quite a bit more complex, and slower than the previous scheduler which was generally deemed fine (and faster, therefore better) for non-interactive "servers", hence the controversy, and made it a natural example for the argument.
Pluggable or tunable schedulers would have been even more complicated than CFS. But even CFS represents an increase in kernel complexity primarily to satisfy one type of user.
That's not to say there's no value - being able to do calculus and such by hand is necessary because you need to understand what's going on, but no one is going to pay me to just sit there and do calculus problems by hand all day.
Work pays you because you understand calculus. You pay school to develop that understanding. School and Work are not the same.
Also, you usually don't get "redos" on an assignment from your boss
Because my boss expects me to already understand what I'm doing. The entire reason you go to school is because you don't know, and you want to learn.
it's expected that you've done what you can to get it right the first time.
You are expected to make mistakes and learn from them.
And the professor can always hand out practice problems or in-class quizzes for the competency check
In class quizzes and practice problems? University is not grade school my friend, "homework" is your "in class quiz" and "practice problems"; its just not in class where its a complete waste of the limited time you have with the prof. I don't know about you, but I got 3 hours of class time a week. Depending on the course, I got another few hours of TA or supervised lab time. That class time was for the prof to explain the material, and answer questions. It would have been idiotic to use that time do practice problems on our own. Everything else was homework. The homework was your practice. Where you are expected to make mistakes and learn from them, discover the gaps in your knowledge and fill them. Some of that you would do on your own, simply by attempting to do it, other times you would make mistakes that you wouldn't catch and that's why someone ELSE is checking all your work. Sometimes alll you need is to know its wrong, so you can go back and sort it out, and sometimes you need someone to talk you through what you are missing. That's the education process.
but the major evaluation should mirror what is desired in the real world.
Why? In the real world your boss expects you to do Y, and assumes you already know the prerequisite A,B,C. School is where you learn your A,B,Cs.
but are the very things that industry constantly complains about with fresh graduates
University a isn't trade school.
I didn't get a degree in computer science so that I could learn how to deploy Active Directory, how to properly configure Apache for security, or program against a given library/API that's popular "in industry", or learn R. You graduate with a degree in comp sci and you should know how computers work, how compilers work, how networks work, how programming languages work, what prodecural, functional, and object oriented are, you'll know about recursion, you'll know about concurrency and resource locking, semaphores and critical sections, atomic transactions, you'll know about AI, or SQL, you'll know how algorithms work, how stacks works, how fundamental data structures work, how to compute performance characteristics, etc. I use a lot of this knowledge all the time, and even the stuff I don't use I'm aware of and can recognize its applicability when it comes up. So when I get asked whether an R program used on inputs of n=3 or 4 runs in 5 minutes to an hour how big can they go... and I have the tools to analyze it and say its O(21^n) and you can get up to be about 6 on a PC, maybe 9 or 10 on a cloud platform for some $$$. (real example by the way from a couple weeks ago). The algorithm was about as good as it could be, so we analyzed the problem itself and came up with a new way of defining the solution space, that could be evaluated and searched in O(n^3) and give us the results we needed. And new we can go to n=20 and beyond on a laptop. That's what I went to university for.
The stuff industry wants, best practices for coding Java interfaces, deep API knowledge in whatever API they happen use that week (net, java, qt, whatever), how to use Team Foundation Server or git or whatever they happen to use th
A lot of people die every year, and flu death are not random. The people most likely to die anyway are the typical fatalities. I'm not saying everything is hunky dory, but just cranking out the raw numbers distorts the story.
Sure, any disease culls the weak and sick first. But a typical bad flu season is less than 50,000. So if we subtract that number out of the 1.8 million... well... its still 1.8 million since that figure wasn't precise enough to meaningfully subtract ~50,000 from it.
3% mortality is horrific; if 20% (typical "bad" flu infection rate) of the USA population got the disease, and 97% survived... that would still be 1.8million corpses in the USA alone.
That's more dead American's in one go than all American casualties of war since (and including) the revolution.
World War I & II, the Civil War, the Revolution, Vietnam, Korea, Afghanistan, Iraq, and all the little ones combined is only 1.3M dead, (with another 1.5M wounded).
1.8M dead is pretty horrific.
Yes, its not civilization crushing levels of horrific like a new black plague would be, but it's still pretty horrifying.
I understand the difficulty, but in the case of advanced math courses, many of them don't have a lot of students in them
Guess it depends on your university. There were easily 200 students in my differential equations stuff.
But yes, some courses were markedly smaller, and you are right that variability in the class makeup does become more relevant in that case.
which deflates the argument those who would receive an F in a curve probably would have received and F anyway
To a point yes. But in those classes, the grading on the curve wasn't some strict... "there shall be 4A, 16B, 20C, and 5F", but more adaptive to the circumstances.
As you said, these classes are getting smaller, and the prof and TAs tend to have enough direct interaction with the students that
"It's better to have projects and homework be the primary evaluator since this is more in line with handling "real world" scenarios - which in theory is the point of giving out grades to begin with."
I disagree.
Thanks to wolfram alpha, stack overflow, and the internet in general, homework is as meaningful or as meaningless as the student chooses to make it. Most professors assigned 5-10% of the grade to homework, and that was usually just for completion, not for correctness.
Homework is an opportunity to practice and develop and understanding of the material being taught. But its up to the student to approach it like that, and use it as the tool for learning it is supposed to be.
Marking the homework is required, and reviewing the mistakes is where the learning kicks in. But actually grading the homework for ones final mark is almost absurd -- its supposed to have mistakes in it -- that's how we find the gaps in our understanding in the first place. Penalizing someone on their final grade for failing to know the material they are still trying to learn is demented.
"Projects" are also rife with problems; with much of the grade reducing to grading the student skills at project management itself. That's an important skill too, but not the one that should be getting tested. (Although to be fair, taking tests is also a skill unto itself, but "test taking" is easier to learn to do well than "project management")
And group projects have all kinds of other issues. In my experience the only people who like group projects are people who treat it like free marks some of their classmates will do for them.
Only if the camera crew and the cast have the same accommodations. Otherwise, by that same logic, the living conditions of the wild animals featured in National Geographic must be "on par with wilderness camping in your mom's basement".
Only if the wild animals can turn to the cameraman and say "fuck this, my legs is infected, fly me home and to a hospital right now, this is over unless you want a lawsuit the likes of which will bankrupt this entire network."
If you have all good professors (a worthwhile goal to be sure) you still need grade curves because the testing is highly variable, especially in fields like advanced math.
It is virtually impossible to come up with a 3 hour exams year after year that are equally difficult, but also substantially different.
It's more practical to make a series of hard exams (knowing that the precise hardness varies from year to year), and then grade on the curve so it doesn't matter if one years exam is 15% or even 50% harder than the previous years.
c) One of the groups had better teachers, so they learned more.
Another poster suggested the same. But that's going to happen whether you grade on the curve or not. A bad teacher teaching to his own easy tests is going to churn out students who don't know as much with the same grades. Using standardized tests doesn't help either, since a bad teacher can teach to those leaving students who score well on the test, but don't know how to do anything but score well on the test.
You identified a real problem, but it has nothing to do with grading on the curve.
I just looked up what exact model my MBP was. It was a MC226LL/A, running a Intel Core 2 Duo T9600 at 2.8 GHz.
Thanks for the detail and the link.
I've sorted out the confusion. The weird thing seems to be whether its geekbench 2 or 3. The T9600 mbp benchmarks over 3900+ on geekbench 2 vs the iphone at 2200; but on geekbench 3 the scores are much closer.
I wonder what that means.
That's why I think it's inevitable that we'll get an ARM-based MacBook at some point.
Bell curves "work" in academic settings because there's hardly any accountability imposed upon tenured professors for how they evaluate students
Bell curves in an academic settings work because they are are fitting the test to the students not the other way around.
The premise is that the student distribution is more likely to be stable than then difficulty course over time.
Suppose two Calculus 101 classes of 200 students each go through a university with two different profs, teaching assistants, and tests. And the absolute grade distribution of the two classes is substantially different.
Is it more likely that:
a) All the smart kids were in one class
b) One class was harder than the other (harder assignments, harder tests, more trick questions, prof less skilled at communication... whatever variables you like.)
The grading on the curve comes from the belief that the differences in the scores of two classes of the same course are in fact differences in the difficulty of that particular iteration of the class, and distributing students on the curve solves lots of otherwise difficult problems, such as:
students shopping profs for grades -- Prof A's tests are easier, so I'll get a better mark in his class. designing tests is hard, designing new tests that are exactly as difficult as old tests is 10x harder. Grading on the curve lets the prof design a test, and then if it turns out it was brutal, the top students still get As, and even some of the F's get bumped up to Cs etc. Or if its unexpectedly easy, the prof doesn't churn out a bunch of inflated GPAs as a result.
I recall getting A's on high level math mid terms in computability and complexity theory where I failed to even attempt a full answer on half the questions.
I understand the frustration some people have with being graded on the curve, and the fear that "even if everyone in the class aces the test" someone still gets an "F" but I ~never~ saw that in practice. My marks generally went up, since the profs were free to be brutal.
those local currencies are money because they are the circulating medium of exchange defined by the government
"defined by the government" is not a necessary criterion for "money" anywhere.
Bitcoin satisifies the only real requirement: that it be a "circulating medium of exchange".
I agree its acceptance is not regionally localized, like a small national currency would be, but given the trivial ability of bit-coin users to transact over the internet, localization is not really a legitimate criticism. If you have bitcoins and an internet connection, then you can exchange or purchase things with bitcoins from a marketplace much larger than the sao tome and principe marketplace.
Indeed, unless you happen to be standing on the tiny Islands of Sao Tome or Principe, bitcoin is much much easier to transact with.
But my MacBook Pro 17 was a top of the line machine 4 years ago
Erm... again I'm seeing lots of scores of 4000+ for a "2009 MBP 17" on the geekbench search; vs ~2200 for a 5S, still very nearly double.
And early 2009 is a good 4.5 years before the 5S releases.
I share your enthusiasm, but the numbers aren't quite there.
It's still pretty impressive that the 5s can match it
Not quite match it, but the iphone 6 very well might.
Although, performance on the desktop cpu has stalled.
The Ivy Bridge stuff vs Haswell is practically the same in terms of benchmark performance. And even Sandy Bridge isn't far behind. There hasn't been a big jump in performance since the Nehalem stuff in 2008. (Think I7 - 940)
So with the desktop moving very slowly, and all the r&d everywhere, including intel, focused on power efficiency and mobile, its no surprise that smartphones are making a lot of progress relative to their larger cousins.
But all that said I agree its still pretty impressive.
Feature-rich and stripped down are opposites. You want features to be available to you on a whim, learn to install new software.
You are missing the point. A stripped down system can still have a feature rich kernel. If I want a feature rich desktop then I need a kernel that has features that enable the sort of high performance UX I need.
Right down to a scheduler that's friendly to interactive user processes. But maybe that scheduler's not as optimal for what you were doing with your server, so now we want a tunable scheduler that can be adjusted towards either.
No. bitcoins are accepted almost nowhere, they thus don't have the liquidity of money
The Sao Tome and Principe Dobra is accepted nowhere but Sao Tome and Principe. With a population of less than 200,000 people, it has far fewer users than bitcoin.
Its still money.
"value" of bitcoins fluctuates *wildly*.
That doesn't disqualify it as money.
those in less regulated countries use the U.S. dollar for anything important.
That doesn't disqualify the local currency from being money.
Because that WAS the intention of the owner: to share their data with random, unknown 3rd parties. That's pretty much the entire purpose of P2P networks.
No. The intention of the owner, and the purpose of P2P was to share the files, not information about themselves.
You can argue that it is a natural function of the software, doing what it was designed to do, but that didn't get Weev anywhere either, now did it?
assuming you either have a buyer or a place willing to take them
If you have that, then you have money. That's the definition of "money".
bitcoins "value" fluctuates.
Another normal feature of money.
bitcoin exchanges aren't banks
Depends on your definition of 'bank'.
For some it qualifies, for others it doesn't. It's certainly not anywhere near as well regulated or reliable as a typical first world bank... but it may not be far off from local banks 200 years ago, or banks in other less well regulated countries even today.
those that use them deserve what they get
Not sure anyone deserves to be robbed. But yes, they were taking a risk, and should have held assets there in proportion to the risk they were willing to take that this and various other things could happen to their assets.
Slashdot Mirror
create a second value x_base=x-1
Where is x_base stored? What if I pass an array as a parameter? Do you pass x_base on the stack with it it? What it if I pass a pointer to an array? What if I serialize and deserialize it?
What if I recursively process an array by passing references to sub-arrays? Do you create x_base on the fly at every level of recusion for every pointer that gets passed?
That doesn't change the observation that zero based is the natural order of things for a computer implementation. Your x_base solution is just trying to mitigate the performance hit, (and is only partially successful). It doesn't counter the argument.
Oops.
Indeed.
It is quite true though that the 0-based thing is entirely an artifact of C (and of course languages that cribbed its syntax).
Its not an "artifact of C" its an artifact of reality.
If you allocate 10 bytes at address X, then:
the first one is going to be at address X, the second one is address X+1, etc.
The x[y] syntax then reduces to address x + y
And so x[0] is the byte at x, x[1] is the byte at x+1, etc.
Now you could 1 base the indexes, where x[1] resolves to address of x, but the implementation of that is:
x + 1 - 1, and x[y] in general becomes x + y - 1
That's an extra subtraction instruction on every single array access.
For larger datatypes it becomes:
x + y*sizeof(type) for zero based and
x + (y-1)*sizeof(type) for one based arrays
That's going to be true for any language, and the question is posed to every language designer: either the programmer counts from zero, or every array access in the language has a subtraction added to it.
For C, which is designed to be a bare bones close to the hardware language the decision was a nobrainer, but even for higher level languages its always a choice between a performance hit, or using zero based indexing.
Another wrinkle is that if you use 1 based indexing, then x[0] is undefined.
Another wrinkle is that if you use 1 based indexing then the maximum length of an array was reduced by 1.
No. Ada begins iterating wherever you tell it to. You can index your arrays from -100 to 0 if you like.
I don't deny that it's useful, but Ada does that by effectively adjusting all the indexes to zero based behind the scenes.
"Thinking that's a feature of programming is a sure sign of a inexperienced programmer."
No, only an inexperienced programmer would not understand that zero based indexing is the 'natural' indexing, and that anything else requires extra processing.
school could do a better job preparing you for work.
It could, but the question is should it.
Your complaint about C++ is like a CGA not knowing how to use Simply Accounting. You got a degree in computing science. Not C++ programming. Yes, C++ programming is useful, and yes, you might have wanted to take a course in that, but its not related to the degree.
Sure, it's great by their standards, but their standards isn't the one people care about at the end of the day.
True, but industry wants the other extreme, and nothing will satisfy them. Bottom line is if they want an employee who can do X. They can train that employee. That's how it works in other industries.
. I knew lots other really cool, advanced CS stuff that their company could use, but unfortunately that's not a substitute for C++ experience (coursework would have been acceptable).
In the upper level compsci courses i took the languages were treated as a tool, not a subject. We were given a 1 day primer and the manuals, but it was a course on "advanced computing topic" not "language". You could have taken it on yourself to learn C++; if you can get a compsci degree you can learn C++. Take a correspondance course, write and release a freeware program. Done.
I'm surprised it took you until the interviews to catch this out. This should have come up earlier -- surely the career counselors would have caught it, or the work experience programs and internship placement stuff, or even just talking with other students and graduates and profs.
My university had a couple C++ electives I could take, including one by correspondence. Usually worth 1 credit. Or you could take C++ separately from university. I took a C++ windows programming course that looked at MFC and COM and the windows event model from an academic point of view. You didn't need them to graduate but everyone was encouraged to take them as they were "relevant" in the job market. The profs were all well aware of C++ in the world, and just thought it was a terrible teaching language for most subjects.
That's no surprise, my post wasn't a hypothetical, but drew on those real scheduler controversies.
The current CFS scheduler was inspired by Kolivas' work, and largely addressed desktop interactivity.
The CFS is quite a bit more complex, and slower than the previous scheduler which was generally deemed fine (and faster, therefore better) for non-interactive "servers", hence the controversy, and made it a natural example for the argument.
Pluggable or tunable schedulers would have been even more complicated than CFS. But even CFS represents an increase in kernel complexity primarily to satisfy one type of user.
That's not to say there's no value - being able to do calculus and such by hand is necessary because you need to understand what's going on, but no one is going to pay me to just sit there and do calculus problems by hand all day.
Work pays you because you understand calculus. You pay school to develop that understanding. School and Work are not the same.
Also, you usually don't get "redos" on an assignment from your boss
Because my boss expects me to already understand what I'm doing. The entire reason you go to school is because you don't know, and you want to learn.
it's expected that you've done what you can to get it right the first time.
You are expected to make mistakes and learn from them.
And the professor can always hand out practice problems or in-class quizzes for the competency check
In class quizzes and practice problems? University is not grade school my friend, "homework" is your "in class quiz" and "practice problems"; its just not in class where its a complete waste of the limited time you have with the prof. I don't know about you, but I got 3 hours of class time a week. Depending on the course, I got another few hours of TA or supervised lab time. That class time was for the prof to explain the material, and answer questions. It would have been idiotic to use that time do practice problems on our own. Everything else was homework. The homework was your practice. Where you are expected to make mistakes and learn from them, discover the gaps in your knowledge and fill them. Some of that you would do on your own, simply by attempting to do it, other times you would make mistakes that you wouldn't catch and that's why someone ELSE is checking all your work. Sometimes alll you need is to know its wrong, so you can go back and sort it out, and sometimes you need someone to talk you through what you are missing. That's the education process.
but the major evaluation should mirror what is desired in the real world.
Why? In the real world your boss expects you to do Y, and assumes you already know the prerequisite A,B,C. School is where you learn your A,B,Cs.
but are the very things that industry constantly complains about with fresh graduates
University a isn't trade school.
I didn't get a degree in computer science so that I could learn how to deploy Active Directory, how to properly configure Apache for security, or program against a given library/API that's popular "in industry", or learn R. You graduate with a degree in comp sci and you should know how computers work, how compilers work, how networks work, how programming languages work, what prodecural, functional, and object oriented are, you'll know about recursion, you'll know about concurrency and resource locking, semaphores and critical sections, atomic transactions, you'll know about AI, or SQL, you'll know how algorithms work, how stacks works, how fundamental data structures work, how to compute performance characteristics, etc. I use a lot of this knowledge all the time, and even the stuff I don't use I'm aware of and can recognize its applicability when it comes up. So when I get asked whether an R program used on inputs of n=3 or 4 runs in 5 minutes to an hour how big can they go... and I have the tools to analyze it and say its O(21^n) and you can get up to be about 6 on a PC, maybe 9 or 10 on a cloud platform for some $$$. (real example by the way from a couple weeks ago). The algorithm was about as good as it could be, so we analyzed the problem itself and came up with a new way of defining the solution space, that could be evaluated and searched in O(n^3) and give us the results we needed. And new we can go to n=20 and beyond on a laptop. That's what I went to university for.
The stuff industry wants, best practices for coding Java interfaces, deep API knowledge in whatever API they happen use that week (net, java, qt, whatever), how to use Team Foundation Server or git or whatever they happen to use th
whoosh...
http://www.youtube.com/watch?v=xuxO6CZptck
A lot of people die every year, and flu death are not random. The people most likely to die anyway are the typical fatalities. I'm not saying everything is hunky dory, but just cranking out the raw numbers distorts the story.
Sure, any disease culls the weak and sick first. But a typical bad flu season is less than 50,000. So if we subtract that number out of the 1.8 million... well... its still 1.8 million since that figure wasn't precise enough to meaningfully subtract ~50,000 from it.
3% mortality is horrific; if 20% (typical "bad" flu infection rate) of the USA population got the disease, and 97% survived... that would still be 1.8million corpses in the USA alone.
That's more dead American's in one go than all American casualties of war since (and including) the revolution.
World War I & II, the Civil War, the Revolution, Vietnam, Korea, Afghanistan, Iraq, and all the little ones combined is only 1.3M dead, (with another 1.5M wounded).
1.8M dead is pretty horrific.
Yes, its not civilization crushing levels of horrific like a new black plague would be, but it's still pretty horrifying.
I understand the difficulty, but in the case of advanced math courses, many of them don't have a lot of students in them
Guess it depends on your university. There were easily 200 students in my differential equations stuff.
But yes, some courses were markedly smaller, and you are right that variability in the class makeup does become more relevant in that case.
which deflates the argument those who would receive an F in a curve probably would have received and F anyway
To a point yes. But in those classes, the grading on the curve wasn't some strict... "there shall be 4A, 16B, 20C, and 5F", but more adaptive to the circumstances.
As you said, these classes are getting smaller, and the prof and TAs tend to have enough direct interaction with the students that
"It's better to have projects and homework be the primary evaluator since this is more in line with handling "real world" scenarios - which in theory is the point of giving out grades to begin with."
I disagree.
Thanks to wolfram alpha, stack overflow, and the internet in general, homework is as meaningful or as meaningless as the student chooses to make it. Most professors assigned 5-10% of the grade to homework, and that was usually just for completion, not for correctness.
Homework is an opportunity to practice and develop and understanding of the material being taught. But its up to the student to approach it like that, and use it as the tool for learning it is supposed to be.
Marking the homework is required, and reviewing the mistakes is where the learning kicks in. But actually grading the homework for ones final mark is almost absurd -- its supposed to have mistakes in it -- that's how we find the gaps in our understanding in the first place. Penalizing someone on their final grade for failing to know the material they are still trying to learn is demented.
"Projects" are also rife with problems; with much of the grade reducing to grading the student skills at project management itself. That's an important skill too, but not the one that should be getting tested. (Although to be fair, taking tests is also a skill unto itself, but "test taking" is easier to learn to do well than "project management")
And group projects have all kinds of other issues. In my experience the only people who like group projects are people who treat it like free marks some of their classmates will do for them.
Only if the camera crew and the cast have the same accommodations. Otherwise, by that same logic, the living conditions of the wild animals featured in National Geographic must be "on par with wilderness camping in your mom's basement".
Only if the wild animals can turn to the cameraman and say "fuck this, my legs is infected, fly me home and to a hospital right now, this is over unless you want a lawsuit the likes of which will bankrupt this entire network."
The overwhelmingly vast majority of calories burned by the human body are burned in just living.
Buf it you burn 2000 calories just by living, and eat 2500, then 500 calories worth of exercise makes a huge difference.
And will have a much better effect on your overall health than just cutting your diet down by 500 calories.
Weight and health are related, but simply controlling your weight by reducing calorie intake won't make you healthy.
If you have all good professors (a worthwhile goal to be sure) you still need grade curves because the testing is highly variable, especially in fields like advanced math.
It is virtually impossible to come up with a 3 hour exams year after year that are equally difficult, but also substantially different.
It's more practical to make a series of hard exams (knowing that the precise hardness varies from year to year), and then grade on the curve so it doesn't matter if one years exam is 15% or even 50% harder than the previous years.
c) One of the groups had better teachers, so they learned more.
Another poster suggested the same. But that's going to happen whether you grade on the curve or not. A bad teacher teaching to his own easy tests is going to churn out students who don't know as much with the same grades. Using standardized tests doesn't help either, since a bad teacher can teach to those leaving students who score well on the test, but don't know how to do anything but score well on the test.
You identified a real problem, but it has nothing to do with grading on the curve.
I just looked up what exact model my MBP was. It was a MC226LL/A, running a Intel Core 2 Duo T9600 at 2.8 GHz.
Thanks for the detail and the link.
I've sorted out the confusion. The weird thing seems to be whether its geekbench 2 or 3. The T9600 mbp benchmarks over 3900+ on geekbench 2 vs the iphone at 2200; but on geekbench 3 the scores are much closer.
I wonder what that means.
That's why I think it's inevitable that we'll get an ARM-based MacBook at some point.
Yay. OSX RT, just what everybody wants.
Sadly its probably true.
Which means you get two sets of students with the same grades but widely different abilities.
That's going to happen whether you grade on the curve or not.
Even if you use standardized tests, the poorly run course might just teach to the test giving the same grades but students with weaker abilities.
You identified a genuine problem, but it doesn't have much to do with grading on the curve.
You can win a million dollars for surviving on a place where people already live!
Nevermind the camera crew that follows them around everywhere.
I think that's on par with wilderness camping in your mom's basement.
Or more likely he was just dumping them out the other end because, for whatever reason, he couldn't absorb them.
Bell curves "work" in academic settings because there's hardly any accountability imposed upon tenured professors for how they evaluate students
Bell curves in an academic settings work because they are are fitting the test to the students not the other way around.
The premise is that the student distribution is more likely to be stable than then difficulty course over time.
Suppose two Calculus 101 classes of 200 students each go through a university with two different profs, teaching assistants, and tests. And the absolute grade distribution of the two classes is substantially different.
Is it more likely that:
a) All the smart kids were in one class
b) One class was harder than the other (harder assignments, harder tests, more trick questions, prof less skilled at communication... whatever variables you like.)
The grading on the curve comes from the belief that the differences in the scores of two classes of the same course are in fact differences in the difficulty of that particular iteration of the class, and distributing students on the curve solves lots of otherwise difficult problems, such as:
students shopping profs for grades -- Prof A's tests are easier, so I'll get a better mark in his class.
designing tests is hard, designing new tests that are exactly as difficult as old tests is 10x harder. Grading on the curve lets the prof design a test, and then if it turns out it was brutal, the top students still get As, and even some of the F's get bumped up to Cs etc. Or if its unexpectedly easy, the prof doesn't churn out a bunch of inflated GPAs as a result.
I recall getting A's on high level math mid terms in computability and complexity theory where I failed to even attempt a full answer on half the questions.
I understand the frustration some people have with being graded on the curve, and the fear that "even if everyone in the class aces the test" someone still gets an "F" but I ~never~ saw that in practice. My marks generally went up, since the profs were free to be brutal.
those local currencies are money because they are the circulating medium of exchange defined by the government
"defined by the government" is not a necessary criterion for "money" anywhere.
Bitcoin satisifies the only real requirement: that it be a "circulating medium of exchange".
I agree its acceptance is not regionally localized, like a small national currency would be, but given the trivial ability of bit-coin users to transact over the internet, localization is not really a legitimate criticism. If you have bitcoins and an internet connection, then you can exchange or purchase things with bitcoins from a marketplace much larger than the sao tome and principe marketplace.
Indeed, unless you happen to be standing on the tiny Islands of Sao Tome or Principe, bitcoin is much much easier to transact with.
But my MacBook Pro 17 was a top of the line machine 4 years ago
Erm... again I'm seeing lots of scores of 4000+ for a "2009 MBP 17" on the geekbench search; vs ~2200 for a 5S, still very nearly double.
And early 2009 is a good 4.5 years before the 5S releases.
I share your enthusiasm, but the numbers aren't quite there.
It's still pretty impressive that the 5s can match it
Not quite match it, but the iphone 6 very well might.
Although, performance on the desktop cpu has stalled.
The Ivy Bridge stuff vs Haswell is practically the same in terms of benchmark performance. And even Sandy Bridge isn't far behind. There hasn't been a big jump in performance since the Nehalem stuff in 2008. (Think I7 - 940)
So with the desktop moving very slowly, and all the r&d everywhere, including intel, focused on power efficiency and mobile, its no surprise that smartphones are making a lot of progress relative to their larger cousins.
But all that said I agree its still pretty impressive.
Feature-rich and stripped down are opposites. You want features to be available to you on a whim, learn to install new software.
You are missing the point. A stripped down system can still have a feature rich kernel. If I want a feature rich desktop then I need a kernel that has features that enable the sort of high performance UX I need.
Right down to a scheduler that's friendly to interactive user processes. But maybe that scheduler's not as optimal for what you were doing with your server, so now we want a tunable scheduler that can be adjusted towards either.
And the complexity begins its lift off.
Yes, the 5s is incredibly fast. With a Geekbench score of 2523, it's faster than a 2011 Mac Mini (which were not slow machines by any means).
Which 2011 mac mini are you talking about?
http://browser.primatelabs.com/geekbench2/search?page=2&q=mac+mini+2011&utf8=%E2%9C%93
The scores for a 2011 macbook range from 5500 to 9500 depending on the model.
So... less than half as fast as a low end computer from 2 years ago? Ok... I'll buy that.
And the 5S is in the same ballpark as a Galaxy Note 3, Sony Xperia Z1, or LG Nexus 5... sure I'll buy that too.
Smartphones have the general performance of entry level PCs from 5+ years ago? Sounds plausible.
No. bitcoins are accepted almost nowhere, they thus don't have the liquidity of money
The Sao Tome and Principe Dobra is accepted nowhere but Sao Tome and Principe. With a population of less than 200,000 people, it has far fewer users than bitcoin.
Its still money.
"value" of bitcoins fluctuates *wildly*.
That doesn't disqualify it as money.
those in less regulated countries use the U.S. dollar for anything important.
That doesn't disqualify the local currency from being money.
Because that WAS the intention of the owner: to share their data with random, unknown 3rd parties. That's pretty much the entire purpose of P2P networks.
No. The intention of the owner, and the purpose of P2P was to share the files, not information about themselves.
You can argue that it is a natural function of the software, doing what it was designed to do, but that didn't get Weev anywhere either, now did it?
assuming you either have a buyer or a place willing to take them
If you have that, then you have money. That's the definition of "money".
bitcoins "value" fluctuates.
Another normal feature of money.
bitcoin exchanges aren't banks
Depends on your definition of 'bank'.
For some it qualifies, for others it doesn't. It's certainly not anywhere near as well regulated or reliable as a typical first world bank... but it may not be far off from local banks 200 years ago, or banks in other less well regulated countries even today.
those that use them deserve what they get
Not sure anyone deserves to be robbed. But yes, they were taking a risk, and should have held assets there in proportion to the risk they were willing to take that this and various other things could happen to their assets.