Follow Slashdot stories on Twitter


Forgot your password?

Comment Re:The elephant in the room (Score 1) 119

Western journalists are under no such oppressive restrictions and deliberately do such things of their own free will. Which system is worse?

Western journalists are poor sobs, with no valuable skills other than networking and rhetoric. If they want to have any sort of career, they need to suck up to politicians and/or create outrage, preferably both.

Comment Re:The elephant in the room (Score 1) 119

There is no left wing party in america.

Yes, there are no socialist or communist parties. There are also no fascist or Christian parties. Great, isn't it? I hope we can keep it that way, rather than turning into Europe.

Look at these numbers and tell us again how left america is?

And if you look at these numbers, you can see how good that is:

Comment Re:The elephant in the room (Score 2) 119

As somebody who grew up in a country with proportional representation, I agree: proportional representation is not particularly left wing, it is something that political extremists in general like, whether on the left or on the right. It turns a parliament into a collection of many small parties with extremist viewpoints and unstable, unpredictable coalitions.

Proportional representation also gives those political parties enormous power over their party members (since "proportionality" is determined at the party level), making independent candidacy next to impossible. If you add government regulation and public financing of campaigns into the mix, you have essentially created a political aristocracy.

In the US, proportional representation is mostly associated with the left because there is a pretty large political minority of progressives and democratic socialists who are frustrated that the majority is rejecting their ideas. They are hoping to gain power through proportional representation.

Proportional representation allowed the Nazis to come to power in the Weimar Republic and has brought many socialists and communists to power in Europe and South America. It is a bad idea, and one of the strengths of the US system is that it doesn't have it.

Comment Re:Knowing when not to (Score 1) 345

You're shifting your claim.

No one I know who does high performance code (such as numerics, real time computer vision, that sort of thing) uses anything but C++, especially for new projects. There is nothing out there that combines the speed and expressivity of C++, and when you know performance is going to be a factor at some point, C++ is the only choice.


Check out something like Eigen or TooN (a somewhat more obscure library which is used in the vision world for things like PTAM). They are very far from C++. The code written down in an editor reads more like maths. There's no explict loops, no explicit memory allocation. They're both high performance libraries used in challenging applications (seriously download and run PTAM, it's amazing).

PTAM is decent C++ code with probably pretty decent performance, but it isn't "high performance numerical code" or even high performance embedded code; it contains none of the performance tuning that such code usually should have.

I think it is dishonest on your part to recommend Eigen, when you know (or ought to know) that it actually isn't high performance and is deeply mired in C++ idiosyncracies. Likewise, I think it is dishonest of you to portray C++ as the only high performance computing language in town, when you know full well that many people are using other languages, and you are simply not even familiar with those other languages.

I do use C++ a lot in my work and I recommend it to other people. But unlike you, I don't feel any need to lie about what it is and isn't. There are many alternatives to C++, and I think people should keep an open mind about what they use.

Comment Re:Knowing when not to (Score 1) 345

So, you agree then that, numerical performance of Eigen sucks for large matrices. For small, fixed size matrices it likely is no better than Fortran either. Syntactically, Eigen is pretty lousy as well: you can write some array expressions, but defining functions on matrices is messy, and slicing and similar operations also are restrictive. Eigen's error messages are impenetrable due to the use of template metaprogramming, and parallelization support in it is close to non-existent.

Fortran these days has built in matrices that automatically use the best possible BLAS/LAPACK backends. You can write Matlab-like array expressions, get clear error messages, and can efficiently pass and return arrays, matrices, and slices. And it provides user-defined data types, operator overloading, virtual functions, and built-in parallel and distributed programming support.

I do a lot of C++ programming (including Eigen) and use C++ because I need to interface with a lot of C and C++ code. But, objectively, C++ is not a very good language for numerical or high performance computing. You can sort of get the job done in it, but it takes a lot longer to write code and ends up much more complex than in a decent numerical language. And while anybody who knows Matlab can write high performance Fortran code almost instantly, becoming a sufficiently good C++ programmers to even get started writing good numerical code takes years.

Large parts of the numerical and high performance computing community never touch C++ at all. C++ is mostly popular in computer vision, embedded systems, and some areas of machine learning.

Comment Re:Bogus (Score 1) 353

So you think an economic question like this cannot be analyzed by objective statistical data.

That's a question of mathematics: there are many questions that provably cannot be answered using objective statistical data.

As a proponent of H-1B's, how would you make the argument? Subjectively? Good luck.

Against Matloff? That's easy: Matloff wants to limit labor mobility in order to increase salaries in his profession and, taken at face value, his own data suggests that limiting H-1B visas would accomplish that. That alone is sufficient economic reason to increase, rather than decrease, H-1B visa quotas.

Matloff publishes in peer reviewed CS journals, and this debate is a sideline of his.

According to Google Scholar, Matloff hasn't published in peer reviewed CS journals for half a decade. What a joke.

Comment Re:Bogus (Score 1) 353

A straw man? It's a claim made by H-1B proponents. ... In other words, like the infamous phrase "best and brightest", Matloff is guilty of rebutting the arguments of the proponents using exactly the terminology that they use. Once more: complain to the proponents.

There are many arguments people make for H-1B visas, some relevant and some not relevant. Matloff choosees to rebut irrelevant arguments, hence he is picking a straw man. Even if his statistical analysis were sound, it would be irrelevant.

Then please give us some plausible scenarios, without which your argument is hand waving.

You seriously claim to be a statistician, and you can't figure this out yourself? Let's leave that discussion to peer review, shall we?

There is almost no mention (and none from you) of what a better way to do this analysis would be.

That's like saying "what's a better analysis of how many angels can dance on the head of a pin"; the premises of Matloff's analysis themselves are wrong, so there is no better way of doing the analysis (that is in addition to the fact that his analysis is also wrong from a statistical point of view).

"Wealth of the nation"? Such a brilliantly precise term from someone who called "shortage" ambiguous. What the hell does "wealth of the nation" mean?

It was an allusion to Adam Smith's classical work "The Wealth of Nations"; look there for what I mean by the term. Furthermore, I don't have to be precise, I'm not publishing a paper. I don't pretend that there is objective statistical data that can answer this question, Matloff does. He needs to make his case.

Comment Re:Bogus (Score 1) 353

While he's currently a CS prof, he used to be a statistics prof, and it shows. He uses hard data properly analyzed

As an immigrant and someone who does data analytics professionally, I have to say: his analysis is logically, statistically, economically, and legally unsound. (Note that his report hasn't even been peer reviewed.)

It is logically unsound because, among other things, it starts with the premise that for the H-1b program to be desirable or economically valid, its workers should be, on average, "better and brighter" than US workers (a straw man), and because it uses arbitrary and unvalidated measures of quality.

It is statistically unsound because he infers that foreign workers are "less bright" than American workers because the "foreign" attribute correlates negatively with measures such as salaries and number of patents; however, it is easy to construct scenarios for which such negative correlations exist even if the workers are objectively still "brighter" than US workers. Inferring that populations of foreign workers are "less bright" based on his statistical analysis is incorrect.

It is economically unsound because he keeps arguing in terms of a "labor shortage"; while such a fuzzy term is often being used to justify H-1b visas politically, it makes little sense as the basis of an economic argument either way. And if one wanted to argue in terms of a "labor shortage", a labor shortage for low-skilled tech workers would be as important economically as a shortage of "the best and the brightest".

Matloff also argues for trying to create an artificial scarcity of workers in his profession by restricting admission of foreign workers. This has a long history in economics. Its effect is to benefit members of that profession, while making everybody else economically worse off. But if H-1b workers do the same work as Americans at a lower salary and free the best and brightest Americans to work in higher paid professions, that is a good thing from an economic point of view and for the wealth of the nation.

Finally, legally, he postulates the existence of "green card indenture", but in practice both H-1b and green card workers frequently experience only brief periods where they can't change employers.

And if Matloff (and let's not beat around the bush, "ebno-10eb" is Matloff) claims having been a statistics professor as credentials in order to lend weight to his arguments, the inference from that is not that his argument is stronger, but that he must have been a pretty poor professor of statistics.

There is some common ground is that if we limit immigration at all, we should limit it to those most valuable to the economy. One way of doing that would be to simply auction off a fixed quota of employment-based H-1b and green card slots to the highest bidders each year. That way, we don't need to get into silly debates as to who is needed, or who is not getting paid enough, and people like Matloff don't get to abuse the immigration system to achieve higher salaries for their preferred profession.

Comment Re:Why? (Score 1) 557

I'm wondering what other images will become illegal because they elicit violence...

In Germany (bastion of free speech that it is), any speech that is "capable of disturbing the public peace" is forbidden. Insulting someone is also prohibited by law. You also can't insult the "organs and representatives of foreign states" (so no talking about Dick's you-know-what). You also may not insult any faith.

For other restrictions, see here:

Slashdot Top Deals

One small step for man, one giant stumble for mankind.