Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 Internet speed test! ×

Comment C++ should be the introductory language (Score 1) 630

I did my own research on this, and went through the top 10 computer science universities and looked at what they taught in their introductory CS classes. Python and Java made up 100% of them, with only one (Stanford) having a C++ option.

Personally, I think C++ should be the introductory language for computer science majors. (Non-CS majors? Sure, teach them Python or Javascript.) Why? Because CS majors all have to learn computer architecture and usually assembly programming is part of learning architecture. It's way, way easier for people to go from C++ to ASM than it is to go Python to ASM or Java to ASM. So a lot of assembly classes I've gone through have backed away from teaching ASM and instead teach C with a touch of ASM in it, which means that their education gets compromised by an attempt to make the introductory class easier.

But research in computer science education shows that you can learn basic computer science principles pretty much equally well regardless of language taught, so we're sacrificing educational quality for no real benefit.

I think most opposition to C++ came from people that learned it back in the day with square bracket arrays and char* strings, none of which really should be used any more now that we have vectors and strings. (And have had for a very long time, really.) Modern C++ is a very enjoyable language to code in.

Comment Re: They simply remember your UDID (Score 1) 115

>Who would have ever thought that a company founded on the principle [sic] of breaking the law in multiple jurisdictions would ignore and circumvent the terms and conditions, to which they agreed, of an entity with which they do business. Whodathunkait.

They're adding functionality that Apple refuses to do. If you cheat in a Steam game, your device and account gets banned. On iOS, apparently, you just uninstall and reinstall and then you can fraudlently order cars all over again.

Might violate the Apple TOS, but they're in the ethical right on this one.

Comment Re:Payment vs Service (Score 1) 903

>As the average net work in a bar jumps the moment Bill Gates walks into it. You wouldn't pretend that Bill Gates has the same standard of living of someone making $7.25 an hour, so why pretend there isn't an enormous gap between schools in wealthy districts and poor ones? There's a reason why no one talks about "failing public schools" in Westchester or the Hamptons.

Which is why teachers around here get paid more to work in bad schools. It doesn't help, though, the research shows. The best teachers still bail out of the schools because they want to work with better kids.

>You know perfectly well that teachers don't start and stop school when students do.

Sure. So do software engineers. How much time do software engineers spend coding on their own free time? More time than teachers spend prepping for class, especially after they've been teaching the class for a while.

>it would be more than balanced by working 50-70 hours a week when school is in session.

On the clock? Hah. No, teachers unions would eat such a proposed workweek alive. If, again, you're taking about other stuff, again, so do software engineers.

>Reasonable? The people claiming this wouldn't touch a teaching job for less than a six figure salary.

Ah, there's the ad hominem. Except you'd be wrong. I taught at a high school just last year, in fact. In addition to running a software consulting business.

>Earning a masters degree, having tens of thousands in student loans to pay off, being salaried and invariably working far beyond 40 hours a week...and that's before even getting to the students. How much would you want to get paid per hour, per kid for being a babysitter, disciplinarian, nurse and social worker.

A master's degree? Are we talking a community college instructor, now?

>And that's before even getting to the actual teaching part, where your performance reviews

What performance reviews? I suspect you're unfamiliar with how the education system actually works.

>Not for a penny under six figures.

You think a person with a bachelor's degree in any subject should make six digits out of college? That's hilarious. You're talking pharmacist-level salary, and pharmacists are a hell of a lot more educated (and attendant student debt) than people with a BS or BA.

>Until they can't find a job that pays off their student loans, at which point it's time to sneer at them for taking on risk they couldn't afford.

How could you even type this? Doctors will take on six digits of student debt because they know they'll be able to pay it off in 10 years and then be very comfortable thereafter. To get a BS around here, it'll cost you about $10k for the first two years in a community college, and about $20k to go to a CSU. $30k in debt can be retired by a teacher off their salary. If they somehow go to Harvard to become a K-12 teacher, then they sign up for one of dozens of debt-forgiveness programs and go work in the ghetto for a while and all their student debt gets bought off by the government.

>You do realize, right, that the reason why doctors salaries are so high is because only wealthy families can risk the six figure cost of a medical degree

No. Again, I don't think you comprehend how student loans work. If you're a poor kid, for one thing, you'll pay close to zero to actually go to college through your bachelor's, and then you'll take on student debt for medical school, which you can work off quickly. Anyone can get a medical degree regardless of financial status.

You're stuck in some sort of 1950s mindset of how education works. I suggest you educate yourself as to how college works these days.

>Uh huh. Found a reason yet for why countries that do far more "meddling" in health care or education than the United States cover all or most of their population for a fraction of the cost?

Are you confusing the tuition paid by students in these countries for the actual cost to educate them? Or the nominal tuition price at a US college with the average price paid? I suspect you are.

Comment Re:Payment vs Service (Score 1) 903

>Our public education system is woefully underfunded

You think so? We spend about $13,000 per student per year, nationwide. (http://www.governing.com/gov-data/education-data/state-education-spending-per-pupil-data.html)

Average teacher salary is $55,000. (http://www.nea.org/home/54597.htm), but this varies pretty wildly by state.

That's a pretty reasonable amount, IMO, considering you only work 9 months a year and get pretty significant benefits.

>higher education is very costly

Anyone *can* go to college, that's how the system is currently set up. Even if it makes no economic sense, the federal government will subsidize your education.

>It would be nice if everyone smart enough to be a doctor or an engineer could just decide to go to school.

You do realize, right, that the reason why doctors salaries are so high is because we impose artificially low quotas on how many people can go to medical school each year?

If you look at costs over time, the two areas that have been spiraling out of control, cost wise, are medicine and college. This is the direct result of government meddling in the field trying to be more fair and just, but really just fucking over the vast majority of Americans.

Comment Re:Mythbusters (Score 1) 422

>It costs money to administer it all. Can that money be better spent elsewhere?

Yes, and if it costs a dollar, it would be well worth it, but if it is a trillion dollars it is not well worth it. So the question is how much? At the Department of Education, they do something similar and the overhead is between a tenth and a hundredth of a percentage point. So yeah, it's well worth it to make a public database. Especially since a lot of people will dig into the datasets for free on their own time, like I do. And if you find something damning, hey, you get a free paper out of it, which is great.

Frankly, your attitude that scientists won't go after replication because they want to chase after new stuff all the time is basically exactly the problem that we have. It's not as sexy as chasing after new research, but since new research is based on old research, this crisis means we're building giant edificies on foundations of sand. We're talking hugely influential, highly cited papers being unable to be replicated. This is a very, very serious matter. We're essentially wasting our money when we get research that we don't know is accurate or not. So our national priorities should change to fix that.

If you think it is career killing to replicate other people's work, I will just say that if the NSF funds these efforts, you'll see replication centers springing up at universities all over the country as they chase that sweet sweet money. Being a professor isn't just about publishing papers, it's about publishing papers and getting grant money. (Teaching? What's that?) New tenure track professors live and die based on grant money, so if you build it, they will come... and fight tooth and nail for it.

Comment Re:Mythbusters (Score 2) 422

>Having all data for all things published wouldn't make a sea change in science.

It would reveal if the scientists were p-hacking to get a significant result. On the downside, there is no downside. The NSF has been pushing for open data for a while, but it's nice to see it encoded into law.

> It would affect less than you expect in practice. It's certainly a bit useful for other scientists, but most scientists are keen on doing their own science.

Which is what needs to change. The NSF needs to fund more replicability grants.

>Has the EPA or NSF ever passed a regulation based off a brand new paper for which there's been no further work in the scientific community?

I didn't say no further work. I said we need to get way from thinking of peer reviewed papers as being actual scientific findings, even if they have p-values included. We used to consider a peer reviewed published paper as being something meaningful, but if it's worse than a coin flip, we need to set up a whole new grant infrastructure for repeating tentative findings that look interesting.

Comment Re:Mythbusters (Score 4, Interesting) 422

>develop plans

Says it all. This bill will actually implement it. It's a good thing. And it can't be overriden by Trump, which is also good.

The Replicability Crisis (https://en.wikipedia.org/wiki/Replication_crisis) is the most serious issue in modern day science. When 70% of published papers can't be reproduced, that means that you are making the correct bet to believe that any published, peer-reviewed, landmark study in a prestigious journal is wrong. And that's a very sad thing that I just had to write. But this is what happens when people set up a system that works the way ours does - people game the system and the trust and credibility of science is hurt by it.

So yeah. We really, really need to be pushing hard for public datasets and replicability of results.

If I were the head of the NSF, I would treat any paper that has not been replicated yet as tentative, and only accept it as a scientific finding until it has been repeated at least once by a third party.

Comment Illegal (Score 2) 120

This should be illegal.

Seriously. If the EFF isn't the right group to go after Verizon, please let me know who is and I'll donate $100 to the cause.

The doctrine of first sale should apply to cell phones as much as it applies to everything else. Our oligopolic mobile overlords have gotten away with being shitty corporations for way, way, too long now.

The saddest line ever penned by man was Stallman was right again.

Comment Re:Failure is always an option (Score 1) 200

>60MPH in San Francisco is going to get you some pretty bad fines most of the time :).

A friend of mine Ubers in SF, and tries to do runs to and from SFO for maximum money. He doesn't live in SF either, but commutes a long way every weekend to work there because the money is so good.

>(1) You're assuming all miles and hours are 'billable', while in reality you would be driving empty towards a pickup and waiting for the next pickup.

There's a pickup fee which offsets this, and in reality you can usually chain together rides.

Also, there's an additional bill per minute if you are in traffic.

Comment Re:Failure is always an option (Score 1, Informative) 200

>There are plenty of people who haven't figured out how much money they're going to end up spending on vehicle maintenance as a result of all that extra driving.

The IRS mileage rate is supposed to be an average cost for operating a vehicle. It is 53.5 cents per mile. Uber pays about twice that per mile in San Francisco. So if you can go at 60 MPH you'll be making about 30 bucks an hour, which is not bad for unskilled labor.

Comment Re:Interesting story (Score 1) 553

> I doubt very much that I could come up with a function to balance a tree out of the blue with no prep or review, nor is there much real world need for most developers to do so.

He didn't have to balance the tree, he just had to check if the tree is balanced.

Pretty easy to do with DFS (which the DHS agent obviously knew):

int depth_check(Node *n) {
    if (!n) return 0;
    int left = depth_check(n->left);
    int right = depth_check(n->right);
    if (left != right) throw exception;
    return left;
}

You could probably simplify it a bit more and use unsigned ints for correctness, but this was off the top of my head.

The calling function would check for an exception being thrown, and return false, otherwise return true.

Comment Re: Poor on $100k? Sure (Score 2) 805

>You're getting a lot better living for the $150k, you're definitely not in the same boat. That's like the people who say, "Oh, my BMW payments are so high, they're forcing me to cut back on my quality of life."

You forget our wonderful progressive tax system. A person with $150k in income and $100k in expenses will also be paying $32,000 in federal income taxes a year, plus state taxes, plus medicare, medicaid, etc. Will effectively be poor.

A person with $200k in income and $150 in expenses will pay $46,000 in taxes plus everything else, and will be running in the negatives every year.

>And even in the Bay Area, you can buy a nice house for $150k a year.

So a $600,000 house? There's exactly four 3 bedroom houses for sale at the $600k price point in San Francisco right now (on Zillow). The average is closer to a million for a single family home. There's a couple elsewhere on the penninsula and Marin, but pretty much everything with these specs is going to be Oakland, Richmond, Hayward, or Concord. I'd rather live in San Diego, thank you very much. (And I have indeed lived in both cities.)

Comment Re:One hour of basketball dunking per day. (Score 1) 142

Our schools (generally speaking currently mandate 3-4 *years* of PE and 0 years of computer science.

Some students are terrible at PE. So what? We make them do it anyway. These might even be the same students that excel at computer science, if the stereotypes are true.

But this isn't even a mandated year of CS. It's a bloody single hour, lodged somewhere in between the 4th and 12th grades. If you think we can't spare a single hour for coding, I don't know what to tell you.

The biggest obstacle to CS education is the sheer fact that nobody is exposed to it at an early age, so they don't know if they like it or are good at it before going to college. This stands in contrast to basically every other major STEM field, where everyone has an opportunity to (or be mandated to) take a high school level class. But only about 1 in 10 high schools even offer CS these days, and the numbers are going down because they're usually not counted for college science requirements.

So, no, this bill really is a good thing. The Hour of Code is so simple even troglodyte teachers can run it for their kids.

Comment Re:Like everything else start with the basics (Score 2, Informative) 312

I like Java, C++, C#, and Python, and think they all work great as introductory languages. C++ gets shit on a bit because there's a lot of bad memories from the 80s and 90s when you had to do a lot of things by hand, but modern C++ is a joy to code in. In fact, if it was up to me I'd say that colleges should teach C++ as their intro language for three reasons:

1) It's as powerful and expressive as Java and Python (with some notable exceptions like split() which you need to invoke Boost for). Smart pointers (instead of raw pointers), vectors (instead of C style arrays) and range-based for loops (to never have out of bounds errors) allows for very fast and safe programming.

2) It is a lot easier to go from C++ to Java/Python than vice versa. Java programmers tend to have a vague grasp on how memory actually works.

3) C is only one step away from assembly. C++ is two steps away (due to name mangling). Java and Python are three or more steps away. Assembly programming, while rare enough these days, is still the gateway to really understanding computer architecture and writing code that works with your architecture instead of against it. Success in assembly should be the goal for a lower-division computer science program.

I also agree with you that most languages take their cues from C++/Java in that they either follow the conventions or deliberately break them. So learning C++ or Java is a really good choice for new programmers for that reason as well.

Slashdot Top Deals

Every young man should have a hobby: learning how to handle money is the best one. -- Jack Hurley

Working...