[ian807]

As dysfunctional as the USA's education system is, blaming it for the programmer shortage is disingenuous. It's about the money. As long as international wage arbitrage exists in other country's favor, the USA will have fewer and fewer programmers. Rational self interested people will pursue high paying work that *can't* be outsourced if given a choice.

Companies, of course, are international, and could care less where there code comes from as long as it's cheap, at least appears to work and makes some manager somewhere look good on a review.

[CTphpnwb]

Programmist, on 20 Jan, 2010 - 12:20 AM, said:

On paper I'd probably lean toward the grad, but in previous interviews I've seen candidates who look great on paper, but are dumber than a box of hammers.

Funny, I'd say the same thing about "experienced" coders. I've had to clean up after too many who write procedural code, and don't even do that well.

@ian807,

Exactly. And I would add that this attempt to blame the educational system is designed to deflect blame from where it belongs: short sighted business leadership.

[Paul-]

seeP+, on 18 Jan, 2010 - 10:02 AM, said:

I love to code and I know math is a plus, but Calculus?!?!

Calculus is one of the more intuitive branches of math, and can be easily related to real life applications. Perhaps you had a bad start with it.

[skyhawk133]

Something to keep in mind about this article. It's specifically about government/military/research. The government isn't going to outsource weapons programming and top-secret computing research to India. Just sayin'

[Paul-]

Programmist, on 19 Jan, 2010 - 08:20 PM, said:

CTphpnwb, on 18 Jan, 2010 - 10:26 AM, said:

I think you need to be qualified to program before you are accepted into a Masters program.

That would be nice, but as long as there are fields of research that require absolutely no programming it will probably never happen. I think this is why so many schools have created Software Engineering degrees.

This is a great point. It is easy to forget that computer science != programming. People who do theoretical work don't care about programming, and rightly so. If a prospective student wants to become a programmer, it's their own responsibility to choose a school and program of study accordingly. If what they study does not meet their career goals, perhaps they chose the wrong degree, rather than a poor school.

[Paul-]

skyhawk133, on 22 Jan, 2010 - 11:33 AM, said:

Something to keep in mind about this article. It's specifically about government/military/research. The government isn't going to outsource weapons programming and top-secret computing research to India. Just sayin'

Oh, you mean fewer people willing to work on the next killing machine? And that's bad? (Sorry, couldn't resist )

[diedev]

Paul-, on 22 Jan, 2010 - 11:46 AM, said:

Programmist, on 19 Jan, 2010 - 08:20 PM, said:

CTphpnwb, on 18 Jan, 2010 - 10:26 AM, said:

I think you need to be qualified to program before you are accepted into a Masters program.

That would be nice, but as long as there are fields of research that require absolutely no programming it will probably never happen. I think this is why so many schools have created Software Engineering degrees.

This is a great point. It is easy to forget that computer science != programming. People who do theoretical work don't care about programming, and rightly so. If a prospective student wants to become a programmer, it's their own responsibility to choose a school and program of study accordingly. If what they study does not meet their career goals, perhaps they chose the wrong degree, rather than a poor school.

tell me! I was in a point to take a CS career, thinking so wrongly that CS = full programing & full nights of code,so I was lose the money of the first payment of the career. I can't stand calculus anymore, I love a part of maths, the rest for the rest.
The mean word is "computer", anybody thinks computer is programming, so it's a mistake to fix among the crowd.

[KYA]

Paul-, on 22 Jan, 2010 - 12:46 PM, said:

It is easy to forget that computer science != programming. People who do theoretical work don't care about programming, and rightly so.

That's a poor assumption. CS graduates "should" be able to program. I haven't seen a single [and if you know of one please share] CS curriculum that did not involve programming of some sort. The point being a CS graduate should be well versed in both theory and application. Theory allows more efficient, creative, and optimal implementation. Will an everyday programmer need to know how to implement an AVL tree? What about an ability to look at an algorithm and guesstimate a rough Big O notation of it? Maybe not, but to say those skills are worthless is just dumb.

Same deal with math. With "typical" business applications, the math level will not go past Algebra II or Geometry. Does that mean you shouldn't take Calculus?

Going down a list of things a CS theorist might do:

Research - research what? Hardware, software, at some point programming/code comes into play
Design - chips? software? GPUs? Memory? Assembly knoweldge, hardware concepts, how soft/hardware interacts requires programming!
Teach - Unless there's a CS curriculum w/o programming, programming is there!


I cannot think of a job where one sits and ponders about CS and doesn't do a lick of programming. If you happen to be aware of one let me know! Sounds cushy.

[CTphpnwb]

Paul-, on 22 Jan, 2010 - 03:46 PM, said:

Programmist, on 19 Jan, 2010 - 08:20 PM, said:

CTphpnwb, on 18 Jan, 2010 - 10:26 AM, said:

I think you need to be qualified to program before you are accepted into a Masters program.

That would be nice, but as long as there are fields of research that require absolutely no programming it will probably never happen. I think this is why so many schools have created Software Engineering degrees.

This is a great point. It is easy to forget that computer science != programming. People who do theoretical work don't care about programming, and rightly so. If a prospective student wants to become a programmer, it's their own responsibility to choose a school and program of study accordingly. If what they study does not meet their career goals, perhaps they chose the wrong degree, rather than a poor school.

I second KYA's question about CS without any programming. I too would be amazed to hear of such a position.

That aside, my basic point is that the assumption that experience is more important than education is not merely flawed, but completely wrong. If experience were even nearly as important as education then we wouldn't have a need for colleges!

[Paul-]

KYA, on 22 Jan, 2010 - 02:51 PM, said:

Paul-, on 22 Jan, 2010 - 12:46 PM, said:

It is easy to forget that computer science != programming. People who do theoretical work don't care about programming, and rightly so.

That's a poor assumption. CS graduates "should" be able to program. I haven't seen a single [and if you know of one please share] CS curriculum that did not involve programming of some sort. The point being a CS graduate should be well versed in both theory and application.

Agreed. That is especially true for a BS degree, which should provide a broad coverage of all the bases. However, for a MS and beyond, people get specialized and will not use on a daily basis a large part of the skills they learned earlier. I know CS graduates with advanced degrees (and I mean successful ones) who during their research did not write a single line of code. Granted, those are extreme cases.

KYA, on 22 Jan, 2010 - 02:51 PM, said:

Theory allows more efficient, creative, and optimal implementation. Will an everyday programmer need to know how to implement an AVL tree? What about an ability to look at an algorithm and guesstimate a rough Big O notation of it? Maybe not, but to say those skills are worthless is just dumb.

I don't know that anybody talked about worthlessness. The discussion was around skills necessary for getting a certain type of job. A programmer job requires a level of proficiency in a certain language. A CS "theorist" with only basic programming coursework will be less suited for that type of job than a candidate who has specific training and/or experience in that language.

KYA, on 22 Jan, 2010 - 02:51 PM, said:

Going down a list of things a CS theorist might do:

Research - research what? Hardware, software, at some point programming/code comes into play
Design - chips? software? GPUs? Memory? Assembly knowledge, hardware concepts, how soft/hardware interacts requires programming!
Teach - Unless there's a CS curriculum w/o programming, programming is there!

No. Again, there is more to computer science than programming. I am not a theorist myself, so I can't argue the point as well as it could be done. But as an example we can think about the analysis of whether an algorithm can be solved with a Turing machine, in the theory of computability. In fact, it is people arguing about made-up problems run on imaginary computers.

KYA, on 22 Jan, 2010 - 02:51 PM, said:

I cannot think of a job where one sits and ponders about CS and doesn't do a lick of programming. If you happen to be aware of one let me know! Sounds cushy.

Think academia, ivory tower...

[Paul-]

CTphpnwb, on 22 Jan, 2010 - 03:41 PM, said:

I second KYA's question about CS without any programming. I too would be amazed to hear of such a position.

Well, with a little help from Google, one can find numerous examples. I spent a few minutes and this is what I came up with. For those who have heard of MIT, here is an article by Scott Aaronson, Assistant Professor of Electrical Engineering and Computer Science. It is about "Perfect Completeness for QMA", whatever that means, and does not contain any programming whatsoever.

CTphpnwb, on 22 Jan, 2010 - 03:41 PM, said:

That aside, my basic point is that the assumption that experience is more important than education is not merely flawed, but completely wrong. If experience were even nearly as important as education then we wouldn't have a need for colleges!

I disagree. Yes, if you have two candidates who have the exact same skills and experience, the one with a degree will always win over the one without one. In cases where you have to evaluate the experience of one against the diploma of the other, often times the former will win.

[CTphpnwb]

Paul-, on 22 Jan, 2010 - 09:07 PM, said:

CTphpnwb, on 22 Jan, 2010 - 03:41 PM, said:

I second KYA's question about CS without any programming. I too would be amazed to hear of such a position.

Well, with a little help from Google, one can find numerous examples. I spent a few minutes and this is what I came up with. For those who have heard of MIT, here is an article by Scott Aaronson, Assistant Professor of Electrical Engineering and Computer Science. It is about "Perfect Completeness for QMA", whatever that means, and does not contain any programming whatsoever.

CTphpnwb, on 22 Jan, 2010 - 03:41 PM, said:

That aside, my basic point is that the assumption that experience is more important than education is not merely flawed, but completely wrong. If experience were even nearly as important as education then we wouldn't have a need for colleges!

I disagree. Yes, if you have two candidates who have the exact same skills and experience, the one with a degree will always win over the one without one. In cases where you have to evaluate the experience of one against the diploma of the other, often times the former will win.

So we're to believe that since he's got Computer Science in his title that isn't a mathematics paper? I see set theory, R=>R mappings, physics, geometry, matrix manipulation and the only possible connection to computer science would be replacing transistors with some sort of quantum switch. Even that connection is very tenuous, being much closer to electrical engineering, which is also in his title. Nice try though.

As for the former winning, it's been my experience that the experienced win because HR has no clue what decent code looks like, and simply makes the assumption that experience trumps all. That's a shame, since experience is both easy to lie about and often irrelevant. You can gain years of experience at something and never get better at it, especially if you work in a large company where decisions are made by committee.

Oh, and as for the Ivory Tower, I happen to be working in one at the moment (doing mathematical modeling of diseases) and even the people who don't code seem to know more about coding than many experienced coders I've cleaned up after.

This post has been edited by CTphpnwb: 22 January 2010 - 07:18 PM

[CTphpnwb]

One more thing about education and experience. I think it's very telling that there's a side discussion in this thread about the importance of Calculus in programming. Experience will tell most coders that they don't need to know it, but education proves otherwise.

With education comes many different kinds of experience that can and do have enormous impacts on how well you do your job, and those experiences simply will not occur if you just train at your specific skill set. In the case of Calculus, if you didn't study it in school then you'd have to be another Isaac Newton to realize its importance to programming. How could you model anything that changes without it?

Honestly, I can't imagine anyone with no college and ten or twenty years of coding experience writing the kinds of code I deal with every day. They'd be lost without an educational background that included something beyond high school math. Here, more experience would be harmful as those with more years of experience would be further removed from math studies, making the subject matter even more difficult.

The fact is that you can't code something you don't understand, and experience doesn't impart understanding, only skill. Skill without understanding is extremely limited whereas understanding can easily gain skill.

[seeP+]

Calculus would be great to know in programming. I don't think it is necessary as far as learning the language. I would think it would be easier to give the programmer the formula of a math problem then to give a mathematician a book to learn how to create a program.

[CTphpnwb]

seeP+, on 24 Jan, 2010 - 02:53 AM, said:

I would think it would be easier to give the programmer the formula of a math problem then to give a mathematician a book to learn how to create a program.

Then you would be wrong. Computer science can be thought of as just a branch of mathematics, like physics. All of the logic you use in creating a program comes from math, so the mathematician starts out ahead of the programmer, with the only exception being their knowledge of syntax and other language peculiarities. To a mathematician, that information is trivial. On the other hand, the programmer who needs a formula handed to them can't even be sure that they're using it correctly or in the right place!

Edit:
If you think I'm wrong about this, pick a forum... any forum... and look at the kinds of help questions people ask. Most of them have to do with programming logic, and even those that don't could be easily solved if the programmer were thinking logically, like a trained (educated) mathematician!

This post has been edited by CTphpnwb: 24 January 2010 - 06:33 AM