"your chosen programming language is not l33t enough"

Massif
aim stupidly high, expect to fail often.

Hee hee, I'll just stir it up a little.

http://scienceblogs.com/goodmath/2006/11/the_c_is_efficient_language_fa.php

Dr Herbie
Horses for courses

Massif wrote:
Hee hee, I'll just stir it up a little.

http://scienceblogs.com/goodmath/2006/11/the_c_is_efficient_language_fa.php

The article is absolutely correct: I wrote my PhD simulations in C/C++, while they were based on an original model in Fortran, which was faster.

Why didn't I use Fortran? I don't like Fortran. Simple as that. My simulations didn't need to be faster, even though some of them eventually took 3 weeks to run. Time was not a factor, correctness was. I was more confident with C/C++ than with Fortran so I went with what I knew.

Herbie

Massif
aim stupidly high, expect to fail often.

Surely Dr. Herbie (if that is your real name Big Smile

. ) if you'd wanted mathematical correctness you should have used something like Haskell.

Which would have involved learning a whole new paradigm, but at least pure functional languages can be mathematically verified. (WARNING, I've just stepped out of my actual knowledge and I'm relying on Stuff People Have Told Me^TM)

WBurggraaf

Massif wrote:
(WARNING, I've just stepped out of my actual knowledge and I'm relying on Stuff People Have Told Me^TM)

LoL:D:D:D:D:DYou mean mr. wikipedia? Wink

. Met him to once Tongue Out

.

Dr Herbie
Horses for courses

Massif wrote:
Surely Dr. Herbie (if that is your real name . ) if you'd wanted mathematical correctness you should have used something like Haskell.

Which would have involved learning a whole new paradigm, but at least pure functional languages can be mathematically verified. (WARNING, I've just stepped out of my actual knowledge and I'm relying on Stuff People Have Told Me^TM)

My PhD was biology, not computer science. I had to use something that the people around me had a chance of understanding.
It wasn't mathematical correctness, but coding correctness. Again, it came down to a language that I was comfortable enough with that I could ensure it was correct.
A danger is learning a new language is not being completely sure that you know how it operates.

And Herbie is my real name. What can I say? My parents were hippies.

Herbie

JChung2006

WBurggraaf wrote:
Never understood why most of the female kind doesn't like programming because there is no space we you can create so much shoes en clothes as in the virtual world of programming.

I suspect a ridiculous comment like this is one good reason why there aren't more women programming.

bart7simpson7
0xCAFEBABE

Haskell is scary Expressionless

100 lines of someone elses code is like reading in binary

serishema
The Last Hacker Chick

JChung2006 wrote:

WBurggraaf wrote: Never understood why most of the female kind doesn't like programming because there is no space we you can create so much shoes en clothes as in the virtual world of programming.

I suspect a ridiculous comment like this is one good reason why there aren't more women programming.

The rediculous comments get anoying.
I also get sick of the oh wow a chick who does programming type comments.

The attitude of (some of) the open source community to us geek girls put me off linux completely. People on linux orriented irc channels eaither rubbish you or DCC you with pathetic obcessive will you be my girlfriend type nonsense.

Thus, i don't go on irc any more and i don't run linux.

ScanIAm
On a scale of 1 to 10, people are stupid.

We have a lot of developers who originated in china where I work (as well as india), and while the ratio isn't 50/50, it is probably 60/40 (men/women).

Minh
WOOH! WOOH!

It's weird. Here's what I'd observed -- I make no judgement as to why -- whatsoever:

> When I went through my CompSci studies, there were very few women in my classes. Maybe 3 out of 100, top.

> It was incredibly rare, when, on my last job, there were 2 women programmers -- in one place -- out of 8 people. Although, one eventually quit programming & went back to school.

> This trend worsen until maybe 3 years ago, when I see more programmers from India. Then the ratio of women was about 50%.

> Of couse, this is the mid-west -- whatever that means.

I hear that CompSci enrollment in college is on the decline, so that can't be good for the diversity. But it makes me wonder if there are some biological traits to be in programming.

Sven Groot
My name has 9 letters. Coincidence? I think not...

Massif wrote:
Hee hee, I'll just stir it up a little.

http://scienceblogs.com/goodmath/2006/11/the_c_is_efficient_language_fa.php

Very true. While working on my thesis I found that the Intel Fortran compiler would usually beat the Intel C compiler on code that was pretty much exactly the same, even when I went in and added the "restrict" keyword (a feature of the Intel compiler that tells it that two pointers don't overlap, I think it's also in the C99 standard) and #pragma ivdep (tells it to ignore any perceived flow dependencies when vectorizing), the Fortran compiler still beat it. It was trivial to defeat the vectorizer of the C compiler, anything more complicated than x[y] would lead it to say "Loop cannot be vectorized: dereference too complex".

Lots and lots of research exists (100s of papers and theses) in solving the problems of pointer aliasing for optimization. But even the best of those can only make a conservative estimate unless they have the full source of the application available so they can do cross-procedure analysis, and they tend to be very time-consuming approaches.

&quot;your chosen programming language is not l33t enough&quot;