April 21, 2013

The decline of women in computing

Colossus Mark 2 computer at Bletchley Park, 1944-45
I can remember when Britain's giant WWII Ultra project that broke German Enigma machine codes was finally revealed in the early 1970s. Bletchley Park employed many thousands of people, yet it remained unmentioned for a quarter of a century after the war. I occasionally stumble upon a pre-1970 text that, in hindsight, is referring to Ultra (such as a reference by Admiral Samuel Eliot Morison to the allies success in U-Boat hunting in the second half of 1943 that is almost taunting the reader to guess), but not much.
TELEVISION REVIEW 
Secret War Heroes, Hiding New Work From Husbands 
By ALESSANDRA STANLEY 
Churchill called the thousands of puzzle-solvers and clerks who spent World War II at Bletchley Park secretly breaking enemy codes “my geese that laid the golden eggs and never cackled.” 
And almost as extraordinary as their work was — some say the decryption of Germany’s Enigma machine hastened the end of the war by as many as two years — their loyalty to the Official Secrets Act is almost impossible to fathom. 
Codebreakers kept silent about their war effort for decades; the British government didn’t officially recognize Bletchley Park veterans until 2009. Nowadays, it is still possible to read newspaper obituaries of 90-year-olds who never told their spouses, parents or siblings what they really did during the war. 
“The Bletchley Circle,” a three-part series that begins Sunday on PBS, finds an imaginative way to give overdue credit to those unrecognized government servants, most of whom were women. 
The series opens in 1943, but it’s actually a murder mystery set in 1952.
Anna Maxwell Martin (“Bleak House”) plays Susan, a bored housewife and mother of two who detects a pattern in a series of unsolved murders. When the police won’t follow up, Susan enlists three former colleagues from Bletchley Park to help her decipher the serial killer’s modus operandi. 

One of the things we learned from the declassification of Bletchley Park was that a couple of years of the history of computers had been missing.

And that leads to an interesting demographic point: because computing was going in rudimentary form during WWII, a lot of women worked on computers.

A famous example is that the most widely used programming language in corporate America during the last 40 years of the 20th Century was COBOL, which was more or less devised by Admiral Grace Hopper. She had worked on the Harvard Mark I computer during WWII.

Her COBOL was notoriously verbose, the Chatty Cathy of programming languages, but it got an awful lot of work done. Not surprisingly, lots of women were COBOL programmers. (My wife was one for awhile.)

Today, the media recurrently gets worked up over the small (and quite possibly declining) number of women in the computer field.

One reason for why women aren't employed as much in computer programming these days is because languages have evolved away from COBOL's Englishness toward abstraction.

But another reason women have gotten squeezed out of programming is that government policy has responded to billionaires' demands that computer programming no longer be a middle class career appealing to American women. Instead, it should be a two-tier business with brilliant male programmers making death or glory bids to gain riches in Silicon Valley, while in the lower tier, American women are replaced by South Asian men via the H-1B visa.

The Gang of Eight wants to nearly double the number of H-1B visas, which will just continue to push American women out of computer programming.

90 comments:

Anonymous said...

1. Why aren't they being replaced by South Asian women?

2. Why is it that in countries that don't have mass immigration, such as Japan, computing still isn't really a woman's game?

Anonymous said...

Slashdot used to run polls of the sex of their readership. E.g., My Gender Is. The results were always overwhelmingly male. The sex ratio at Slashdot is similar to that of Wikipedia contributors.

ProNorden said...

If a single man has a reliable family-wage job, it leads to a 'wife & family' and building community. If a single woman has a family wage job, it leads to ...her waiting for a man with even better prospects...and just waiting until death.

Red Pill Theorist said...

I think an interesting aspect of this trend is the greater elasticity of female participation in sectors of the labor market. There's a lot of evidence that women are much more sensitive to working conditions than men are, in large part because they're the second income rather than the first. Another question is where all those COBOL programmers went, ie why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes and female programmers? Obviously it's related labor supply elasticity, but that doesn't seem robust enough to explain everything.

Anonymous said...

What is the precise kilotonnage of weed that you bong-huffed prior to writing this piece? 6? 86?

"why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes"

Because women hate programming and are bad at it. Am I the only one here not stoned?

Claudia Z said...

You can limit the number of unattractive South Asian programmers and thus make the field more attractive to American women, but is that a good thing? Does that benefit society as a whole?

Women don't need to be programmers if they don't want to be surrounded by uncool Asian peoples. They can do something else.

If you don't love programming, don't do it.

Anonymous said...

I'm a graduate student in computer science, but I know nothing about COBOL. I mostly program in C++.

I get the impression that the sort of work these supposed female "programmers" were doing decades ago was fundamentally different than modern software development. How many of those British women knew anything about cryptographic theory? I'm guessing none.

I don't buy the immigration angle here. Women today don't go into computing for the same reason they don't go into any other STEM field: their interests and abilities lie elsewhere. You try too hard to make everything about immigration, Steve.

Anonymous said...

I think an interesting aspect of this trend is the greater elasticity of female participation in sectors of the labor market. There's a lot of evidence that women are much more sensitive to working conditions than men are, in large part because they're the second income rather than the first. Another question is where all those COBOL programmers went, ie why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes and female programmers? Obviously it's related labor supply elasticity, but that doesn't seem robust enough to explain everything.

It could be an intellectual [meta-intellectual*?] version of two phenomena of modernity which we discuss a great deal here at iSteve:

1) Geographical clustering and isolationism [contrary to an original promise of the WWW]

2) Group maximization strategies within a specific activity [e.g. generic "Potels" seizing control of the motel industry, or Scots-Irishmen coming to dominate the various fiat electron scams, to include the very creation of the fiat electrons in the first place]

On the flip side, I know a rather prominent female academic physician who says that once a specialty within medicine gets a reputation as a female ghetto, you can't get any of the male medical students to consider it for their residencies anymore.

She says that once the chicks start to move in, the guys depart en masse.

Bottom line, though, is that there is something about modernity which seems to be encouraging and accelerating these sorts of trends.



*Here I'm thinking of the fact that we no longer have "chatty" programming languages which would mimic the way that women go about thinking [and resolving] the problems in their lives, and ergo might actually appeal to girls and young women as a offering the possibility of a skill worth mastering.

Actually, the more I'm thinking about it, the more I'm wondering whether Java [or any other "C-family" language] should only be used for teaching introductory programming to boys, but that COBOL should be used for teaching introductory programming to girls.

Huh.

Maybe we've made a big pedagogical breakthrough here at iSteve.

Anonymous said...

Task specialization has changed the role of women in computing during my work career. In the 70s and 80s there were many women writing custom COBOL and PL/I code in big business and government. These women typically had subject matter expert (SME) knowledge of a particular business area and used that knowledge to analyze business problems and write COBOL or PL/I to address them. This was typically done in a batch environment, there was little expectation for the immediate gratification of real-time processing in most areas of business and government at the time. However, IBM developed CICS to support rapid on-line transaction processing in real-time. "Green screen" terminals abounded in this new world and the increased complexity kicked off the task specialization. Either you could learn CICS or not and typically this ability determined whether you would be "functional" or "technical." The new graphical user interfaces in the 90s accelerated this task specialization and women usually ended up as "functional." Most of the classic computer nerds and subsequently low-paid foreigners were "technical." Interestingly enough, these days there still is money in "functional", e.g. getting PeopleSoft to work for the HR department at General Dynamics. Much of the "technical" work has been shipped to Bangalore. Much of the H1-B visa action is "functional" or "managerial" Indians required to make the offshoring work in addition to high-end "technical" folks to lower US contractor billing rates.

There was a recent media uproar in Canada involving the Royal Bank of Canada and outsourcing of a department to India. It appears RBC was bringing in Indians on temporary work visas to learn the Canadian jobs prior to the outsourcing and the public outcry resulted in a mea culpa from the head of RBC. Not an unusual process, BTW, just unusual media attention and a bizarre public apology from RBC.

Mitch said...

Instead, it should be a two-tier business with brilliant programmers making death or glory bids to gain riches in Silicon Valley, while in the lower tier, American women are replaced by South Asian men via the H-1B visa.

Part I: Blame google and the insistence on advanced math. Advanced math doesn't have a whole lot to do with about 90% of computer programming, but it keeps out a lot of women.

Part II: death of IT in corporate america, which used to spend billions on custom software solutions. IT In the 80s had tons of people, both men and women, with English and history majors who were extremely good programmers. But in the mid-90s, funding for IT dried up. Instead of having a big inhouse team, it went to consulting companies. More up front money, less overhead. Consulting companies required lots of travel-although for years, people (again, men and women) could just be a specialist in their particular field in their particular geographic area.

And then Bill Gates paid billions flood the zone with cheap Indians. Rates plummeted in the early oughts.

So people who had a technical bent but no real interest in advanced math could make a very nice living as a techie for about 20 years (raises hand). But women, as a group, don't much care for advanced math, don't want to travel constantly, and are more interested in solving process and management problems with technology than in inventing new technology.

It's too bad. IT in the 80s and early 90s was fun.

Mitch said...

Another question is where all those COBOL programmers went, ie why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes and female programmers? Obviously it's related labor supply elasticity, but that doesn't seem robust enough to explain everything.

It wasn't just COBOL. Women were highly represented and well-respected in all areas of apps programming (higher level), business processes and workflow management. There were fewer network and DBA specialists, but they were still a good chunk of the field (20% rather than 40%, say).

No, it was the change in the work environment and education requirements that drove the changes, I'd say.

sykes.1 said...

Adm. Hopper spoke at Ohio State a number of years ago (80's ?), and she claimed then that 80% of all the computer was written in COBOL. The reason being that almost all code was business or finance related, and COBOL, whatever its other defiencies, was well suited to that task.

Large amounts of FORTRAN and BASIC are also still being wriiten.

TGGP said...

Female programmers seem to be disproportionately foreign. An example of the "Swedish paradox", where the genders are forced to be more similar in poorer countries, while richer folks are free to "self-actualize" into greater divergence.

Anonymous said...

Computer science is like most technical fields--take a few years off when your kids are little and your technical knowledge becomes obsolete; you'll never be able to get back unless your former employer wants you.

Anonymous said...

I interview applicants to a top tier tech school from some affluent communities in the midwest. The boys interested in computer science are heavily influenced by their fathers. They've watched them work at home, upgrade computers, etc. and have learned to program outside school. I just don't hear that from the girls. The gender ratio in the programming classes at the local high school is what you'd expect too. All in all, I'd guess that women in CS enter college with less related experience, which puts them at a disadvantage from the beginning.

Anonymous said...

While it is true that there were women programmers in computing's early days, I think you are somewhat exaggerating the numbers and role of women. Note that all programmers weren't necessarily writing the actual code, but many were doing secretarial type tasks such as organizing and operating the punch cards.

Anonymous said...

I'm not convinced by this idea that programming is becoming less wordy.

It's true that a modern programming language tends to have a more symbolic core syntax than the old imperative ones, but the programs themselves are much more verbose and linguistically complex.

In fact, the reduction in the number of reserved words in the language just means that there are more English words available for use in the program in ways defined by the programmer. Fluent interfaces are the norm.

The dominant way of thinking about programs now is functional programming, in which you build up your own language of functions yourself.

It's true that to do very simple things: add numbers, say, you write fewer English words and more symbols in modern languages than you do in COBOL, but the point is you just don't do very simple imperative things in modern programs; those programs have been written. And in any case, if you wanted a verbose, wordy way to do those sorts of things for some reason you could easily create it because functional programming languages are completely extensible.

Whatever your program does, typically you will have created your own fluent interface that you have written yourself using your own words.

I am a programmer and I think much more carefully about the names I give things now and the English fluency of the code I write than I did 15 years ago.

Francis

Class factotum said...

I hated COBOL. We had to learn it, along with Fortran, in my BASIC class in high school. It took six pages of code just to tell the computer to turn the page.

David said...

Anonymous #1, because South Asian men are about on the level of white women, while South Asian women are below this level.

Cail Corishev said...

You described the change at the end perfectly, Steve. They're turning it into a feast-or-famine business for American programmers. It's becoming harder and harder to find work as a "good programmer." You can still be a great one and do very well, because there aren't many great ones overseas. But "good" programmers, like good workers in other fields, can be replaced by larger numbers of much cheaper replacements. So just making a reasonable living writing good code is out. Be great and make big bucks, or find something else to do.

To the first commenter, Steve explained why: programming languages have become more abstract in the process of becoming more powerful. COBOL is to languages like C, Perl, and LISP as writing a grocery list is to working through a geometric proof. Women don't do modern programming for the same reasons they mostly don't do math and logic, regardless of what country they live in.

It may also matter that computers have become powerful enough to allow women to do other things with them, such as graphic design, without needing to get into the nuts and bolts of computing at all. So a woman (or man) with some interest in using computers has many friendlier options besides programming.

Anonymous said...

"..why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes and female programmers? Obviously it's related labor supply elasticity, but that doesn't seem robust enough to explain everything."

Obvious is the development of alternative programming languages which resulted in products that better served paying customers.

cipher

Cail Corishev said...

Another question is where all those COBOL programmers went?

Well, presumably many went home and raised kids after the war, and there was a stretch when few women were in the industry. By the 70s and 80s, when women were getting back into the workforce in droves voluntarily, COBOL was still used on a lot of legacy systems, but all the new work and teaching was focused on newer languages such as Fortran, Pascal, Lisp, and C, all of which were much heavier on logic and abstraction than COBOL was.

Anonymous said...

My mother was a computer programmer from the 1960s on, ultimately working for the government. There were a lot of female programmers then.

Even when I started working on Wall Street in the mid-90s at a financial information firm all the programmers were women.

What is funny is that the coder community today is obsessed with getting girls involved in coding, as though they are trying to get blacks into hockey or something,

Anonymous said...

"Why is it that in countries that don't have mass immigration, such as Japan, computing still isn't really a woman's game?"

I have a coworker who recently worked in Japan on a hardware-software project.

The Americans had to do all the software because the Japanese software as "women's work". "Real men do hardware." Seems to contradict the quote above.

I think this explains to some degree in Japanese failure to do well in the smartphone market.
Robert Hume

Anonymous said...

A lot of this has to do with the "design" of university departments.

In the beginning, say late 50s through late 60s, 4 types of departments competed to train programmers and teach computing: Math departments, EE departments, library "science" departments, and community college business/accounting type departments. Women and Cobol were heavy in the last two.

The math and EE departments basically won. The modern CS department for the last, say, 35 years has basically been a math/EE fusion. Math and EE are heavily male-oriented. The CS filter courses tend to be heavy math/algorithm type courses.

Library science is having a comeback, but in a more purely mathematical form, as "data sciences", aka Big Data.

For the longest time community colleges kept Cobol and teaching women to program alive, but this has died down as the need for businesses to write their own programs internally has died down. There was a time when even a modest-sized business would have its own programmers and write its own programs. A women could learn to program and, no matter where she moved, she could work near where she lived. She didn't have to move to silicon valley.

Like economics, CS also developed something akin to physics envy. If you weren't working on a hard mathematical problem, you weren't doing real work. Cobol was pretty easy, just about anybody could do it! So it was largely uninteresting to researchers. The CS departments were not software engineering departments and they sure weren't business record-keeping or accounting departments (which women seemed to like). We seem to still be a long way off from having a real field of software engineering (women in general don't seem to like any form of engineering).

Today there's much less need for companies to write their own internal programs. (BTW, I think many in silicon valley today have a unhealthy attitude that all programming is really done in a few places like silicon valley, and of course the hicks in other places couldn't possibly program. I sometimes tell them it's the other way around, there was a time when programming was common everywhere across the US. Often they don't seem to really believe me.)

Of course now a US woman would be insane to go into programming as a primary career. She'd be competing with maybe a billion or so young single men who would almost be willing to eat grass before giving up the idea of getting rich programming. Not to mention all the married third-worlder two-fer couples willing to take a chance working for whatever the captains of commerce provide.

hingerr said...

"..waiting until death."
Huh? Who do you hang with?
There's more to life than husbands children (for some anyway), and most women I know are too busy to realize they are waiting...maybe that's the prob. Anyway, some guy has to agree with her plans.
Now my sister married in her 30s, two kids, and is a computer programmer, six digit income, learned on the job, started in the 70s. Her husband was a RC vehicle mechanic HS grad. Don't ask me how they got together--well, he was good looking.
She got her GED while I got a Masters in a non-high-income field. She is still doing very well--far better tha I-- but she now works for an Indian owned company. Said company, in the midst of a great depression here, is bringing in more of "their" people. And the traitors we call the U.S. "government" wants more them.

DCThrowback said...

Steve: a master of the immigration jiujitsu argument.

Too bad nothing, not even arguments about women or minorities being hardest hit, can stop those who hate certain Americans so much they have to stop at nothing to change it.

Anonymous said...

But another reason women have gotten squeezed out of programming is that government policy has responded to billionaires' demands that computer programming no longer be a middle class career appealing to American women.

That's not why women aren't programmers. Even without immigration, women would not be programmers. A lot of the women "programmers" were not really programmers as we understand the term today. People who put the punch cards into the mainframes could be called "programmers". It's sort of like calling the women who were working in munitions factory during WWII "defense engineers" or something.

I understand the point is to make arguments against immigration, but you don't have to affirm feminist lies to do so.

And you're unlikely to win over feminists to anti-immigration views by telling them that more of them would be able to have exciting careers writing code.

a Newsreader said...

During WWII they didn't even have COBOL, or any non-machine computer languages. Most computer programmers spent their day translating equations (written by the physicists and mathematicians) and such into ones and zeroes. The skills required for such work were roughly the same as those of telephone operators and secretaries.

Nowadays, computer programs called compilers and assemblers do the work programmers used to do. What we call computer programming (or software engineering, or hacking, or computer science) now is the domain of highly trained nerds. For whatever reason, you don't find lots of women in the nerd crowd.

As an aside, another obsolete profession that was once dominated by women was called computer. They would sit at a desk all day and write out computations. I wonder why nobody complains about these women's jobs being displaced by technology. Must be sexism.

Anonymous said...

The Gang of Eight wants to nearly double the number of H-1B visas, which will just continue to push American women out of computer programming.

I see what you did there.

Auntie Analogue said...


It should not be overlooked or forgotten that Bletchley Park's successful codebreaking operations were formed on the singularly brilliant innovative work of Alan Turing, who happened to be gay. Its arguable that in WWII the brilliant intellectual output of one gay guy, by shortening the war by up to two years, saved more Allied - and enemy - lives - than any other individual saved.

Kaz said...

http://www.visasquare.com/h1b-visa/case-details/facebook-inc-610870/data-warehouse-operations-engineer-20569.html

Some of these H1B jobs pay surprisingly well..

E. Rekshun said...

"American women are replaced by South Asian men via the H-1B visa."

American women and American men are replaced by H-1B immigrants. But women could care less. They're more interested in fluff jobs in HR and government.

E. Rekshun said...

I earned a BS in Computer Science in 1986 at a private New England college. My class was about 25% American female; 0 foreign females or males. In '98 - '00, I took several graduate level CS courses at a large southern state university as electives for my MBA. Those classes were about 15% female - all foreign; no American females and I was the lone American male.

Anonymous said...

So the codebreakers kept silent about these things all those years.

I wonder how many other things Government workers have kept silent about?

Anonymous said...

Some iSteve-ish baseball links in the news today:

The Kennedy Meth
[was Roger Maris on methamphetamines?]

Baseball Software Can't Score What Jean Segura Did Friday
[runner running backwards stumps software package]

Also, iSteve favorite Edward O Wilson says you don't need to know math:

Terrible Advice From a Great Scientist

pat said...

I taught hundreds of computer science students. About one quarter of them were women as I remember. I never taught programming even though that's was my 'day job'. I supervised maybe a hundred programmers at one time or another - not a single one was a woman.

The only female technical professional I ever hired was a database administrator. The salary we were offering wasn't competitive so I took a chance on an under qualified woman in the hope that I could develop her. Nope - total disaster. That was private industry.

In government I had over a hundred women working for me and only one man. He was the only one who could program. We contracted out to EDS for the actual programming. EDS as everyone knows was largely all ex-military and male. All those women did semi-technical jobs that were largely clerical. No private industry organization would have had any of those positions.

You are right about the English like syntax of COBOL. It was created as a self documenting language. The extreme opposite was APL a very concise language with a math like syntax. It was lovely and elegant but you couldn't figure out what it did an hour later.

Mainstream languages - both the American FORTRAN and the European Algol - were more algebraic.

Every woman I've ever known liked geometry but hated algebra. I can't be the only one to have noticed this. The female brain seems to come a cropper on the shoals of algebra. So modern algebra like languages like C, C++, and Java seem to be out of reach for them.

I don't think women will be making a comeback in programming and I don't think immigration policy matters here.


Albertosaurus

David Davenport said...

Another question is where all those COBOL programmers went, ie why didn't COBOL and its appeal to women generate a self sustaining ecosystem of female friendly codes and female programmers? Obviously it's related labor supply elasticity, but that doesn't seem robust enough to explain everything.

You didn't fully read Steve's column and the hypothesis Steve presented:

One reason for why women aren't employed as much in computer programming these days is because languages have evolved away from COBOL's Englishness toward abstraction.

Some history: In the old days -- 1940's, '30's, and earlier -- there were rooms full of female and some male clerks with pencil and paper, adding machines and tables of logarithms -- human DO UNTIL loops -- trying to solve difficult equations numerically and iteratively. In other words, rooms full of people were employed to try to get the same effect as a computer program.

The human DO loops knew how to do certain limited, repetitive calculations. They didn't necessarily entirely understand the mathematics on which they were working... ballistics trajectories, radar-guided intercept courses, implosive force needed to start fission reactions, and the like. Mathematicians and scientists, mostly male, understood the math.

Early electronic -- more accurately, electro-mechanical computers -- such as the Bletchley Park machine required several human attendants. For data input, the machine had to be fed with rolls of paper tape punched with holes. The entire program could not be run at the same time. Instead, the program had to be subdivided into segments. To start up each segment, humans would have to load additional roles of paper tape or punched cards.

COBOL, which stands for something like Common Business Operating Language, was never much good for scientific or mathematical computing, and was never used much for the nerdy stuff. Here I part from Steve's story. COBOL's generational software companion, FORTRAN, was the programming language for science and math. Formula TRANslation was never a distaff specialty.

Nowadays, I suppose that Excel spreadsheets accomplish much of the bookkeeping clerical work that COBOL and COBOL programmers used to do. The contemporary equivalent of a woman COBOL programmer is an administrative assistant who does Excel and maybe some database scripts.

The sequence is: human DO loop -> computer tape and card feeder -> COBOL programmer -> office ladies with spreadsheets.

The long range trend is make such computer-related jobs redundant. However, government and quasi-goobermental corporate jobs such as Human Resources keeps getting bigger and more numerous, so former COBOL programmers have moved on to office jobs with newer job titles.

DR said...

"One reason for why women aren't employed as much in computer programming these days is because languages have evolved away from COBOL's Englishness toward abstraction."

If this hypothesis is correct we should see more women programmers in lower level and procedural languages like C++, C#, Java, PHP and Perl. We should see fewer women programmers in more abstract and functional languages like Haskell, Scheme, Clojure, Erlang, and R.

I don't know if it is or not. Anyone know where to find data like this?

eh said...

Nice try, Steve. Did you expect Barbara Boxer to go all anti-immigration now?

Anonymous said...

Does anyone else hate spreadsheets?

How about a simple searchable database instead for many tasks?

Steve Sailer said...

To sum up my historical argument:

During WWII, there were a surprising number of women present at the creation of computing.

Some of them went on to careers in computing.

One of them, Grace Hopper, created a very female friendly programming language.

COBOL became the American corporate standard.

And American corporations employed, at least from the 1970s onward, a lot of American women to write code in Hopper's female friendly language. This provided a lot of moderately well-paid jobs 9 to 5 jobs to women.

Over time, programming careers in America have evolved toward a Go Big or Go Home model where the best American programmers (largely male) compete for high pay and, potentially, riches, while drudge work is outsourced, or insourced to foreigners assisted by the H-1B visa system.

candid_observer said...

The thing about COBOL that permits less capable programmers to flourish has relatively little to do with how English-like it is -- though that does make it friendlier.

It is really the class of problems it ordinarily handles that mek the difference. COBOL is, of course, oriented toward business problems. Historically, those have included such things as doing payroll, and printing reports from large datasets. These do NOT involve complicated or abstract algorithms, or, say, the use of pointers, or scientific calculations.

It's really the handling of complicated algorithms, machine architecture, high level math, and the engagement of the architecture of enormous and enormously complex software systems, that can render programming hard. Those functions are virtually never a part of a COBOL program.

Anonymous said...



It should not be overlooked or forgotten that Bletchley Park's successful codebreaking operations were formed on the singularly brilliant innovative work of Alan Turing, who happened to be gay. Its arguable that in WWII the brilliant intellectual output of one gay guy, by shortening the war by up to two years, saved more Allied - and enemy - lives - than any other individual saved.


Impossible.

Back then all gays were either killed by KKK dudes or blacklisted by uptight WASPs. Oh, wait...maybe this is just another example of diversity before there was diversity.

Cail Corishev said...

Some of these H1B jobs pay surprisingly well..

For now. When illegals and guest workers first started replacing Americans in the slaughterhouses, they probably got paid very well too -- for a while. Once they became the norm, the wages could come down and push the rest of the Americans out.

Callowman said...

COBOL was promoted as the programming language that would make everyone a programmer. However, it was inevitably densely logico-mathematical, just like every other language, and thus played to the same cognitive strengths as every other language. The only difference was, in COBOL, the logic was masked by verbiage, making it ever so slightly more difficult to use.

I'm guessing the higher percentage of female workers early on was due to the persistent need for data entry.

H1B visas - now there's an anti-immigration argument! Girls: not so much.

Cail Corishev said...

If this hypothesis is correct we should see more women programmers in lower level and procedural languages like C++, C#, Java, PHP and Perl. We should see fewer women programmers in more abstract and functional languages like Haskell, Scheme, Clojure, Erlang, and R.

I don't know if that's true, but I suspect it would be the case, although the numbers are probably low enough in all of them that it would be hard to see past the sampling error. I'd expect to find the most women in Java and PHP -- Java because it's been heavily pushed by PHMs and schools for a while, and PHP because it's ubiquitous enough that it's used by a lot of non-programmers to write programs.

Anonymous said...

Feminists say it's wrong for women to say in an abusive relationship.

Why are American conservatives in an abusive relationship with neocons?
When neocons were without a home after exile/expulsion from liberal/Democratic community, American conservatives offered them a new home in the spirit of having a bigger tent. So, what did neocons do? They took over and purged all the elements of the right they didn't like. Many of them work in cahoots with Obama and liberal Zionists to remake conservatism into 'gay marriage' and 'amnesty'-embracing ideology.

So abusive, yet American conservatism remains in this relationship. How many bruises and bleeding must conservative suffer under this abusive boyfriend to finally call it quits?

I suppose... even though neocons are abusive, they buy us nice jewelry once in awhile, and we are so dazzled by their wealth.

Kaz said...

@Anonymous of 12:47PM

That's probably a bad example. Alan Turing was convicted of being gay and killed himself because of all the shit that came with it, i.e. the chemical castration.

Anonymous said...

I'm kinda offended by Roissy-isms in some of the comments.

I'll take this one from Red Pill Theorist...

"They're the second income rather than the first"

No. Wrong. Normal, Healthy White Women Don't Like Asian Engineers--Male or Female.

Before I read Steve and Vdare, I was in college as an engineer with my white girlfriend and we had some very Steve-Esque private conversations. We considered ourselves to be The Last Of the Americans in our Classes.

The white woman interracial marriage rate is very very small. We don't like working with minorities, and White Men shouldn't Want Us To (because white men should want white women only surrounded by white men for obvious impregnation reasons.) I even remember being at work and this asian guy kept hitting on me and it was like Ew Ew Ew Ew.

So That is The Reason.

White women are really quite loyal (unlike white male engineers who are more than happy to marry asians women...Asian Women are Story of Bravery and Love for those of you who missed that)


Anonymous said...

*Impossible.*

*Back then all gays were either killed by KKK dudes or blacklisted by uptight WASPs. Oh, wait...maybe this is just another example of diversity before there was diversity.*

There's a foot in the mouth. But then I'm so fucking tired of hearing about Alan Turing that I forget not everyone's heard of him even once.

Lady Engineer said...

@ pat - One of the freshman weed out classes for electrical engineering majors in the early 80s at my college was a one semester APL class. I remember lots of arrows and parentheses and the solution could be done in one extremely long line of code. I used to have my male (can't say I knew any female CS majors back then) CS dorm buddies help me with the homework when I was stumped.

Whiskey said...

Women in an industry tend to flock in there where men who are desirable mates cluster. Thus, nurses and technicians work in health care around commanding, demanding, a-hole doctors. Meanwhile, if you're a guy, would you be a nurse? Really? A female-dominated job? That alone gets you penciled in "gay" and hurts your mate prospects.

One of the reasons women departed IT in droves, was that men went from commanding, demand, high-status professors and mathematicians, to nerdy guys wearing pocket protectors who had little status and prestige.

Even in education, there are coaches, principals, and such who are commanding, demanding, and authoritative. In a society marked by near absolute female freedom, the premium paid on male sexiness is also nearly absolute. Women simply won't put up with unsexy men dominating the male part of the workforce. It is the equivalent of guys surrounding themselves with women who look like Kirstie Alley.

Anonymous said...


She says that once the chicks start to move in, the guys depart en masse.


Funny. Eleanor MacCoby in The Two Sexes: Growing up apart and coming together made the same observation.

Anonymous said...

Me: On the flip side, I know a rather prominent female academic physician who says that once a specialty within medicine gets a reputation as a female ghetto, you can't get any of the male medical students to consider it for their residencies anymore.

She says that once the chicks start to move in, the guys depart en masse.


YOU: The white woman interracial marriage rate is very very small. We don't like working with minorities, and White Men shouldn't Want Us To (because white men should want white women only surrounded by white men for obvious impregnation reasons.) I even remember being at work and this asian guy kept hitting on me and it was like Ew Ew Ew Ew.

So That is The Reason.


So Men don't like being around women at work [too much politically correct Nurse Ratched horseshit to have to deal with], and white girls don't like being around Asian guys [at all].

Sounds like maybe folks are choosing their careers not on the basis of what they would ENJOY doing for a living, but rather on the basis of FLEEING from the co-workers whom they know they would HATE having to work next to for a living.

Which, in turn, would tend to agree with another big iSteve theme over the years, Affordable Family Formation: Young white people can't afford to have babies because they haven't yet saved enough money to be able to afford to purchase a home in the tonier neighborhoods with the good schools - i.e. young white people might otherwise LIKE to have babies, but they can't yet afford to FLEE the blacks & the mexicans & the muslims who infest the affordable parts of town.

Kaz said...

"The white woman interracial marriage rate is very very small. We don't like working with minorities, and White Men shouldn't Want Us To (because white men should want white women only surrounded by white men for obvious impregnation reasons.) I even remember being at work and this asian guy kept hitting on me and it was like Ew Ew Ew Ew."

You're a weird guy aren't you?

Anonymous said...

"One of the reasons women departed IT in droves, was that men went from commanding, demand, high-status professors and mathematicians, to nerdy guys wearing pocket protectors who had little status and prestige. "

No, when I went into programming in 1984 a lot of my fellow trainees were female. They were attracted by the high salaries - several had left maths teaching jobs - and the high status of the job.

Nearly 30 years on, the few young programmers who exist (most IT has been offshored) are probably no less nerdy than we were - but they're certainly not in it for the money or status. Hardly any are female. Because 'IT professionals' have gone from being valued employees to being just another resource to be exploited.

(And the teachers have mostly gone back to teaching)

Anonymous said...

"Every woman I've ever known liked geometry but hated algebra. I can't be the only one to have noticed this. The female brain seems to come a cropper on the shoals of algebra. So modern algebra like languages like C, C++, and Java seem to be out of reach for them.

I don't think women will be making a comeback in programming and I don't think immigration policy matters here.


Albertosaurus "

on TIMSS 8th grade girls from 22 countries outperformed boys on algebra, in only 3 boys did better. The story was reversed on geometry.
albertosaurus, you so funny!

Anonymous said...

"Would the Real Sarah Hanson Please Stand Up?"

http://www.theatlanticwire.com/technology/2013/04/would-real-sarah-hanson-please-stand/64412/

"What happens when a 19-year-old programmer extensively covered by media outlets this week probably doesn't exist?

Here's what we think we know about Sarah Hanson: she's a 19-year-old who dropped out of an unnamed college after her first year to start Senior Living Map, a start-up that helps people find senior living centers without the help of a care adviser. She allegedly lives in Seattle, Washington. She also allegedly sold 10 percent of her income over the next ten years to an anonymous San Francisco angel investor in exchange for an $125,000 investment in her company, via an auction website. Other than that, there's no trace of her existence on the internet. Anywhere.

Which seems weird. For a 19-year-old who has allegedly been coding since she was 12-years-old, you would think there would be a Facebook account, a Twitter account, even a LinkedIn, maybe. There might be something, but there's not. And it wasn't until after VentureBeat's John Koetsier published an interview with the alleged Hanson that the evidence tipped him off that this person he's emailing with -- not speaking with in person or on the phone -- may not exist."

pat said...

Steve is right about computer languages and women and I'm even more right.

One reason there were probably so many women involved in Ultra was that there were so many preliminary clerical like steps in preliminary processing. At about the same time in America we had a similar nexus of intellectual effort at Los Alamos.

Richard Feynman was there and his job was 'computer'. It wasn't until later that machines inherited the name. The senior computers organized a room full of subordinates (menials) who would calculate answers. You would for example have a classroom of technicians arranged in columns and rows who passed partial answers from one to the other. A whole room full of men - in the American case - would be needed to invert a single matrix.

I imagine that this was the sort of thing they were doing with all those women at Bletchley. The early computing machines could do some tasks but most of the mathematical grunt work was done by human grunts.

Remember every person reading this blog gets their paycheck from a system coded in COBOL. Almost all American business processes are written in COBOL.

Let me mention documentation again. That will help explain why. If you are tasked with maintaining a large software system you typically get an order to enact some change. The first thing the systems analyst does is to assign a coder to do the research. They have to figure out what the code does, how it works and what other processes it effects. In a big system it takes much more time to do this research than it does to write the new code.

Big business systems are filled with code written decades ago. A lot of it does nothing. There are patches on patches. Occasionally management will authorize cleaning out the dead code but such efforts can cost hundreds of thousands of dollars and produce no immediate results at all. These are the kind of projects that get pushed aside.

If you write a routine that solves a problem a week later you will have forgotten how you did it. If it was clever it will be even harder to figure it out again. That's why production coders are told to document their work. There are several ways to do this. There are separate notes and there are in-line comments for example. COBOL was created as a self documenting language. Its syntax was prolix on purpose. The idea was that it could be read by the new coder even if the old coder had neglected any other documentation. It could also be read - at least a little - by an ordinary business person.

Women of course score higher on the verbal parts of IQ tests and less well on the spatio-math sub tests. So in this sense Steve is exactly right. COBOL is a female friendly language.

But as we all remember with the Millennium Crisis its hard to find COBOL coders anymore. When I was teaching at a local JC they were still teaching RPG - a completely obsolete language. Academics like to teach what they had once learned themselves - process that assures that students at public institutions will always be learning technology a generation or more out of date.

Nobody has ever written a website in COBOL. The early Microsoft web technologies used VBScript which was based on BASIC which was in turn based on FORTRAN. The COBOL prolix English style syntax has died out because Grace Hopper's next project after COBOL was a similar style DBMS - a non- relational DBMS. That was a dead end. ADA was another language that the federal government tried to establish the way it had established COBOL. That effort failed. Then they tried to create a similar massive governmental networking scheme with ISO-OSI. That too failed. The TCT-IP protocol suite was victorious.

All the big federal attempts at replicating the female friendly COBOL phenomenon were busts.

Albertosaurus

Anonymous said...

Sheryl Sandberg, of Larry Summers and Facebook fame, sure seems to get her face in the papers a lot. What it has to do with computers I'm not clear.

media computer expert said...

I think I'll watch this show cuz that Sophie Rundle is really stacked

sexist said...

Steve: why do you think in The Life of Julia she was a "web designer" rather than a database query debugger? All things being equal women would rather be doing something even minutely more glamorous. Let's not now pretend a generation of mathematically-capable distaff have been shortchanged by the big bad Patriarchy again--at the end of the day, they're chicks and will tend to think & behave in chick-like ways.

Anonymous said...



Which, in turn, would tend to agree with another big iSteve theme over the years, Affordable Family Formation: Young white people can't afford to have babies because they haven't yet saved enough money to be able to afford to purchase a home in the tonier neighborhoods with the good schools - i.e. young white people might otherwise LIKE to have babies, but they can't yet afford to FLEE the blacks & the mexicans & the muslims who infest the affordable parts of town.


Okay, but what about Europe home to lots of white people who get lots of gov't freebies to make family affordable and until recently there were damned few dangerous minority neighborhoods, yet American whites outbred them while at once being ethnically cleansed from cheap housing.

Anonymous said...

I'd guess that programing used to a 9 to 5 job which paid a decent wage. Increasingly these days, it isn't either of those things.

Evil Sandmich said...

Yes Mr. Sailer, no one has taken down your point that women will take a hit, perhaps disproportionately, from the gang of eight's plan to screw over the white middle class so that the white upper class can make a buck. They actively hate white working men, so maybe if it's framed as a 'women and minorities hardest hit' argument it might stick, though it hasn't so far (the 'women and minorities' don't seem to care much that they're getting boned).

Anonymous said...

Twenty years ago, when my son was in Boy Scouts, I had to ask my wife not to go camping with us. She learned to cook over a wood fire in Mexico as a girl, and could not stand to see what those boys did to the food they were cooking.

So, she would run up and "help". As soon as she did the boys would turn and walk away and go play. Mom was here.

When I retired in 1997, I took an ICS course in COBOL, out of curiosity. I dinked around with the, I think it was, RM-85. I learned you could actually do trig functions, but it was very very slow.

I read a year or two that actually a significant percentage of lines of programming on the planet are still today in COBOL.The large companies use it so much, and it costs so much to change anything, it still occupies a lot of programming time. There may be more efficient languages, but there are so many lines in use, it would cost a major fortune to convert, so it has never been done.

When they worried about the 2K bugs, they were running all over the place hiring COBOL people.

Anonymous said...

"People who put the punch cards into the mainframes could be called "programmers".


No, this is flat wrong. I knew plenty of women programmers in the day.

The women who put punch cards into computers (and hung tapes) were called operators. Not all operators were women. Because these people were called operators, when their function was automated, the system that replaced them was called an operating system.

(I was an operator for awhile. Sometimes in small shops programmers had to operate the computer as well.)

There were lot of women actually programming. Maybe 50%. There was also keypunch, which was 90% female or more.

An important reason for the high numbers of women in the field back in the 60s and 70s was that you needed to learn how to program, but you did not learn to program in computer science departments back then... because there were not that many computer science departments at the time.

So career paths were different. People got into programming via all kinds of paths. It turns out than in a lot of areas, women were very competitive as programmers. It also turns out that computer science doesn't necessarily have that much to say about the core logic of, say, a payroll program. The accounting department did, and there were lots of women in the accounting department. My sense is there still are.

So one thing that's different today is we now suffer from, "oh, of course we need to hire someone who has a BS from a CS department if we need a programmer, no one else could possibly code."

Computer science in the sense we know it today is about a lot more than programming. The stronger the department the more it tends to be a subdepartment of mathematics and the stronger it tends to filter women.

Anonymous said...

"Note that all programmers weren't necessarily writing the actual code, but many were doing secretarial type tasks such as organizing and operating the punch cards."

No, many were writing the code. I worked with many women programmers in fairly large shops. Some shops probably had more women programmers than men. Programming really was gender neutral in the height of the Cobol age (say, 1970).

I do think a lot of women moved into programming from keypunch. They started by reading the Cobol code they were keypunching and said "heh, I could do that".

Anonymous said...

"Because women hate programming and are bad at it. Am I the only one here not stoned?"

No, you're just factually wrong. Women might hate programming today, but they just didn't then. I mean I was there. They didn't. Programming was just a job skill. Women were pretty good at it in many areas.

Perhaps interactive terminals, computer labs, the modern internet culture, and commercial programming has somehow made the whole programming culture different? Doesn't seem like it could be the whole explanation, though.

An important point is that the modern long-term software license (and making big-bucks in programming) didn't happen until very late, say 1967. Up until then you didn't make money programming, you made money doing something else, programming was just a necessary side skill, like running a sort machine.

I think women hate what the modern CS department has become and young women now know that programming is a field in which they will run into "aspie" young men who will always beat them at programming (and can be unpleasant to boot).

One point demonstrates that women can be just as good at day-in day-out software engineering as men. Take a random sample of the Indian and Chinese programmers in silicon valley. Say in a large software shop like Oracle. I think you will find that something like 25%-40% are women. They seem to do ok. Often they are part of a "two-body" package (both husband-and-wife are in silicon valley to make money).

If you don't think there are plenty of women writing code in silicon valley, you are mistaken. I'll grant you that they might not be doing it if they weren't making a lot of money.

Anonyia said...

"Women in an industry tend to flock in there where men who are desirable mates cluster. Thus, nurses and technicians work in health care around commanding, demanding, a-hole doctors. Meanwhile, if you're a guy, would you be a nurse? Really? A female-dominated job? That alone gets you penciled in "gay" and hurts your mate prospects."

Men are actually entering the nursing field in droves right now. Probably because it is widely touted as an in-demand field with good pay.

Anonymous said...

One point demonstrates that women can be just as good at day-in day-out software engineering as men. Take a random sample of the Indian and Chinese programmers in silicon valley. Say in a large software shop like Oracle. I think you will find that something like 25%-40% are women. They seem to do ok. Often they are part of a "two-body" package (both husband-and-wife are in silicon valley to make money).

Of course it wouldn't be that the Chinese and perhaps also the Indians are coming from a group with a few point higher average IQ?

I have met a few good Chinese female programmers in Silicon Valley and some non-descript Indian female programmers.

Of course, we are no longer writing payroll programs and you tend to have to understand really hard stuff.

Mr. Anon said...

"Anonymous said...

The women who put punch cards into computers (and hung tapes) were called operators. Not all operators were women. Because these people were called operators, when their function was automated, the system that replaced them was called an operating system."

That's interesting. I didn't know that.

Anonymous said...

Another point about women in computing.

It's common to think of computing as something that happened after WWII, but the computing industry, in particular business computing, goes back much further. For instance, IBM:

"...was founded in 1911 as the Computing Tabulating Recording Company (CTR) through a merger of three companies: the Tabulating Machine Company, the International Time Recording Company, and the Computing Scale Company."

CTR changed it's name to IBM in 1924. It was automating Wall Street in the 1920s and automating the US government in the 1930s that made IBM the colossus of computing. All this happened before electronic computers.

What did they use before electronic computers? Tab shops, sometimes, but not that often, called unit record computing. Tab computing resembles dataflow computing. Typically an entire floor or room would be converted into what was effectively a giant dataflow computer for card processing. Tab equipment had plugboards that enabled different calculations to be performed (as in a reproducing-calculating-punch) or formatting printing. Sorting and reformating (repunching) machines did a lot of the core work. Any database problem could be solved with cards, it just might take awhile...

I'm just old enough to have used tab equipment and have known tab "programmers". I think the social aspect around tab shops was entirely different than modern programming. There was a level of complexity, but it could be mastered by nearly anyone. You did not have to get a BS in "Tab Machine Science" to program tab equipment. You did not operate tab equipment in isolation. Due to things like keypunch, there were always a significant numbers of women in tab shop computing.

BTW, the Los Alamos project computing was largely based on IBM tab shop equipment, not human computers. There were books written in the 20s on how to do what we now call numerical analysis on tab equipment.

Most of the people involved in defining Cobol (for Pentagon accounting and record keeping, essentially) were women with tab shop experience. Many of the women Cobol programmers came from tab shops. It was an easy transition, as Cobol could be seen as an easier way to do tab shop computing.

Fortran (scientific and math based computing) was always about 10% of computing, as compared to Cobol's 90%. This breakdown between math-based and non-math based computing is likely still true today, though the languages have changed. University-level education today is targeted at the "Fortran 10%", minimizing the number of women. There is now no career path for large numbers of native-born women to enter the field.

You learn something new everyday:

"The term "Super Computing" was first used in the New York World in 1929 to refer to large custom-built tabulators that IBM had made for Columbia University."

Anonymous said...

"Of course it wouldn't be that the Chinese and perhaps also the Indians are coming from a group with a few point higher average IQ?"

Being able to compare the three groups, I don't think there's a detectable difference in raw intelligence. What's different is that the silicon valley female programmers are all about economic advancement, big time. They (and their husbands) are under tremendous family pressure to bring in the money. If anything, the focus of modern silicon valley women on income seems to me to cause fewer of them to be in the field because they have an innate enjoyment of it. I would describe most of them as "adequate" programmers. We're not talking coding greats like (maybe) Linus Torvald's crew of core lieutenants here. (All male, I might add, and I don't think any are Indian, a few are Chinese, most are boring normal white guys.)

(Also, isn't average Indian intelligence somewhere in the 80s?)

Anonymous said...

"...you tend to have to understand really hard stuff."

Really? Can you list some of this stuff? Perhaps the problem is we have failed as a field and made things difficult that should be simple. Are you saying that all programmers are researchers?

But what I've seen more of in silicon valley is simple ethnic nepotism. It is particularly easy, because it goes hand-in-hand with "we have a horribly bad reputation as a place to work, a sweatshop, I don't know what we can do... guess I'll have to call up some distant cousins."

Auntie Analogue said...


To:
Anonymous 12:47 & Anonymous 1:35 -
The following tells of the thanks Turing got from H.M. Government - which didn't need a KKK to harm gays - for his having worked himself into a frazzle throughout World War II to save millions of lives.

Relevant snip from Wikipedia:

"Turing's homosexuality resulted in a criminal prosecution in 1952, when homosexual acts were still illegal in the United Kingdom. He accepted treatment with female hormones (chemical castration) as an alternative to prison. Turing died in 1954, just over two weeks before his 42nd birthday, from cyanide poisoning. An inquest determined that his death was suicide; his mother and some others believed his death was accidental"

(And I apologize for having misspelled "It's" as "Its" in my original, earlier comment.)

Anonymous said...

Some posters here are perpetuating the myth that Indians are (in general) good programmers. They are not.

http://www.computerworld.com/s/article/9134122/China_dominates_NSA_backed_coding_contest_

Anonymous said...

"And it wasn't until after VentureBeat's John Koetsier published an interview with the alleged Hanson that the evidence tipped him off that this person he's emailing with -- not speaking with in person or on the phone -- may not exist." "

stuff white males like

Cail Corishev said...

Perhaps interactive terminals, computer labs, the modern internet culture, and commercial programming has somehow made the whole programming culture different? Doesn't seem like it could be the whole explanation, though.

Probably not, since programming was becoming a male nerd thing before most of those things came about.

Programming simply changed into something much more math- and logic-heavy, and therefore male. COBOL programming back in the day was mostly about reading records (sets of data) from the keyboard or storage, doing some sort of modification or analysis on it, and writing them back out. As someone said, the modern equivalent is Excel spreadsheet processing, and plenty of women still do that.

But what we call "programming" mostly changed into how to move dots around on a screen in certain patterns as quickly as possible, modeling things in 3D, doing rudimentary AI, and the like. It also got closer to the hardware for the sake of speed in many cases, which means closer to the absolute logic of 1s and 0s, which women don't like.

It's not so much that it's harder (though it probably is) but that it's different. A simple tic-tac-toe game isn't hard to write, but it takes math to layout the display and some simple AI logic for the computer player. It's not much like the record-keeping that COBOL was designed for, so the languages have changed to fit the job. It's much more like algebra than like accounting.

Was it necessary for programming and the languages to get harder? Well, yes, but it's hard to explain how. I did procedural and object programming for years, but the day I learned about first-class functions, I realized I'd been handicapped. The day a programmer learns recursion is a huge step: whoa, what can I do with this? The hard stuff developed in languages because it's powerful and lets us do things that couldn't be done without it.

Cail Corishev said...

Some posters here are perpetuating the myth that Indians are (in general) good programmers. They are not.

My impression from talking to typical Indian programmers on programming forums -- guys who are getting paid to program in the language in question, quite often with a "degree" from what they call a college -- is that you could get the same results by rounding up a bunch of 14-year-old American kids who play a lot of D&D or Minecraft and putting them through a one-month crash course in programming. No, I'm not exaggerating.

Cail Corishev said...

Although foreign programmers didn't push women out of the business, Steve has discovered an interesting parallel here. As someone pointed out, in the days of mainframes and computer scarcity, there was a separation between the people who did the difficult work of developing the algorithms and the people who did the laborious work of entering the code into the computer. Most anyone could be trained to do the latter job, and many women did it.

Faster, cheaper hardware and better interfaces changed that, so the guy designing the program could put it directly into code and get immediate results, rather than waiting for someone to schedule time on the mainframe and bring the results to him. That eliminated the go-between position of operator to a large extent.

Now, companies like IBM and Microsoft want to go back to a model similar to the old one, except instead of operators you have barely-trained code monkeys. You still have a few geniuses at the top coming up with the new ideas and figuring out how to make them work, but then you have rooms full of operators writing simple little snippets of code, each doing one task specified by the geniuses. Essentially, they want to build programs the way you built cars before robots: a few brilliant guys to design the new model and how it goes together, and a factory full of people who know how to bolt on one piece or attach one panel.

That wouldn't be a problem, except: they want the factory filled with foreign workers because they're cheap. I don't think it's the best way to produce software (open-source doesn't work that way, and it produces heaps of ridiculously good code), but if the code factories were filled with Americans getting paid like auto workers, that'd be good for America, and I'd be fine with it. It's the quest for cheap foreign labor -- and the bad-mouthing of Americans workers that they do to get it -- that I have a problem with.

Red said...

Women are very good at repetitive tasks, such as translating code to punch cards. Once programs became very large, difficult to debug, and required a deep understanding to code properly women dropped out. I work with 2 female programmer. Where as I love debugging and fixing problems more than anything else, they hate and become overwhelmed when things don't work right.

pat said...

Anonymous wrote:

TIMSS 8th grade girls from 22 countries outperformed boys on algebra, in only 3 boys did better. The story was reversed on geometry.
albertosaurus, you so funny!


I'm glad that I have kept you amused. But perhaps some of the risibility is simply that you chose your evidence so carefully.

Girls develop a few years earlier than boys. So at something like age 14 American girls are a little taller than boys of the same age - causing great anxiety in the boys. But of course boys soon run into their own maturational spurt and most end up taller than their female classmates.

Something similar happens in math ability.

Sex differences on quantitative skills do not appear with any consistence prior to the 10th grade. Meece and Parsons - Sex Differences in Math Achievement Toward a Model of Academic Choice - Psychological Bulletin Vol. 91 No 2.

That was the first study I found with my quick Google search. There are surely others.

By the 12th grade (high school senior) boys are clearly better than girls in math and of course girls are better than boys verbally. SAT and GRE scores attest to this phenomenon. This was what Steve meant by saying COBOL, being English like, is female friendly.

You seemed to have cherry picked the one statistic that shows the opposite situation. I am in fact wrong about many things often. But that's because I write a lot of quick comments and I often don't slow down to check my facts. It is not because I'm trying to deceive anyone. I was commenting on a sex difference that is so well known that I didn't think I needed to bother with a reference.

We need I think, however, to check on you.

Albertosaurus

Anonymous said...

Meanwhile in the IT factories in India:
'The failure of technical education and a glut of IT graduates are fuelling résumé fraud and a burgeoning background verification industry'

http://caravanmagazine.in/reportage/chequered-pasts

Truth said...

"TIMSS 8th grade girls from 22 countries outperformed boys on algebra, in only 3 boys did better. The story was reversed on geometry.
albertosaurus, you so funny!"


" I was commenting on a sex difference that is so well known that I didn't think I needed to bother with a reference."

I don't know there, Bertie. In a discourse, references are considered a good thing. To the casual observer, he/she appears to have Deandred you here.




Anonymous said...

Redmond 1978

Anonymous said...

The problem with programming is that, rail against this all you want but it's still true, is that it's a kind of creative writing. Less math, really, and more recognizing patterns in the problems that you're trying to break into bite-sized* pieces.

How many staff writers does it take to come up with a Tom Clancy novel? Just one - as long as he's as good as Tom Clancy.

A good programmer is worth 10 fair programmers, a great programmer is worth 10 good programmers, and the Epic programmer you want as your project architect is worth 10 of them. And there are two levels on top of that. A single idea on a napkin from Ken Thompson can change the world (Google UTF-8 story).

Women in those top three tiers? As rare as Women Fields Medal recipients.

* The pun would not have been pardoned.

shivratri said...

"TIMSS 8th grade girls from 22 countries outperformed boys on algebra, in only 3 boys did better. The story was reversed on geometry.
albertosaurus, you so funny!"


" I was commenting on a sex difference that is so well known that I didn't think I needed to bother with a reference."

> I don't know there, Bertie. In a discourse, references are considered a good thing. To the casual observer, he/she appears to have Deandred you here.

I guess casual observers aren't good with logic. The test scores of 8th grade girls are irrelevant when discussing _adult_ women in the workplace - particularly when the trend is reversed as they approach adulthood.

Notice how people who don't have logic on their side resort to inane comments and inane youtube clips to obscure the fact that they are bad at logic.

Anonymous said...

Anon 2:57 PM, don't diss Truth, homie is one of Sailer's most awesome commenters. He keeps the joint popping. Don't hate, appreciate!