Hacker News Reader

Comment by jojobas | Hacker News Reader

Also wondering what was the nature of of this work in the first place.

Hundreds of women familiar with assembly programming couldn't have appeared of thin air to go mentally execute their code and punch it into cards.

View on HN

There's a short book that covers this period of British computing history: Programmed Inequality: How Britain Discarded Women Technologists and Lost Its Edge in Computing, by Mar Hicks.

It's more history than sociology, though the title suggests otherwise. There are interesting anecdotes from women who worked from the 1940s-1970s for British intelligence, the General Post Office, and, yes, Dame Steve Shirley.

Note that the interviewer of Dame Shirley is this same Mar / Marie Hicks. link to the full transcript is http://www.computerhistory.org/collections/catalog/102738707

My guess is most of them, like Grace Hopper and Margaret Hamilton in the US, had math or physics backgrounds. Even assembly is comparatively easier than, say, formal logic or partial differential equations.

Just having learned maths or physics doesn't magically make you able to code, let alone blindfold-code.

Code is math. Variables, functions, and expressions are all mathematical terms.

The main difference is the time between feedback on your code. Back then, you'd only know if there was an error in your code when it failed to run. Now we get it almost instantaneously. But that's not how it always was.

Where I've studied abstract algebra, TFCV, partial diff. equations and quantum mechanics were mandatory courses, normally passed by everyone. Not all of them were cut to follow code paths, recursion etc (although a large fraction thrived with coding and are now in the industry, much larger than that of the general population). I don't think you could grab a random 60's mathematics BSc's and have them punch in business-ready code, male or female.

One more example the early spaceflight guidance computers were completely designed / engineered on paper in graph form, with mathematicians and engineers doing all the hard math work.

Assembly programmers would then turn the graphs into assembly using the best technique at the time, at moments having to use NOPs to cool down the ram.

For these people the source of truth was the graphs and the actual assembly was throwaway.

> assembly is comparatively easier than, say, formal logic or partial differential equations

Can you explain how, exactly, one compares the ease of assembly to the ease of formal logic, such that one can then make a definitive and sweeping statement about that comparison?

It is perhaps an opinion more than a fact, but an opinion that is likely widely shared by people who have done both. Computer assembly languages are not conceptually complex in the same way that higher level math is. The difficulty of assembly comes from its relative obscurity, the proliferation of instructions and architecture variants, and the process of analyzing a problem into the required steps. These are things that can be overcome with patience and reference manuals. At its heart, assembly instructions are doing simple things.

At the risk of an analogy, I would describe assembly programming as memory-intensive, and math as compute-intensive.

Having done both, I have to agree with OP that assembly language is far easier than PDEs.

maybe ibm wouldn't hire them back after they got married or had a baby

It was a booming market, why wouldn't they?

I guess it was a possibility, but then it's a prime example of capitalism fixing/exploiting the pay gap.

Even as late as the 1970s, it was not uncommon for women to automatically get fired once they got married. Some jobs, like secretary, nurse or teacher, were deemed suitable for unmarried women, but married women had to stay home, raise kids, and run the household. This left a lot of highly capable women outside the workforce.

A company specifically created to hire those women was a brilliant idea.

In the US the labor force participation rate in 1975 for women was 45%; today it is 57%. Was it really that much worse in the UK?

that 45% was to a significant extent poor women doing unskilled labor: maids, factory workers, cooks. not office workers

however, shirley did not found her company 'freelance programmers' in 01975

she founded it in 01959

the ensuing 20 years were among the most successful years in history for feminism in the usa and uk, marked by, among other things, rapidly growing female participation in the labor force

Even in 1959, the US labor force participation rate for women was 37%. So since that time, there’s only been about a twenty point increase.

Most people were poor, and married, and with children, so that would make sense if the real issue was underrepresentation in some upper class jobs.

if you were a computer programmer, high school teacher, or secretary before you got married, it might be the case that the only jobs open to you after marriage were factory worker and maid, both involving a substantial risk of sexual abuse and resulting marital rupture. even as a programmer, shirley reported experiencing 'being fondled, being pushed against the wall'

also, if you had a baby, usually you got fired because you couldn't come to work for months or even years (unless you could afford a nanny), and then you had to start your career again from the beginning at a new job

underrepresentation in upper-class jobs was a real issue, but far from the only one

roughly speaking, they considered it unethical

Reply on news.ycombinator.com