Comment by LanceH | Hacker News Reader

Comment by LanceH | Hacker News Reader

Parent Full thread

Reaching out and providing scholarships and generally giving a leg up is not how this gap will be closed.

It has never been about teaching women to code, except at the bare minimum "I can get a job at a giant company where I'll write 2000 lines of code in a year."

It's about right clicking with your mouse to see if there are any other options available. Googling for your answer before you ask someone else. Asking questions on stackoverflow. Hacking and reading at night to figure out something that may or may not be for work. Recompiling the source code to make it bend to your will.

There is a price to be paid, and it doesn't have to do with money. As was said thousands of years ago, "There is no royal road to Geometry." It is the same today. All these solutions that don't involve paying that price are destined to fail. College age is too late. Middle school may be the last opportunity for any real percentage of women to be affected. It's not about completing classes.

I think that's hogwash. The things you suggested:

> It's about right clicking with your mouse to see if there are any other options available. Googling for your answer before you ask someone else. Asking questions on stackoverflow. Hacking and reading at night to figure out something that may or may not be for work. Recompiling the source code to make it bend to your will.

These are things people can be trained to do.

The real problem here is that software engineering is still in the dark ages. We overvalue people who have been "hacking since middle school" because we can't systematically articulate the skills a good programmer should have and how we can train people in those skills. But this is purely a function of the immaturity of the field. There is nothing magic about software development that requires it to be perpetually stuck in the dark ages of ad-hoc craft.

You can't train someone to be good at something they fundamentally don't want to do, and you can't train them to be passionate. I've been hacking since middle school, and I'm a terrible programmer, even after a couple dozen programming classes ranging from Visual Basic to C++. Anyone would be able to see this within five minutes of talking to me. Fortunately, programming isn't what I do for a living (at least not today). I'm much more interested in breaking things than building things.

The point is, if you want to understand how computers work, you'll put in the time and effort and experimentation that's required to learn that. And you'll be better for the experience. If you aren't comfortable experimenting with your computer, you won't have the experience of doing something really cool with it. You won't have the experience needed to understand at an innate level what is going on. I couldn't explain half of what I know about computers. It's just simple familiarity. It can't easily be taught except through trial and error and experience.

Give a man a fish, teach a man to fish, or give a man a fishing pole and let the man figure out how to fish on his own. Which one embodies the hacker spirit more?

I agree 100% that you can't teach passion - that has to come from within. However, we certainly can do awfully better at encouraging that passion when it emerges rather than dutifully stuffing it back in the bottle and turning a promising college career right back into the pursuit of an MRS(tm) degree.

And if that takes a degree of well-meaning but mostly awkward affirmative action until the majority of brogrammers have either retired or drank/snorted/driven themselves to destruction, that's just the price of changing the status quo IMO.

I used to think otherwise, and all of these heavy-handed efforts to pull women into Computer Science once made me roll my eyes at their ineffectiveness, but all I had to do was listen to how a lot of my younger co-workers still judge women. Same $h!+ new decade alas...

You're making a basic logical mistake, which is that the fact that someone didn't hack in middle school means that they wouldn't enjoy being a programmer. That's a logical hole that let's a lot of sexism through. It enshrines a status quo where we consider it acceptable to have dramatically fewer women in the profession simply because of gendered socialization at earlier stages of education.

I never stated that, or even implied it. I'm not the original person you were talking to though. I actually stated that just because you do something in middle school doesn't mean you'll be good at it later on (I'm not a good programmer, remember?).

I'm just saying you're missing the point. The point is, passion can't be taught, only encouraged. People hacking since middle school are probably passionate about it. That doesn't make them better than someone who became passionate in college, minus the decade head-start. It's not about teaching people to become programmers, it's about getting people interested in experimenting and learning on their own.

We overvalue people who have been "hacking since middle school" because we can't systematically articulate the skills a good programmer should have and how we can train people in those skills.

You did a great job of explaining the unease I'm starting to feel around the whole "hacking since middle school" ideal. Besides the gender and class bias that result, it's just lazy thinking.

I say this, of course, as a straight white male who has been hacking since elementary school.

> We overvalue people who have been "hacking since middle school"

Nothing specifically Dark Ages about this. I think that's true with many fields. It's the theme of the child prodigy. A painter, a violinist, an athelete would all be given special notice if the excelled at these skills at a young age. One would expect an excellent engineer to grow up playing with legos or building toy bridges.

Perhaps this is a practice that doesn't do us any good, and should be done away with, but it is not at all limited to programming.

As to engineers, while people might expect that correlation to exist, no company actually cares whether a candidate played with legos as a kid. It would be absurd to even bring it up in an interview. Engineering companies expect that any skills you need to do the job will either have been taught in school, or will be taught to you by the company.

What's "dark ages" is grouping programmers together with painters or violinists or athletes rather than other engineers. It's the result of not having a systematic theory of the skills necessary for the profession, and not having a way of taking people with the requisite mathematical aptitude and instilling those skills in a formal way.

Do middle-schoolers who will be doctors take apart animals in their spare time? Do the future-lawyers sue people?

Violinists, painters (presumably the artist type, not the outside-of-my-house type), and athletes are in career paths vastly different from engineers, doctors, or lawyers. There are very few people who actually support themselves with careers in the former.

Even though I have been coding since 2nd grade, I'd hate for that to become some kind of career filter to keep people out.

The career path is different, but mindset is very similar. As far as mindset goes, I find I have a lot more in common with the artist or musician than with the doctor or lawyer.

And much like the artist or actor, but probably unlike the lawyer or doctor, I'd probably pursue my technical passions (at least as a hobby) even if there was very little money in it.

As someone who grew up thinking he wanted to be an attorney (until I realized how much paper work there was), I grew up participating in debate club and mock court. My sister grew up wanting to be a doctor and would research various illnesses and learn all about them; she is now in school for medicine. This sort of thing might not happen in every field, but the two you pointed to are probably not the best examples, since both of these behaviors are pretty common for kids interested in law and medicine.

A "practice"? It's called reality.

There's a reason why most top professional athletes started their sports at a very young age. It isn't coincidence or societal convention.

That analogy doesn't fly. Top professional athletes are truly exceptional in a way programmers are not. They're the 0.01%. They're a good analogy to maybe Fields Medalists, not people working at Facebook.

I mentioned "top athletes", since they're a very visible part of the curve where we can demonstrate that longevity in a field from an early age increases likelihood of success.

I would expect that all levels of athletic performers would benefit likewise - although they won't end up in the top 0.01%.

We like to flatter ourselves and think that programming isn't a skill like athletic performance, but that's just because it makes us feel better since it's harder to measure.

It's not so much about skills (there are no skills that a 25yo hacking since 10 would have that a 55yo hacking since 40 wouldn't), but about selection: people who were more interested in computers than anything else at 10 -- and were spending all their time doing that, for absolutely no monetary or other reward, are going to be far better at doing programming (again, at any age) than someone who isn't that interested. And when there is an opportunity for such an interest to show, it would show early. Making sure that opportunity is there for most people would go a long way towards equalizing the gender ratios.

I can get a job at a giant company where I'll write 2000 lines of code in a year.

Wait, what? 2000 lines of code per year? Having never worked in a "giant" company, is that typical? I do that in about 10 days at a normal pace, and can do it in five if I'm really pushing it to meet a deadline or whatever.

Anyway, I thought 4K LOC/month was pretty standard for programmers of all stripes, or ~50K LOC/year. Is it really just 2K?

/shocked

The Mythical Man Month gives 10 lines per day: http://www.cs.usfca.edu/~parrt/course/601/lectures/man.month....

Some of the best coders I've seen probably have negative LOC/month counts.

I am a huge fan of deleting code. And I've landed commits that killed thousands of lines in one go, so I'm pretty sure I'm in the negative.

Yes, 4k LOC/months is standard, regardless if you write assembly, haskell or C. whether you work through a mill of bureaucracy or you are in charge of your own project. These 2k LOC/year coders must surely be unproductive.

I'm sorry, but that's just a false blaming statement and completely counterproductive to the conversation. There is absolutely zero evidence correlating LOC and productivity.

About the only time you could link LOC to productivity is in the very early stages of a project - where some type of visible progress is needed to plan for a possible feature set for the release.

Sarcastic statements are considered to be negative blaming statements. They don't help the conversation.

I think they were being sarcastic.

These "LOC/month" comparisons are pretty useless, because it depends on the programming language and the project. One can churn out loads of code in languages like Java and C#, whereas dynamically typed concise languages usually need less lines of code.

It depends on what you are doing, pure maintenance work is very very low LOC.

Wait, are you saying that college is too late to learn how to program?

I didn't begin programming until college (I'm a white male in the US), and I've gotten paid to do it ever since.

That's me on the AUTHORS list for Google Chrome's v8 engine [0] (I ported it, and Node.js, to Solaris if your curious), so I'm at least competent enough to do that.

Anyway, that's proof that it can be done, and at a level that make a difference somewhere. :)

That said, I don't like to code, so maybe that has something to do with why I started late. I was into computers from pre-school onwards, though.

[0] https://code.google.com/p/v8/source/browse/branches/bleeding...

I didn't begin programming until after college, actually. But I had a math degree by then (theoretical math, not even applied). I had developed that curiosity and drive, however, and I think college is too late for that.

edit - I mean for large numbers. The exceptions are always wonderful to see.

another edit - I also don't think that most programming is about programming, it's about figuring out how things work. Just learning to program doesn't help with programming very much. "Figuring out how things work" needs a name, and that's what needs to be taught.

Actually you are a good confirmation of the bias. You didn't start until college, and you don't like to code.

I interpreted it as if the individual doesn't have an innate curiosity and need for answers, how things work, how to tweak things, an endless desire to learn. If they don't have that "hacker spirit", then it's unlikely they are going to be immensely successful or game-changing.

Anyone can get a job doing coding from learning it at any age.

Is college too late to learn arithmetic? I would equate the two; in fact, I have wondered for a while if we could actually replace arithmetic lessons in elementary school with programming lessons (and have students learn arithmetic by implementing it in code). You still have a malleable brain in college, but picking up an entirely new way to think becomes harder as you age.

Gee, I don't know. Grace Hopper learned to program quite late in life and she did ok. Maybe it's not WHEN people start to code, it's HOW they start to code and whether they learn while doing it? Big-time athletes don't appear because they were kings of the middle-school playground for 10 years, they participate in an increasingly intensive program that winnows out pretty much everybody who participates.

I don't think that we see a huge number of failed programmers ending up as insurance salesmen or financial advisors because they blew out a knee.

I think the argument is that statistically speaking, it's too late to produce passion in large quantities. It's not just about one person discovering code, it's about getting enough to close the gap.

I'm not sure I agree... many people discover things in college that they've never heard of and subsequently major in. But it's at least debatable.

Reply on news.ycombinator.com