Hacker News Reader

Focusing on frontend is pretty typical of code academies. My guess is because it's easy to inspire someone when they build an interface to some they'd actually use. Unless you're already a developer, backend systems are further from things people actually do. It's also an easier sell because it's easier to explain to people what they'll do--you just show them.

Because then it will be regarded a "softer" engineering domain and the market will implicitly price the alleged lack of rigor accordingly. Just take a look at the gender ratio for "human computer interaction (HCI)" students in CS academia, aka UI/UX designers. It is important that we don't funnel aspiring female engineers into a less rigorous domain just because it is more popular with their gender. The education system should not reinforce stereotypes at such an early stage. If they still want dedicate their career to fiddling with CSS variables after 4 years of education, great. They can do that after mastering foundational CS concepts so the system doesn't accidentally segregate and create a caste of second-class workers.

You’re really just saying that a CS degree should be used to gatekeep frontend work.

Sorry if my comment came off that way, I meant the opposite. I am a strong believer in making software engineering more accessible and as anti-accreditation as you can reasonably get. I even believe that the current medical accreditation system ought to be changed to be more in line like law where medical schools admissions are not allowed to solely control the production of doctors.

What I meant regarding aspiring female engineers is that those going into the traditional CS route should not have their career choices be subtly influenced by existing industry trends and the tertiary education system should do everything it can to prevent that.

That sounds good, but I think their point was that it works in the other direction too and you're ignoring it just because it's status quo. Right now, instead of becoming an expert in HCI, everyone is funneled into classes about database indices because some department chair thinks that "real" computer science is bit-twiddling and cache invalidation. You must specialize in the C before you can learn about the H in HCI.

Why not the other way around? I think by far most of the problems that make development difficult are human problems. Why shouldn't we push all the CS students to study psych and writing first?

Firstly, I don’t agree with the notion that frontend is easier or that women shouldn’t pursue frontend. But I will say that it’s a repeated trend that when women enter an industry, pay for that industry in particular goes down (this happened to teachers for example).

> But I will say that it’s a repeated trend that when women enter an industry, pay for that industry in particular goes down (this happened to teachers for example).

Could this just simply be because women are increasing the labor supply, and thus driving down the cost of labor in that field?

Whilst that may be true, it's common in many fields for women to get paid less than men doing the same jobs. This has become such an issue that in some places organisations and now required to report on their gender pay disparity, e.g. https://www.theguardian.com/world/2021/dec/29/gender-pay-gap...

Whenever I've looked closely at such statistics, it's pretty much never been the case that men and women are doing exactly the same job.

That is usually a symptom though, e.g. paying men more on average because they are in more senior positions on average.

> I don’t agree with the notion that frontend is easier or that women shouldn’t pursue frontend.

If the comparison is between teaching "CSS, HTML, and Javascript" and "writing CRUD apps in Python," sure, they're about equal.

I think people get their wires crossed a bit though with backend tasks where you might need to debug a race condition, or like, be a Linux expert. Since companies typically pay for backend resources, they're also more likely to assign engineers optimization tasks to reduce costs. These are basically Senior Backend SWE tasks, but people don't stop to consider what a comparable Senior Frontend task might be.

And thinking about it, they're the kind of thing you'd also need to be a Linux expert at. Like, half the problems YSlow flagged on my projects were things Sysadmins needed to handle, like slow DNS, or ETags. So they typically got foisted on me, rather than the frontend engineer.

> it’s a repeated trend that when women enter an industry, pay for that industry in particular goes down

Honestly, I have no idea how one disentangles supply and demand effects, from cultural and regional effects. One random HBR article suggests that comparing locations by woman partipicaton rate reverses the trend, to where more women _increases_ local median wages: https://hbr.org/2018/01/when-more-women-join-the-workforce-w...

I've done both at a decently high level (leading teams to do both, if that means anything to you), and I can say for certain that there is an entire order of magnitude fewer problems working on the frontend than working on the backend.

I can also say hiring for a frontend dev does not mean they can do backend work, whereas a backend dev (in my experience) is able to jump into the frontend with relatively little friction when needed.

YMMV of course, but this has been what I've see in in my ~20 years.

I have the opposite experience. My experience with junior engineers out of the modern CS education is that the current leetcode based system is impressive at churning out a backend engineers who can write functional boilerplate code and algorithms. As long as you have a senior engineer supervising the overall architecture and doing code reviews for security issues, fresh grads are basically easy to replace cogs in the backend factory. Frontend on the other hand requires much more experienced candidates to deal with resource constrained systems, especially when it comes to animation heavy and fat client/WebAssembly apps where every frame counts.

Leetcode algorithms for backend? I don’t think so, most backend devs should be experts in leveraging their team/project’s chosen DB infra. Which mostly comes down to designing a sane data model and selecting appropriate indexes. And in my experience it’s very hard to find people who can even do that well.

Do you work in HFT? I consider that basically a subspecialty of systems programming.

> Do you work in HFT? I consider that basically a subspecialty of systems programming.

Did you peek at my comment history? How did you figure that out? Not HFT, but trading systems yes.

It is hard to think of another industry where someone's impression of daily-grind junior backend work is "leetcode" instead of "Why does this stupid json parser choke on floats that don't have any digits after the decimal!? Let's try a different one."

Resource constraints? That is not something 90+/-9% of frontend developers need to care about. Backend devs will hit resources constraints a lot earlier than frontend devs will, anyway.

No? Server and cloud resources have almost zero marginal cost at most medium sized and above companies unless you are doing something truly intensive like video encoding. For most web apps, you can go a long way by just scaling horizontally and throwing more ram at the problem.

On the hand, you don't have any control at all for the frontend, especially for consumer products.

Ah, spoken like someone who hasn’t actually tried to solve a backend problem by throwing more hardware at it. :) It does not work, and is very expensive, jumping from tier to tier. Your memory leak will just keep leaking, your inefficient query will just keep gobbling up RAM, your unoptimized traversal will still take too long, until you find the actual issue and fix it.

Modern web browsers have abstracted all of the resource management problems away from the app developer. The worst you can do is crash the browser, which is a pretty easy problem to avoid.

Is this because women entered the industry or because more workers entered? Or some other reason?

Front-end development is not "easy" or "trivial". This misguided perception is exactly what leads to hacked-together websites full of div-tag soup, zero accessibility, and forcing their users to run huge amounts of pointless JavaScript which ruins their devices' battery life. Please everyone, stop doing that and learn to build professional "front ends" to your webcrap.

It would help if the getting started documentation on the major entry points didn't advise the opposite. How is a beginner supposed to know? Most of the people they'll encounter along the way think it's completely reasonable to pull in a gigabyte of unaudited dependency upon unaudited dependency for a new app before they've done any kind of design work to determine what they might need.

> What's the difference between "front-end" and "frontend"?

It's just a style/proofreading issue. You knew that.

Is it? The author says one is easier than the other, that feels more substantive. I do assume there's some kind of proofreading issue, but it's ambiguous and I don't know how to resolve it, hence my mention.

> Front-end development is now seen as "easier" then backend or frontend.

This makes absolutely no sense if "Front-end" is the same as "frontend" and you knew that.