Hacker News Reader

Comment by jmfldn | Hacker News Reader

This comes up regularly here and I always feel that I need to point out the elephant in the room each time. From a previous post...

"The fact that most engineers are young is much more to do with the numbers in the profession doubling every five years in the last twenty years imho. If you were an engineer twenty years ago, there are now sixteen times more engineers. Your peer group is going to look pretty small therefore. All of these younger engineers are going to get older too of course, and the demand won't keep on growing at this rate, so inevitably the ratio of young vs old devs will even out."

https://news.ycombinator.com/item?id=28294734

I don't deny that there are other factors but this simple numerical fact accounts for a lot. The vast numbers of 20 and 30 something developers around now are not all going to become consultants, managers, founders.

View on HN

This is a good point and is one possible explanation for the apparently lack of movement in the SO poll 10 years on

The SO survey is an absolutely awful representation of the community, because it mainly represents the young inexperienced engineer and students who can't do much without googling.

The longer you are in the industry, the less time you spend googling stuff. I'd say I probably end up on SO once every 2 months. There's probably huge swathes of devs who didn't even know there was a survey or never visit SO because they can generally get on with their job

> The longer you are in the industry, the less time you spend googling stuff.

Only if you stay in your comfort zone. Memorising all the symbols in Framework X isn't going to help when you have to use Framework Y.

Mostly disagree - With basic background knowledge of the overall ecosystem, problem domain & tooling, figuring out a new framework or API is as simple as newing-up each type and exploring the public interface via something like intellisense.

I can learn some dependency faster by throwing it into a blank unit test project than I can by reading over the (likely outdated) documentation.

Intellisense will find you a solution but it won't find you the solution.

I have learned to Google things even when I think I know the answer and I've done it that way many times before, because often there's a better way that I've been missing out on all this time.

> Intellisense will find you a solution but it won't find you the solution.

Yes, and what is the penalty for finding a bad solution and then having to try another one?

For me, the cost is only my time. My customers don't see any of these fuckups because they only live on my computer until I get things sorted.

> I can learn some dependency faster by throwing it into a blank unit test project than I can by reading over the (likely outdated) documentation.

Does your greater experience also come with the emotional intelligence that allows you to see how arrogant that sounds?

I'm definitely not some wizened master but I'd also agree with the statement. Good updated fully reflective of use documentation is not the majority of libraries and even the ones that do manage to do a good job often miss some critical change till someone raises an issue or a PR.

> Does your greater experience also come with the emotional intelligence that allows you to see how arrogant that sounds?

How something sounds is largely a function of the person listening to the thing.

In fact, the primary determinant of how X sounds to person A is the nature of X; otherwise we would be unable to use data to make inferences about the nature of our world.

A few people have already disagreed, but I want to add that in some areas the more you specialize, the more unique your work. "Memorizing new framework" is fully within my comfort zone, it's just API surface.

Finding useful information to actually build those frameworks is much tougher, and only gets tougher the deeper you go. StackOverflow is very much suited for newcomers and engineers working at the top of the stack. My work became harder and thus less googleable (which bums me out, honestly).

I disagree. Even if you make a complete left turn and only work on something tangentially related, you still likely have a wealth of knowledge that you can build on. I doubt you're going to make many moves in your career where you 100% start from scratch

Disagree. If you’re not using a Framework X or Y, suddenly SO becomes a lot less interesting. This doesn’t necessarily mean that you’re staying in you comfort zone.

As an aside, totally subjective, but Google’s results have been degrading for a while - it’s not that you won’t find what you’re looking for, but it’ll often take a bit longer. Better resources exist.

I super rarely use SO and even googling is relatively limited. Am in my first year out of University as a Software Engineer in a reputable conpany and get nothing but praise in my performance reviews. I have no idea what these memes about copy&pasting code from SO mean? Like, are they serious?

For my day-to-day job, I've visited SO under a dozen times in the last few years. Too much of what I deal with is some proprietary vendor SDK or chip where the data sheet is only released under NDA for SO to be of much use.

I'm an embedded C developer. Right now, there are 366,183 questions tagged "c", 9,312 questions tagged "embedded", and 2,282,444 questions tagged "javascript". I don't use JS. Not that many embedded C developers are using SO. So I tend not to bother with SO.

I'm an embedded dev too and I just barely use SO. Too often the questions are either out of date, not relevant or questions get flagged as duplicates because a tangentially related question has been answered for a different architecture and compiler

I think you’re probably right. I’d bet that most of the survey respondents are relatively new, or frequent and established contributors. The data are likely skewed towards these individuals and not representative.

I may end up on the site if I’m googling something, but I can count on one hand the number of times I’ve asked a question, posted a reply, or voted on an answer in the past 5-6 years.

SO is still a valuable resource, but, and I realize this is selfish, community participation stopped being interesting or useful a long time ago.

Also older devs might engage less with SO community and platform. Get what they need and move on. Where as younger one might think there is something to gain for being more active there.

The other one being who is being asked; do 30+ year olds still frequent SO and participate in the surveys?

And the other one is, do the SO polls mention how many people replied?

Exactly. Especially considering that the plot shows only percentages.

For the author of the blog post it would be better to draw conclusions about 'developers disappearance' from a plot showing absolute numbers.

I don't know if I buy that argument. The number of new developers isn't much higher today than it was already in the 90s. I know plenty of former developers in a wide variety of jobs, and it looks like most people simply drop out in the first 10-15 years. Being a developer is not a good long-term career option, unless you really like the job. It demands much, but the pay isn't competitive, except for a small minority.

" The number of new developers isn't much higher today than it was already in the 90s"

All the data suggests this is wrong by any order of magnitude. The web was an absolute game changer.

> The web was an absolute game changer.

In the fact that it made rank-and-file CRUD harder. The web's state-less-ness and screwy anti-wysiwyg UI "auto-flow" made CRUD into a convoluted whack-a-mole mess even its mother hates. It takes about 3x as long and 3x more code to develop a CRUD app in web now. We have to focus on tech & UI minutia instead of the domain. We de-evolved and nobody seems to care. The Jetsons were shot in head point blank, replaced with the Flinstone DOM.

The web may be is a godsend for e-commerce and social networks, but it sucks rotting eggs for office CRUD: the wrong tool for the job.

The web was an absolute game changer in the 90s, and it attracted many new people into the field. A lot of those people left tech due to the poor job market in the 2000s.

"Being a developer is not a good long-term career option, unless you really like the job. It demands much, but the pay isn't competitive, except for a small minority."

1. Being X is not a good long-term option, unless you like the job. 2. Not sure where you base you viewpoint, but I have a hard time finding a more competitive and global job market that favours candidates. 3. All high paying jobs demand much. Otherwise they wouldn't be high-paying.

My impression is that the top 20% of the job market pays competitively, while the bottom 80% doesn't.

The average developer writes and maintains custom software for large organizations. While that kind of secretarial work is useful and necessary, it's not particularly valuable and doesn't pay that well.

Not particularly valuable.....until your system goes down and you cant access the data or pay your employees or activate your machinery, or miss vital deadlines, etc etc.

A job may be vital without being particularly valuable. Just ask any of the "essential workers" who had to continue working on-site last year. The value of your job depends more on the costs of replacing you than on what would happen if nobody does the job.

I dont understand you definition of valuable? Do you mean in terms of revenue generated? Is that what you mean by valuable?

I mean economic value – what people are willing to pay for something.

For example, drinking water is incredibly important. Without it, we would die in a matter of days. However, you can't create much value by producing clean drinking water (branding it is another story), because people are not willing to pay that much for it under normal circumstances. As long as people can reasonably assume that someone will produce drinking water anyway, its value remains low.

In the same way, many jobs are essential but don't produce much value under normal circumstances. There is usually another person or another company willing to provide the same good or service.

> My impression is that the top 20% of the job market pays competitively, while the bottom 80% doesn't.

Truth is, it's the top 20% that creates the most value.

"Not competitive" for a software developer is still a pretty high paying job by most peoples' standards.

> but the pay isn't competitive, except for a small minority.

What pay is though? I make significantly less than numbers discussed in tech-centric places, but I still make more than I could doing anything else given my age/education.

Does a software job at some government contractor that pays 90k really demand much more than a civil engineering job at some government contractor that pays 80k?

Reply on news.ycombinator.com