Ask HN: I love coding but hate software engineering. What's next? | Hacker News Reader

Ask HN: I love coding but hate software engineering. What's next?

7 pointspacman_6 comments

I like computers. I love coding. But I hate working as a software engineer. I dislike having to deal with big codebases, legacy undocumented code, pricky code reviewers, slow bloated IDEs/frameworks and things like that. I've tried (for years) but now I must admit I'm done.

I want to shift my career into some area where my coding skills would be appreciated as a tool to enhance my work, but not as the goal of my everyday work.

What are my options? I've heard Economists and Biologists are using coding to crunch data, but that's it. Is there any other option?

View on HN

Over the last 40 years coding has become a toxic. The fact that the number of women in the field continues to fall tells you everything. And why you are unhappy.

I would suggest not working at places where there is a critical mass of CS types. Find an engineering company where the product isn't software or worse adware.

It sounds like most of your issues could be resolved by working for startups / small companies that have greenfield projects.

The 'pricky code reviewers' could be addressed by having an honest conversation with your colleague(s).

If you absolutely can't persist and have the money to take some risks, you could consider the scientific computing space which in some cases still uses FORTRAN, though I imagine most things are in C now there.

Wouldn't the legacy codebase be a turn off for this individual?

Plus, no shade intended as I do scientific / academic code professionally, but for historical reasons a lot of scientific code has one goal in mind; be as effing fast as possible, no matter the cost. Often times projects like WRF or BLAST achieve this end to an impressive extent, but there is a reason the most readable code is not always the fastest.

The latency / throughput budget of a project can be divvied up in a lot of different ways. In your mobile app it's likely best to avoid optimization for clarity; so you might overspend in the name of maintenance or easy of extension. That said, if this was the 80s and you were attempting to sequence genes, you are probably thinking "clarity be damned we'll just rewrite it later when computers get fast."

That is what a lot of scientific code looks like. It is written just as the idea is possible, probably a little before it is possible. As a result there are esoteric optimizations everywhere and questionable architectural choices as far as the eye can see, all in the name of "making it actually happen."

I mentioned this is another thread the other day, but every "scientific programmer" job ive seen basically asks for a scientist who just happens to be able to program.

How about something in research?

Reply on news.ycombinator.com