Hey all you happy systems programmers, I have always been interested. Can you enlighten me? I currently do fullstack web / data monkey stuff

What aspect of systems programming would you like to be enlightened about? I do a lot of consumer electronics programming, which I think falls into that category.

I grew up learning programming on a Commodore 64, where getting anything done meant twiddling bits and writing to registers of various hardware devices. Writing device drivers today can provide the same sort of thrill, since the work is quite similar. A lot of my time in high school and college in the 90's was spent learning all about UNIX and Linux, so it feels very natural today when I develop user-space tools that bring value to the user. Outside of nostalgia, there's plenty of bleeding-edge work to keep things interesting -- audio/video processing, extending platforms such as Android, security, etc.

Sounds exciting. I want to know how to make a switch, if one can. I am in the same position (mostly webdev) as the gp. I do understand that "systems programming" is very broad.

I get the impression that most people I've worked with in the CE world come from non-CE programming backgrounds, so I doubt there are major obstacles. The biggest help would be familiarity with C/C++ programming, and being comfortable with the practices needed to use these languages properly (e.g. explicit memory management in C, RAII idioms in C++). (Next-generation systems languages like Rust can't come fast enough, in my opinion.)

You can stop responding now - apparently the surveymonkey has a limit of 100 responses on their free plan, and upgrade would cost me hundreds of dollars. Anyone with a paid account is willing to redo this survey later? (we're already at 205 responses after less than 12 hours, of which only a 100 is shown).

For much nicer formatted results, see here:

https://www.surveymonkey.com/results/SM-VT2RQJNR/

(Sorry, the original post is too old and I cannot edit it anymore).

The results are confusing. I can't tell which bar is which number, 1-5.

Indeed, thanks for pointing out. I've reconfigured the results page, I think the visibility is pretty good now. You need a new link to see it though:

https://www.surveymonkey.com/results/SM-VT2RQJNR/

Yes, very confusing. The key is at the bottom, but it would help if each bar was labeled.

The bars are ordered 5-1 from top to bottom.

I was a little unsure of which category to choose for my work as a developer of consumer electronics software. "Embedded" means different things to different people -- some people reserve the term for microcontroller-level devices. "Systems programming" seems closer, but the "/ Desktop apps" bit had me running back to the embedded category.

For what it's worth, I think it's a great part of the programming world to be in. :)

I picked multiple, based on what I have to do. I write NLP software, so I ticked front-end (make GUIs in Angular), back-end (make REST/WebSocket APIs for gluing our software together), AI/ML (that is the core of most advanced NLP software), and systems/desktop (since we need to package and deploy our software in self-contained packages).

I hope I didn't screw up the metrics. :/

I did the same as you. For instance, when I do webdev I'm mostly unhappy (web frontend sucks big time for me!), but when doing desktop or other, I change my "happiness" numbers.. Don't know if it screws the metrics as well..

I see numerous problems.

Shouldn't there be 2 separate questions? "What field are you in?" and "Disregarding salary, how happy are you with your field?" Would prevent confusion.

How did you come up with this list of fields? Where does academia fit in? Why are systems programming and desktop apps in the same category?

If I made it into two separate questions, I doubt I'd be able to make a meaningul report out of it with survey monkey. Plus, I don't think it's that confusing really - if you give your answer for a given field, it means that you're working in it.

There is no academia (as in academic career track) - I am only asking about programmers, not researchers. If you work in academia as a programmer (say implement stuff at CERN), you can answer according to the nature of your job - ex. if you help scientists write simulations, you select "scientific/numeric", if you mainly work on massive data pipelines, you select "big data" etc.

As for coming up with a list - I compiled it myself, trying to bundle together work that seems close enough in nature. For me, desktop programming looks similar to systems programming - i.e. it's mostly C++/C/other compiled languages, working against OS APIs, no garbage collection, expectation of reasonably low latency.

I'd be interested in a satisfaction by industry - or better, a combination of industry and field.