Does anyone regret going back for another degree in CS?

[TheGrkIntrprtr]

I have an unrelated degree, BSc Econ, and I love the programming (R, SQL) side of my job as a data analyst. For personal reasons, I'd like to remain working in the public sector, so I can't really transition to a dev role internally without the credential. Data engineer roles don't seem to exist here either. Have to make a decision of whether to resume a second bachelor's in CS in September, or give it up entirely. Can't do part time studies, got an A- in Gregor Kiczales' CPSC 110 course at UBC based on HtDP, which was GREAT, but too stressful while working full time. I'm in my late 20s.

I was spending some time learning Unix on OpenBSD, as well as Vim and Python, and enjoying it. But I can't help feeling I'm wasting my time without the piece of paper, if my goal is to be able to spend my days building software - even if it were maintaining some old crusty old system in Java, I think I'd enjoy it more. And the time I'm spending on that means I'm not spending it on keeping up my skills for my data analyst role, like brushing up on ML (though I'm not really interested in this..) etc.

The problem is, the opportunity cost of going back is hard to stomach. Giving up a decent salary, paying tuition, moving away (3 hrs away or so) from family members after being away from them for so long, it all seems so costly that I'll regret it. However, if I don't go through with it, I think I'd regret that too - that I didn't take the chance to be doing something else for the next 30+ yrs that I'd enjoy more.

So, I could use some advice on a way forward. Is being an excellent dev even feasible if it isn't your day job? And if anyone has gone through something similar, I'd love to hear how it went for you - did you regret going back to school, or not, and why?

For anyone who reads this, thank you for your time.

[codingdave]

> I'd like to remain working in the public sector, so I can't really transition to a dev role internally without the credential.

I've been working for the public sector for a decade now. There are plenty of dev jobs in this sector, if you seek out those of us who sell SaaS software to government entities. Many of the jobs I'm aware of in this sector are remote jobs, almost none of them require a degree. The software needed by public entities is also different than private - we tend to deal with specialized needs and regulatory environments, so you can walk in the door with added value to your employer when you already understand the needs of this sector.

[cactus2093]

It sounds like you are already confident in programming and maybe just have a few gaps in your knowledge. You can fill those with online classes or just as you read and learn things over time. In my opinion, going back to formal school for years would probably be a waste of time when there are so many programming jobs that don’t care about a degree.

Doesn’t the public sector contract out a huge amount of their technology anyway? There are tech companies that do things like help cities optimize bus routes, or sell Saas or data products that governments use (e.g. palantir).

Surely you can find a private sector programming job where you’ll have a much bigger impact on solving public sector problems than working for the public sector. Maybe even start your own company down the road.

And I have to think that since you do have a college degree just not in CS, once you get a few years programming experience under your belt you could even get hired for a public sector programming job, although I don’t have any experience with that and don’t know for sure.

[pestatije]

Have you considered a postdoc or masters degree? It's just one year, plus they're flexible on time schedules.

Also, formal employers like public service usually consider your postdoc subject as equivalent to degree subject. And I have colleagues with chemistry and physics degrees who got the job just because they did a computer science subject.

[TheGrkIntrprtr → pestatije]

The Georgia Tech OMSCS is one option I'm considering which would be more flexible, but I'd definitely need some more courses under my belt before they'd accept me. Unfortunately it'd probably take longer, 3 yrs or so part time I'm guessing.

The kind of 1 year MS in CS masters requiring little or no background that exist in the US aren't really available in Canada as far as I know.

[jphamtastic → TheGrkIntrprtr]

I would take a chance and apply if you have a fair amount of experience. I hadn't taken a class in CS ever before but talked about my experience on the data science team at work. It ended up being good enough for me to get in to OMSCS.

[pid_0]

The income loss and cost of a degree is almost never worth it.

[tubularhells]

No, it was not worth it. Lot of time and effort wasted on things I don't need to know/use 95% of the time. The other 5% I could just ask a co-worker or the internet, and get an answer.

University is overrated. If suggest you start working on you own pet projects, as your enthusiasm is sure to die once you have to program for a living instead of for recreation.

[jbjbjbjb]

Is there any flexibility on the public sector requirement? And is there a possibility that work experience counts as a credential of sorts? Perhaps you could work in the private sector for two years or so and come back to the public sector.

[TheGrkIntrprtr → jbjbjbjb]

I have some health issues that require expensive drugs, so the health benefits are really helpful, even here in Canada. Some close family members are in a precarious position (mental health issues), and I have some anxiety issues myself, so I really value the job security. Then again, everyone has issues, so perhaps I am just especially risk averse. Probably out of place on HN, but it's a great forum for learning.

I think work experience in place of a degree only works if you've been kind of grandfathered in unfortunately, so a related degree is usually a hard requirement.

[taraharris]

For personal reasons, I'd like to remain working in the public sector

Being so agreeable is easier in the short-term, as it allows you to avoid making difficult decisions. The cost is your long-term happiness.

[dfraser992]

I don't know how much this will help...it is a bit free form (need some coffee). After 30 years as a dev/QA/tech lead, I went back to school to get a MSc in data science/ML/AI. I was sick of development work and needed to do something different (the last job was with a crummy startup etc). I thought I'd teach myself ML stuff, but realized a piece of paper as an official credential would probably be necessary to impress HR departments.

And going back to school (aka a vacation) sounded great - I would get exposed to all the basics in a compressed format vs. wasting time fumbling about teaching myself. Sort of like why I went to film school 20 years ago... which is another story again similar to yours.

One track of the MSc program was for people doing a conversion course, so I got out of a few classes thanks to all my previous experience (you might be able to do that). In the process of doing original research for my thesis, I realized at this point in my life I liked doing ML research - I'd been sliding into "research" over the past few years of jobs, so it all made sense. And academia is a different environment than industry, so something new to deal with. So I've been hanging around the uni doing research type work, TA jobs, etc. for the past few years.

So what might you learn about yourself by going back to school - versus getting even more stuck in the rut/groove you are in? I was 30 years old only a decade ago (I am actually 50...) Time has a way of speeding up the older you get. Do you want to be doing the same thing a decade from now? Or be in management?

My advice is, based on what you said, and whatever unconscious biases I might have, is to go back to school. You are certainly young enough to take a detour for a few years (like I did with film school) and having a CS degree along with your data analyst experience will make you pretty valuable. It sounds like you really like programming, so I wouldn't try to suppress that (like I did with writing screenplays - IT is much more time/cost-effective and profitable unfortunately and it is so very similar to writing actually).

Weigh the personal costs against the probable long term benefits of learning something you really like and which is more valuable career wise than sticking to data analysis - are they really that costly? Fundamentally, programming is a craft. There are plonkers who do it only for the money, but they are not ... I get the impression you want to be a craftsman, so you should be willing to suffer for your art :) It is the only thing that will keep you motivated when you are on a solo death march.

And if you really do need credentials to get a dev job in the public sector, then you need the degree. Perhaps it is as simple as that.

[TheGrkIntrprtr → dfraser992]

Thanks for this great reply. Yes, I think you're right - at least with the constraints I've set for myself, if I want to build software as a full time job in the public sector it seems like there's no way around getting the credential.

However, I'm guessing I might need to start at the bottom again. Probably wouldn't be able to just waltz into a full dev position with my data analyst experience I'm guessing. That'd mean accepting a huge decrease in salary.

Maybe learning enough outside of work to begin contributing to open source projects that interest me would be enough to scratch the itch.. or perhaps some freelancing/building software to sell on the side? Is it feasible to really get good at it if isn't my full time job though, I'm not sure..

One thing I was thinking of doing, if I decided against the degree, was really digging into some open source data engineering projects like dbt and airflow. Even just reading the source code and getting a good understanding of it if I can't contribute. My thinking was that even if no positions labelled as "data engineer" exist, it might do something to help me maneuver into a data analyst/science labelled position where modern data engineering is really what's needed.