>And that doesn't even begin to cover graphics and computation, compilers, algorithms, operating systems, data structures, artificial intelligence, and on and on.

Maybe if you are attending top3 country schools, but for the rest all those topic you can get good or better knowledge on the internet

What's your point? You can make that argument for almost any degree. The discussion was whether or not OP would LEARN anything from a CS degree since they've already been employed as a software dev. The answer is a resounding yes.

> Maybe if you are attending top3 country schools

This is bog standard stuff for any school, not "top 3". Working in software doesn't mean you know jack shit about anything. I also notice calc 2 or 3 isn't included in the list.

My BSc degree taught lots of things that would be useful for an average software developer with some experience. MSc programs tend to be more research-oriented and thus less practical unless you work in some R&D niche that matches what you studied 1-to-1.

Practical BSc materials included other programming paradigms, mostly pretty advanced declarative and functional programming, equivalent to big chunks from CTM and a typical Haskell book. Also deep relational algebra and calculi, including query optimization, and low-level driver programming, concurrency models, and networking. Further to that, compiler construction, amounting to significant chunks of the Dragon Book, etc.

This is something you can self-study. But, obviously, an institution will provide good structure and credentials. Depending on the cost, it might make more or less sense. The OP might want to look at good EU universities where tuition is free for EU nationals and coursework is taught in English at BSc level.

    > significant chunks of the Dragon Book

You did this as an undergrad? I am surprised. That is normally graduate level.

Yes, my undergrad school was dominated by some compiler researchers. We didn't cover it all, by any means, but we built lots of lexers, parsers, and interpreters. There was an elective that went through the rest of the book. Crafting Interpreters would have been a better textbook, but it wasn't available at the time.

I suppose it isn't entirely paradoxical, but I disliked the dragon book, even though I rather liked Hopcroft/Motwani/Ullman's (theoretical, non-compiler book) Automata Theory, Languages, and Computation. (Turing award winner) Ullman giveth and taketh away.

Purple dragon (Aho/Lam/Sethi/Ullman) is probably better than red (or green) dragon though, as it adds some interesting topics.

In an undergrad course, I'd probably prefer a simple book that uses recursive descent.

> I can guarantee there's a huge amount of knowledge in a proper CS degree that they don't have.

Idk, I don't buy it. I have a master in Chemistry, but ended up a software engineer 8+ years ago with no formal CS education.

Way after starting to work, I started digging in most CS topics and I'll just plainly say that: I know more about networking, OSs, math, programming languages, type systems, algorithms and ds, system design etc than fresh graduates I interview, let alone people that graduated ages ago.

Unless you've been living under a rock, most universities out there programs are public, lecture and notes are there, and there's an overabundance of excellent full courses on YouTube and simila.

The only thing that a degree tells me is that you went through a series of tests and passed them, tells me absolutely zero, nothing, about what you retained in years studying.

It's not surprising, in chemistry it was the same. There were many people that would ace all exams but literally forget stuff or not really understand it just few weeks after an exam.

Firstly, congratulations (seriously) on your efforts. Firstly for getting a degree in chemistry (not easy) and secondly for a rigorous approach to self-improvement.

I would however suggest that you are not a prototype for self-taught programmers, much less self-taught programmers with no other degree (and thus no exposure to academic rigor.)

Of course the OP might be as advanced as you are, but its fair to suggest they likely aren't.

Personally I learned to program as a child, and then did a comp sci degree. And I got a lot of value from that which I've used through my career.

Of course (these days) there are lots of resources online so the OP could learn everything he needs "self taught", but degree programs formalize this, and give you a degree at the end of it. (And formal degrees are still a valuable hiring filter.)

    > I would however suggest that you are not a prototype for self-taught programmers

Twenty-five years ago this was bog standard. Most people who were programmers did not come from a computer science background. They were self-taught. What is so different today that they cannot do the same? If anything, the biggest difference is the "gates are higher", but the work is barely different.

I work in a field where most of my customers are self-taught, often (but not always) degrees in other fields. I can only use my experience as a reference, ymmv.

What ive noticed is that people have different strategies to learning. Most adopt a "just in time" approach. They learn only enough to complete the task at hand. There is very little curiosity in the bigger picture, or the fundamentals.

This is the opposite to the formal approach, which grounds the learning in theory.

can people learn the fundamentals in a self-taught way? Obviously yes. Do most people learn this way? Id suggest not.

The contrast is to students at college who are always asking "hey am I learning this? When will I use this?" It's not always obvious where they become important.

In my own career I was able to leverage my fundamentals training and translate that into value to those who just want to complete a task.

25 years ago computers were mostly interesting to "geeks", and those looking to program them were even more so (ok, that's really 30-35).

Yes, it started to shift 25 years ago already, but what has obviously changed significantly is that software development became a good paying job, which attracted even wider masses not caring about the fundamentals. There are certainly more self-taught programmers who dive into those fundamentals in absolute numbers, but relatively they are more of a minority. But that means that there are even more who do not care in absolute numbers.

You have a point I did not consider. Having studied advanced chemistry, math and physics as a chem master obviously taught me a lot how to approach CS.

There's that, but more than just how to approach a technical topic you put in the effort to lean to such a depth on your own. Most people will not bother (if they even know how in the first place).

Also it sounds to me as though you are fully capable of independent research. Even among those who hold masters degrees that's not guaranteed. A significant part of PhD programs is ensuring that students are capable of that.

I think you are more of an outlier than you realize.

I hear this refrain quite a bit: "But what have YOU retained?" and it almost comes across as weirdly accusatory. I can't speak to everyone's experience, but since we're tossing out anecdotal evidence, I've also acted in the role of interviewing applicants for software engineering positions (granted this was over a decade ago). Occasionally we'd discuss CS topics just to get a feel for the applicant and most of the applicants with BS/MS in computer science knew their stuff.

I mean if you aren't retaining a majority of what you spent 4 damn years studying then what was the point? I don't claim to have an eidetic memory but I recall the theory/application of most of what I studied in university. A few months ago, I had to whip out the Taylor series for a particular problem and I pretty much hadn't touched the concept since taking Calc 2.

It also seems like you along with a number of people seem to be addressing a point that I never once made.

Once again I suppose I need to restate the original argument because basic logic apparently missed the cutoff.

The argument was not:

- something something the vast wealth of knowledge on the internet replaces the need for your stuffy ivory tower something something

- something something my horatio algers narrative elevates me beyond the filthy CS grad mugbloods something something

- something something but what did YOU RETAIN Leonard Shelby something something

The argument was that the OP would not LEARN ANYTHING NEW in the course of acquiring a formal degree since they've already started working professionally. If the OP got a job in frontend development (for example), it's highly unlikely that they've ever encountered any of the aforementioned subject material I mentioned.

That you N=1 spent your formative years poring over scrolls of the SICP and The Art of Computer Programming like you were preparing for your bar mitvah is not generalizable to the vast majority of people working in IT.

> the OP would not LEARN ANYTHING NEW

Do you perhaps mean "useful to them"? One of the most common complaints is that CS degrees include "useless" stuff. Based on that observation alone it's difficult to believe that OP will have encountered all the relevant material in the workplace.

But then OP said ME/EE. Surely you don't mean to suggest that working as a software developer will have covered that material to any significant degree?

Edit: Nevermind, I see we agree and I misunderstood you. I'll leave the comment though because the point stands.

I think of working problem sets as training in the sense of gym training rather than job training.

Gym training is also good for your brain of course.

You know more than fresh graduates, but fresh graduates don’t absorb 100% of the material being taught — more like 50% at best, and many promptly forget that knowledge after taking the final. As someone who went back to school after about five years in SWE industry, I did (and still am) learning things in CS classes.

Granted I did know a lot of the material already, and I do go to Berkeley which has a comprehensive CS program, so YMMV. My main point is to not compare your knowledge to fresh grads knowledge to determine what they are taught.

    but OP will 100% learn a great deal from a BS/MS in Computer Science.

You missed one of the biggest points. OP want's to do a mechanical / electrical engineering degree. So most if not all of what you're saying goes straight out the window. Actual EEs with a degree tend to go into software because they can make more money, not using much of their degree except in some specialist roles.

That's not to say that they wouldn't learn the things you said they would learn in an actual CS degree. But it it's not what OP is talking about at all.

I say that as someone that does have a Bachelor and Master degree in Comp. Sci.

Sigh. It's also not what was I addressing AT ALL. The line of reasoning was:

sircastor: You (OP) already have a career in software. I’m here to tell you: you probably won’t be learning much you don’t already know.

Me: Yes they will. (Irrespective of whether they choose to study mechanical engineering, electrical engineering, or computer science.)

Fin

And FWIW, OP did mention CS as a possible field of study.

I hear this all the time on HN. I am nearly a neckbeard at this point as a programmer. I have never once used any linear algebra in my career. Discrete math: Only a tiny bit, and mostly to pass HackerRank/LeetCode questions for interviews. Looking back, the highest impact (to me) undergraduate comp-sci course was intro to algorithms. You really use those over and over again, even if you are just "using the library" (vector/map/etc.). It helps to know what and how they do it, even if you cannot write it from scratch yourself.

While you haven't used linear algebra, someone working in MLE would probably find it useful to understand. Now I'm not saying you need to take a linear algebra class to understand matrix/vector operations, but it would be useful as it comes up.

When I did my degree I tried to do as many courses unrelated to the things you listed as possible.

I signed up for as many mechanical and electrical engineering courses as possible, plus control theory, VLSI design, FPGA design, robotics and optimization.