The books and blogs at least taught me basic syntax constructs like variables, function definition, loops, etc. But I couldn't understand how to package them up (compose) to use them in solving problems.
Then I started studying through Berkeley's 3 course intro CS61A, CS61B, CS61C. They have all materials in the open and if you are following the most current iteration of the course, they even post solutions to the problems. It was godsend for me.
Now, I am confident to learn more CS topics using courses from CMU, MIT, Berkeley, Stanford, etc. The thing that was holding me back was the lack of confidence of programming and understanding really what a program was doing.
Here's my take: if you are interested in computer science, the combination of these classes is great. If you are interested in being a software developer or building things quickly to solve problems there are far better resources out there.
You might be good, you might get some stuff done, but you won't be great.
I say this as someone who has been around this industry a while. You don't need a fancy degree from a fancy school, but you do need a solid basis in theory. Some of the best developers I know never went to college, but they did learn the theory on their own.
And some really poor developers did go to college, but their college didn't emphasize theory, and so they aren't very good when something happens just outside of what they learned, or a new problem presents itself.
Computer science mostly requires a pencil and a piece of paper.
Programming is the act of making a computer do what you’ve written down on that paper.
Software engineering is figuring out what do you need the pencil for and once you’ve done it, making sure computers are doing what programmers think they told them to.
Could you please elaborate on this? :)
Valid criticism. Think for most students, 61A has too steep of a learning curve for an introductory class. For people without any programming experience, it’s a bit too much too fast and with a class size of 2000+, you struggle for resources.
For those with prior experience, it feels like unnecessary training wheels. You learn a little bit of Python, SQL, and Scheme. You have to do environment diagrams (What Would Python Do) on paper, and learn how to use a debugger. Projects and tests have mad-libs style fill in the skeleton code.
In my experience, 61A is a tasting menu of computer science: you take it for the experience, and leave hungry and confused – but excited.
> I tested out of 61B.
OP, you really missed a lot by testing out. This class is half theory, half learn by doing. For example, you learn about red-black trees or Dijkstra's in lecture and then go to lab and implement it.
I know theory and algorithms seem bland at first, but look at the applications. https://www.cs.princeton.edu/~wayne/kleinberg-tardos/
Favorite aspect of this class was designing and then writing a smaller version of git. You had to write a design spec that was checked off, and then your implementation had to pass a test suite.
If that’s not software development, then I don’t know what it is.
> 61C taught me how computers actually work and simultaneously made me quit the major.
61C was not one of my favorite classes because of how low level it is (compilers, RISC-V, concurrency, etc.). But if you stopped taking classes before upper division, you missed out on a whole gamut of “practical” classes like algorithms (CS170) and operating systems (CS162).
Such as?
Genuinely interested.
It's beautiful like SICP, without being too difficult or technical.
It teaches one thing well, functional decomposition.
My 80s high school had no programming classes to speak of, but in my own tinkering, I'd done a lot of BASIC & Hypercard Hyperscript tinkering, & I'd looked over enough assembly & Pascal to get the gist. I felt, "I'm good at computers, I can learn what I need there on demand, I've got to major in college in other deeper subjects that I can mainly learn from those departments." And, "being a software developer" alone seemd too narrow an ambition - I thought I'd instead be "someone-good-at-software" in some other kind of innovative industry.
Still, I took CS60A because I didn't want whatever watered-down stuff they'd provide to non-majors – Pascal, I think, at the time. Within a week or 2, CS60A convinced me: there's more interesting depth to 'Computer Science' worth learning.
That continued through CS60B, and by the time I'd done CS60C & the combined EE intro theory & lab course – EE1(?) then, I think, unsure its current equivalent – I'd felt like a circle had been closed on my understanding of how it all works, from logic gates (which I'd understood since a 5th-grade science project & then "Rocky's Boots") through to high-level stack-based languages.
I was hooked, & then knew CS would be at least one thing I'd major in. I ultimately double-majored it with Economics.
Then once out in paid employment, I'd work with people with Bachelors or even Masters degrees from schools not quite-as-famous for their Computer Science programs, and those people would be missing basic knowledge & intuitions about algorithms, data-structures, & OSes that'd been well covered way back in the Berkeley CS60A/B/C intro series.
Of course, the depth of tools & experience available to kids through high school age is wildly different, & better, now. And, if your highest goal is professional training as a software developer, college itself (& college-style courses) may not be ideal, compared to just doing as much as you can, as fast as you can, in the most-challenging workplace that'll take you, filling in gaps with self-study as needed.
But if you want a high-quality college-style education in Computer Science, with the breadth & depth for lots of possibility-frontier work – not just the surface stuff evident to a younger or outside view – Berkeley's curriculum was an excellent eye-opener for me. For precocious students, it should probably replace any "Python & C++" coding in AP high school classes.
I'm genuinely curious, why did the course make you quit? Was it the material or the professor, or something else entirely?
They probably more interested on how the computers can work for them. These students most likely to work in the banking or finance sectors where the payments are good and bonuses are awesome. Don't got me wrong though, they are good students and engineers, but under-the-hood of computers are not their main interests or passions. For example, someone can become a car enthusiast but can't be bother to open the engine bonnet or pump the tires.
[1]CS 61C. Great Ideas of Computer Architecture (Machine Structures):
We still have not understood how to tach coding. The subject seems to be completely non linear, and some people get stuck and cant move on and the teachers cant figure out why and they cant figure out why, it could last 1 month or could last 5 years, and at some point some random thing they read or watched completely unlocks them.
(For me the thing that helped me was Richard Buckland's Higher Computing lectures: https://www.youtube.com/watch?v=hE7l6Adoiiw&list=PL6B940F08B...)
The best way to learn is to keep trying and particularly trying to understand what you do not understand.
As Feynman said on his last board: What I can not create, I do not understand.
So if you are one of those people who is stuck, or you watch CS61A/B and it does not work for you, just keep looking and trying.
In the case of them not "getting" pointers, for example, getting stuck might reveal they need to go back and solidify their understanding of the memory model, or how data is represented. These are good things to know you don't know!
Even getting stuck on yak shaving can pay dividends – having the student carefully compare their command to the instructions, learning what does and doesn't work, and patiently building experience with their tools. Even if they feel like they're doing nothing, they're probably learning.
So I think the major problem with self-study using online CS materials isn't people getting stuck per se... it's that those materials are no substitute for teachers and a cohort of learners. Some people do just fine without that support! Others do not.
I agree, its like 'level up' in a video game :) but I think it is very hard to not despair when you are hitting your head against the wall for too long. If you check the learningprogramming[1] subreddit it is full of stories of people giving up.
It seems in programming, maybe because the pay is so high, the intersection between good teachers and good programmers is not very high.
I read a some of papers investigating `how to teach mastery` and was comparing tutors vs tools, and the very best results were from tutors + tools (sorry I tried to find links in my history but couldnt). Tutors beat tools, Tools beat not getting any help, but even the best tools were not substitute for a teacher.
Maybe now with chatgpt the tooling will greatly improve, for example having an random error most kids just give up and wait for next lesson to ask a question, but now they can just past it in chatgpt, and most importantly they can ask chatgpt the stupidest and most embarrassing questions that they would never ask in class, or to their teacher.
Is this not the nature of learning in general? Why is it supposed that learning things in the linear A-B-C-D fashion is even possible for most humans, rather than supposing that most people would need to revisit certain topics before learning new topics, e.g. A-B-A-D-C instead of A-B-C-D?
For example if your native language is subject verb object (cows eat grass) it is quite linear to learn subject object verb (cows grass eat) languages (e.g. Japanese), you put in time, and you make progress. There are other subjects where when you get stuck you cant move on, and the pedagogy and androgogy systems we have came up with for math/physics and etc are getting better and better to understand what people don't understand and how to move them further. Which on its own is quite problematic when you have a class of 30 kids and you are moving with the 'average' kid which does not exist.
In the same time, programming is fairly new, and teaching it is still evolving, educators still disagree on what is important and in which order (reminds me a bit of this Feynman video about 'Greek' versus 'Babylonian' mathematics https://www.youtube.com/watch?v=YaUlqXRPMmY)
There was a great example someone used, the amount of people who get confused by the equal sign, and some actually never go through understanding references and values for many years:
x=5
y=x
x=6
print(y)
a=[]
b=a
a.push(1)
print(b)
Then one night on a project I was just staring at the code at 3am…and it suddenly all made sense. Got A’s in all my CPSC courses after that too.
The same can be said about any subject or skill.
If you go to any class, you'll find some people are interested, others aren't. No matter the subject, be it sport, physics, math, chemistry, art, history, etc. Same teacher, same material, different reaction from students.
There's no artificial "ceiling" at play that's specific to coding.
There are countless reports of people 'not getting it' after studying for 4 years, they cant make fizzbuzz, and I dont mean they dont get some nuance of it, they literally cant even start typing.
I dont think there is artificial "ceiling" to coding, I think we still have to learn how to teach it, and particularly how to teach it to a mass of people.
We also can’t teach math well for the masses, there’s a long history of failure. Like New Math or the STEM pipeline. Embarrassing stuff really and I don’t know what a good solution is. It’s one reason I don’t totally hate the new approaches to how math is taught that get lambasted on social media (this is not New Math which was much earlier) because the past isn’t working well enough.
Here’s my post about how to audit 61A : http://blog.pamelafox.org/2022/07/how-to-audit-cs61a.html?m=...
I'm reminded of an experiment I was told to do when I was learning how to be a better cook, and I complained that I didn't understand what bay leaf was adding to a dish. Someone suggested cooking some plain white rice with some bay leaf, and I could finally taste the peppery notes it added. Sometimes you need a neutral substrate to see the details of the thing you're trying to learn.
61B made me feel like I knew how to program.
61C taught me how computers work.
A decade or so later 61A is actually the most useful of the 3 courses now that I’m mid career.
I might recommend that you take 61b first before the other 2 if you’re just starting out
Then I cam across the book How to Design Programs aka HTDP. I was skeptical at first. But I eventually started and worked through the second edition of the book available freely at https://www.htdp.org .
It was so beautiful. The teaching languages and writing examples before writing code. It made me really understand programming.
Now I can pick up any language (not that I become a language expert overnight) in a short time. I only need to map the concepts to this new language. And that's that. I can even pattern match my ideas from higher order functions to imperative looping constructs.
I have another book in to TDR list. It is Norvig's Paradigms of Artificial Intelligence Programming.
I got A's in all of my upper-level CS courses at Cal. Cryptography and compilers were a breeze compared to the CS61 courses. My average grade in the CS61 series was a B-. The semester I took CS61B (data structures), my grade on the final was 21, out of 100. The average grade was 16, and the highest score was 35. (The 35 is a VC now and is fairly active on HN.)
Before coming to college I had tons of programming experience and two interships. I wrote websites, backends, real life applications shipped to customers, school club websites, I even taught C++ in my high school's IEEE chapter. CS 61A completely changed my perspective on how to think about code, I found it so valuable that I could graduate just with 61A, study everything else (algorithms, data structures etc) myself and be fine with it (this is likely an exaggeration of course, there is some value to finishing a traditional CS curriculum).
My only recommendation is to learn programming in a more useful and productive language like C++, as the thing you do not want to do in CS61A is to simultaneously learn to program and learn the concepts in those courses. The right way to think about it is those courses are trying to make you go “ohh that’s how things actually work”.
I had previously tried to learn Java as my first language and never could make sense of all of the "public static void main(String[] args)" and handwaving involved to write a simple hello world.
An interactive website https://seeing-theory.brown.edu that takes you through some of the most critical probabilities and statistics concepts.
Statistics 110: Probability
If you want a more advanced overview of Probabilities and Statistics, this course from Harvard University is an excellent introduction:
https://www.youtube.com/playlist?list=PL2SOU6wwxB0uwwH80KTQ6...
This is MIT Course 18.06, taught by Professor Gilbert Strang.
https://ocw.mit.edu/courses/mathematics/18-06-linear-algebra...
One of the best linear algebra courses that you’ll ever find.
Essence of Calculus
This is Grant Sanderson’s excellent take on calculus:
https://www.youtube.com/playlist?list=PLZHQObOWTQDMsr9K-rj53...
I will forever worship the ground that man walks on for his mathematics help.
https://news.ycombinator.com/item?id=37305485
I haven't gone through all of them but I definitely recommend 3blue1brown and betterexplained
I think you'll learn tons by starting a personal project, tiny, small, daunting, whatever your appetite for pain is, and just doing that.
I thrashed away on my crm like saas for 6 months before the first demo customers were using it. And then worked on it for another year till it was sellable to less friendly companies.
Recently I picked up a JavaScript book and read it. While I learned a few things, I was shocked how much I knew from what I learned from stack overflow and just random reading and thinking during the time of building that CRM.
I was fortunate to take - - CS60A with Stuart Russell - CS60B with John Ousterhout (now Stanford) - CS60C with Paul Hilfinger - CS150 with Randy Katz
I believe (after watching videos online) that CS10 The Beauty and Joy of Computing is also a wonderful intro course to take before the CS61 series.
My daughter took CS61B with Paul Hilfinger and he was so gracious to take a picture with the 3 of us together. I have it here on my office desk!!!
I’ve heard a lot of good things about Harvards CS-50, not sure how that compares.
The book is amazing, but it is also very extensive/overwhelming and I found that the lectures provide a better guiding structure for myself and the lecturers are also quite entertaining at times. The website also provides additional material related to the course such as exams, projects and lecture notes.
Source: I lectured 61A in 2021