The blood, sweat, and tears I had to put in to (probably) scrape by an A was truly excessive. Professor Eggert is a very brilliant man and I look up to him; with his contributions to open source and Linux, it was a pleasure to be taught by such a remarkable human being.

However, his lectures are not exactly the clearest and it gets confusing and dry sometimes. READING THE ASSIGNED READING BEFORE LECTURE IS A MUST. On days I don't, I get lost within 5 minutes of his lecture and there, getting kicked in the dust, steamrolled into confusion. However, if you do approximately understand what's going on in the textbook before you come to class, you will be able to make the connections necessary to understand a particular concept.

Go to discussion and office hours - that's how I was able to at least decently understand the material in general, as they help you clarify the details and such. Swathi was incredible; I wasn't in her discussion but I came to her office hours. She was clear and friendly, and if she's unsure, she won't give the incorrect information by appearing she knows everything; rather she will ask your email and she will take the time to carefully explain the concepts! My TA, Anurag, was pretty good too, as he explains the concepts in an easier manner to grasp.

The tests are insane. I got below the median on the first one and barely above median on the second one. But I but pulled through the final as I took full advantage of the above and pulled a high 80%. Again, please use your resources to your advantage and you won't regret it. The tests will test you on nuances and stuff you have never seen before but able to be solved based on what you learned, so guess and try to write a lot and pray for partial credit.

I honestly slaved away the last 3 weeks of the quarter, seeing the light in the tunnel. So here's something: if you think you might get destroyed in a class, don't lose hope since the final can save you.

What can also save your grade in this class, with the fiery destruction that is the exams? Do well on all your labs, and do your homework! Homework may be 1 percent each but every point counts in case the raw score gets so bunched up together than 1% makes the difference between letter grades.
For labs, try to get a 100 on all. Get help from TA, work with friends, and Google. Also try to get as high of a speed up as you can in the last lab, the OpenMP lab. I got a 2% extra credit (max possible) so that might also saved my grade. This year the highest was 16x speed up so try to aim for that, or higher, if you can.

To the people taking it fall, best of luck to you. Don't lose hope from failing a midterm. It happens, but as long as you score above average in the end, don't worry too much.

DISCLAIMER: I got an A in this course, don't judge me.

LECTURES: As told by all the previous reviews, reading the textbook before lectures is essential because Eggert barely went over the important concepts. The lecture notes (handwritten by Eggert on sheets of paper, projected onto the screen) are barely identifiable. Still, try to note down everything that appears on the screen as they may magically appear in exam questions.

DISCUSSIONS: Very helpful. My TA, Shikhar Malhotra, did a great job covering all those materials skipped by Eggert and offered a lot of help for the Labs.

LABS: Doable, but time-consuming. Lab1 and 3 are significantly harder than the other two. The Lab specs has extremely vague wordings; fortunately TAs clarified most of the problems on Piazza.

HOMEWORKS: Easy, but only worth 5% of the grade.

TESTS: Hard. Each test usually consists of 1-3 coding problems and the rest are short-answer problems. The coding problems are straightforward and doable (similar to those appearing under the 'bitwise' tag on Leetcode), but requiring knowledge of the data representations. The short-answer problems are where students lose the majority of their scores. Wording for the problems are often very ambiguous, making it easy to misinterpret the problems and get a lot of scores deduced. There are some very open-ended problems (e.g. "is this computer structure better than that structure? ") with generous offerings of partial credits; there are also problems that are almost impossible to answer because the materials in the book and the lectures are far from sufficient and only Eggert himself knows the answer. Answer 3-4 problems correctly and getting partial credits for the rest would result in a score slightly above median.

THOUGHTS:

Professor Eggert is an extremely smart and talented person. However, his unparalleled understanding of computers is both his strength and his drawback. His expectations of his students are somehow unrealistic compared to his effort put into teaching the fundamental concepts.

Concepts related to computer architecture are often esoteric and abstract: what are virtual memories? How do caches work? What gets popped onto the program stack? What exactly is an I/O stream? What are deadlocks, and what's the difference between semaphores and mutexes? Thus what most students truly need is someone who can walk them through the process slowly and patiently, providing them with the necessary intuitions. However, Eggert assumes the book can do all those "dirty jobs" for him so that he can go on to talk about some extra material. As a result, most students are left in their own confusions after the lectures.

I would not advise against taking Eggert's courses, as they are great indicators of individual intellect. However, throughout this course I learned mostly from the textbook, google, Stackexchange and slides from other colleges' CS departments, not from Eggert himself.

Browsing through the slides for this course's equivalents at other colleges, what I saw were clear pictures and explanations of what's going on inside those computers. I saw animations of program stacks along execution of each line of code; I saw graphs depicting the mapping between virtual memory and RAM; I saw paragraphs of technical explanations, bolstered by handful of practical examples.

Saltiness rises in my heart, as I am starting to doubt whether this course actually paid attention to students' learning experience. I am not a fan of such pedagogy.

Eggert is a fast-paced professor who goes beyond the bare minimum. His classes are hard, but you will learn a lot.

There is a LOT to learn from this class. However, a lot of the learning process comes from: First reading the book. Second, understanding his lectures completely (Protip: When he writes his midterms topics are DEFINITELY guaranteed to cover his last 2 lectures during the week. But who knows, he might be reading this now). Thirdly, really understanding his lectures using supplemental material. That means other colleges' websites, notes, slides, stackexchange, stackoverflow, etc. Also the labs aren't completely clear and there are bugs in the labs, but it's nothing too horrible. However, for the amount of effort and the grade I got compared to other classes at UCLA and their curves, it wasn't too bad. Thanks Eggert :)

P.S. I think Eggert personally looks over finals and may consider giving some leeway if you're on the edge between two grades. But I'm not sure, that's what I hear from people anyways.

OK. I gave this class bad reviews because he isn't a clear professor, and his lectures are all over the place, not helpful, and hurt my brain to sit through. Never the less, I figured out how to do well and I am not a genius / overachiever in all aspects in life, I just care about CS and wanted to do well. This is what you have to do: Read the book. I don't care what other people say, read the book-- if you CAN, do it BEFORE lecture. He jumps all around, and the days I was most confused and went on my phone most were days when I didn't read in advance. When I read in advance, his words in lecture helped extend the detailed textbooks ideas. Textbook is very detailed but AMAZING. they don't try to dumb it down but they take it step by step. Heres what Eggert wants: Someone with Intuition. You DEVELOP this intuition by reading and asking questions. Non stop. There are many times when I never felt clear about a concept but I still was able to score high on a question about the concept. Eggert knows people are always confused in his class. He cares about you having intuition about what the concept it. He cares that you can see a question on a test that is not something you've heard of before, and have the intuition based on your reading to answer the question even without seeing the concept prior. Heres the great thing: ITS AN EQUAL PLAYING FIELD. NO ONES HEARD OF THE CONCEPTS HE BRINGS UP ON THE TEST. there are MANY right answers. YOU need to read the textbook and ask questions in class in order to develop some type of language to use during tests that shows you are at least TRYING to grasp the general idea of concepts in the class, you don't need to know the "right " answer. Even if the answer is wrong, if the intuition is good or backed up by good logic, you WILL get points. Explain as much as you can on the test and you WILL get partial points, and they go a LONG way . also, he never prepares you or teaches you what are in his labs. You need the internet to help you for labs.

I think the material in this class is really good. It gives you exposure to a lot of software that will relevant your future classes and jobs (i.e. Linux environment, etc.).

That being said, this TWO UNIT class is a lot of work and honestly pretty hard, since I'm assuming most people who take this class have never used Linux, git, ssh, threads, dynamic libraries, or done C programming (i.e. malloc, realloc, and free). A few people I'd assume would've touched on python a little bit, but not enough to make the project associated with the language any easier. Because of your noobiness, you're most likely going to find your entire quarter consumed by this TWO UNIT class because learning something new every week is really hard. A good number of people cope with the workload by just cheesing the class and using online sources to make every project rather trivial... but that kinda defeats the purpose of this class, which is to get acclimated to a lot of software at the steepest learning curve possible.

So that being said, I would say that the time and work that you put into this class is worth it because you learn a lot. It is kinda lame because Eggert could do a much better job at structuring the class to make it easier for the noobs who didn't know what they were getting into (i.e. be more elaborate in his specs). After all, I was one such noob. However, it is definitely a good class to take, and you'll realize this even more so as the years pass by during your journey to becoming a software engineer.

N.B. The finals are pretty random because they're all made by TAs. Some of them were jokingly easy while others were 60% the level of a typical Eggert exam. When I took it, I think had the hardest final (my TA was Lauren Samy). Honestly, it doesn't really matter, because if you know the material and did all the projects legitimately, you'll do fine.

Although the class is poorly structured and definitely way too much work, I have to say I appreciate how I was taught to learn the material: with the steep learning curve from covering new topics every week. I am currently in my first software engineering internship and I have to say that the thought process from 35L has allowed me to acclimate to the steep learning curve of the tech stack used in the company I am interning at.

The assignments are tough since everything is new; I suggest to go through fire and take cs33 too at the same time (I did this plus cs33 Eggert): this is because later on, the linking and threads concepts overlap between these two classes.

For the final, it is luck as your TA writes it. I would suggest printing TA slides, assignments (using small font like 4 point font), ascii table, emacs reference card, regex cheat sheet, and Computer Systems: A Programmer's Perspective (literally, the CS33 textbook). Again, go through fire and take cs33 too since I wouldn't have been able to answer some questions on the final had I not learned it in cs33 or has the book as reference.

Good luck, and I came to UCLA with ZERO programming experience and I was able to pull an A going through 10 weeks of fire and suffering. You can do this; just put in ALOT of work! I certainly think it is very helpful for internships and software engineering in general.

Paul Eggert is a good professor. Understanding his lecture is a challenge but you will learn a lot from his lectures. He is just very knowledgeable and has so much to say about computer system. His test is hard, and sometimes you will find it what the hell the question is asking. I personally like the way he constructs the test, because I think in general, his test is testing on how well you can use what you already know about computer system and apply to his problem, so as long as your answer is making sense to the problem, you will get most of the points. Another reason is when the test is too hard , everyone is on the same boat, you don't have to worry about that half of the class is gonna score 70~80 above, because when you feel like you do horrible and half of the class is gonna feel the same way. My advice: read the book !!!!! it is gonna save your life!

It's honestly been a while since I've taken this class, so I can't really give a really detailed overview of what the class was like. However, I would highly recommend taking Operating Systems with Eggert because he really knows his stuff. His philosophy towards software, his passion, his humour... I really think he's the embodiment of what all prospective software engineers should aspire to be.

So, I would recommend Eggert because he's a really awesome person and you learn the most by being around awesome people. That being said, Eggert is not exactly the best "teacher" per se and he will use a lot of computer science terms without bothering to define them and sometimes, his lectures will be really hard to follow because his knowledge of the material trumps your knowledge of the material a thousandfold. Also, the class is a lot of work and maybe a little overwhelming at times.

However, as Professor Eggert frequently emphasizes, a lot of software is about tradeoffs. In the situation here, what you're trading for your "free" time, unimportant hobbies, and uninspiring social life is an opportunity to learn a very cool subject at the steepest learning curve possible. I would say that's a worthwhile trade.

Edit: Also, I took this class back in Winter 2015, which is the last class based on the old curriculum. The labs have been changed decently so the class might not be the same experience as before.

Seriously the worst professor I have ever had in this school(though it's my first year in UCLA..)
1.He requires a crazy amount of reading prior to each lecture(ranging from around 40 pages to around 100 pages). The book is almost all plain text, with very little graphical illustrations, no highlight and no summary. And it's dry and convoluted. But the professor just assumes you read all the stuff in advance and have a very good understanding about it. Otherwise there is no f*cking way you can understand things from his lectures!
2. As I said it's very hard to follow him during his lectures, but the worse part is that it's even harder to go through his lecture materials after class. Unlike Smallberg his lectures are not podcasted and unlike Nachenberg he has no slides. All he does in class is writing down messy and unorganized notes. And when you go back to try to study from your notes, you will just get completely lost!
3.The exams are ridiculously hard and irrelevant to what what we saw and did in the book.
4. Oh and there are the awful labs... Mostly irrelevant to the lectures, and super hard. Honestly it asks you to find and fix bugs in some long and convoluted codes. It also requires you to have a very solid background in C and GDB, non of which we have covered in depth in lectures.
5. The TAs are horrible. Sometimes when you ask questions to the TAs the TAs won't even know the answers. Also the office hours for the TAs are poorly organized. Most of them will have office hours on the same day, or even at the same time period. Sometimes before the midterms or project dues you won't even be able to find a TA for help.
6. Feel like the piazza platform will be more effective if the professor himself can actually participate in answering students' questions:)

In summary, the sloppiest, laziest and most unorganized professor I've ever seen! He absolutely puts no effort in ensuring that students can actually learn from his class. I don't know why people will say they learn much more from Eggert's class??? Feels like I don't learn anything from him(I don't think that's bc I'm too dumb though since I got As in my Maths, Physics and CS31,32 with ease.) So PLZ dooooo avoid him if you can. He will be your absolutely worst nightmare! Trust me!