The recent reviews don't lie; the class, and professor rosario, have improved since a majority of the reviews were left. I took it online and we had 5 homeworks, 1 extra credit project with spark, a midterm, and a final. All of which were pretty straightforward, and the curve prof rosario gives is pretty generous. Additionally, I never felt like prof rosario was belittling students like some previous reviewers mentioned; I asked some pretty dumb questions in office hours and class, and never once did prof rosario make me feel like an idiot (even though I probably am at times). I'm not too interested in databases, but I'd recommend most CS majors to take some offering of this course to get a fundamental understanding of databases.

I remember seeing Rosario get a lot of from prior Bruinwalk reviews and actively tried avoiding him - then I took his class and realized how WRONG those reviews were. He's completely revamped the curriculum post 2019 and it's solid.

Rosario is awesome. His lectures are clear + engaging and follow the slides in lock step (he posts slides prior to lecture, annotates during class, and re-posts after).

Best thing about him is he works in industry as a Data Scientist / SWE, so every topic is covered with practical, real-world examples that we can actually use.

And he's a nice guy who truly enjoys teaching, and he handled COVID better than any other professor I've seen so props!

For the first 10 weeks I really wanted to give a good comment for this professor, because his average response time on Piazza is 40 mins and he is always willing to support students (he gives us an extra 1 hours to do midterm), and his (kinda) engaging lectures. His homework (which is 40% of the grade) is generally not hard, except for the last one or two, and he gives extra credit for the one mini-project.
However, things changed. You wouldn't expect the final would be much much harder than the midterm. In midterm, we have many SQL related problem which is clearly defined in class or in textbook (or slides). Well, in the final, even though we have 24 hours to work on it, many friends of mine and me have to work at least 10 hours for a supposed 3 hours exam (he told us this exam was written for 3 hours). There are a bunch of explanation kind of questions that you can't really find the related information in the textbook (or slides or lectures), and even the textbook doesn't go in detail for those concepts. So, you have to really really think out of the box, and like use what you have learned from the past 20 years maybe, to compose an answer, not to mention that he gives few partial credit for answer that is not in his solution. Putting it simply, it's not just testing what you learn in class, it's like asking you to publish a paper for some of the questions. THIS IS INSANE.
All in all, it's not the kind of easy-A CS elective class you are looking (maybe not even easy-B). However, if you really like database systems, the lectures are good (JUST DON'T TAKE THE EXAM). So if you have other choices of professor teaching this class, go for it.

Rosario has really chilled out since last year. He was a friendly and helpful instructor during Spring 2020. The homework assignments were really doable, and the tests weren't insane.

He sucks and I hate him a lot

TL;DR - do not take CS143 with Rosario. Just not worth it. I made a Bruinwalk account just to post this review in the hope that I can save someone from this class.

Rosario has a massive superiority complex and it seems like he teaches only to get a power trip. He assumes everybody is against him and treats students as such. He is rude and condescending and does not want to help you learn unless you suck up to him at office hours. He cares more about being right and powerful than helping us learn what we joined his class to learn. Tests/specs are unclear, he doesn't like to answer questions, and is borderline abusive on Piazza. A friend of mine was outright cyberbullied by him and others on Piazza, administration did nothing.

I as well as many of my friends have complained to administration about him so I am frankly astonished he is still teaching. This man should not be teaching. Do not think your experience will be different -- do not take this class.

Rosario is actually pretty nice, approachable, and knowledgeable. But if you anger him you will get the hardest final exam of your academic career and wish that Eggert wrote the test instead. Spring 2019 got pwned lol.

Bad!!!!!!!!

---TLDR---
In my opinion, Rosario's intentions aren't expressly bad, but he often acts in ways that don't help anyone. He frequently characterizes things he doesn't seem to really understand as 'unnecessary', 'impractical', and 'not relevant to industry'. He frequently tries to act cool - and sometimes is cool - but often simply acts in unreasonable ways.
Most of the things I wanted to say are already written on bruinwalk, but here's some more stories.
---Stories---
When discussing some dependency stuff, topological sort came up and he made a mistake in explaining the material. Mistakes are honestly fine, but what isn't fine is that when a student questioned it during lecture, he quickly disregarded it and became increasingly upset as more students realized he didn't know what he was talking about. He then proceeded to yell at the student, rush through the slides, and then, near the end, reference his meme of a crying baby to state that everyone is picking on him (the "picking on" text was underneath a picture of a crying baby on the slides). I guess people have their off-days and its not a big deal, but he refused to ever clarify the material, and then put a question on the final regarding it. I feel that's not very cool of him.
I might give him a pass for this whole topological sort fiasco if it weren't for the fact that for the midterm, he said "we will only make passing reference to them" regarding recursive and window queries! And then he put them both on the midterm! (Rosario often says that he fully specifies what is on his exams - this is not true, do not believe it.)
Moreover, he said in office hours once that triggers and constraints will be extremely important for the test, but they only came up as a tangentially for one out of 200 points!
Also, he said that B+-Trees were not to be tested (because he severely messed up teaching the topic, and was unable to explain the difference between a B+-Tree and a B-Tree during lecture, and even the homework on the subject was broken because the insertion-style was ambiguous and inconsistent lol), but then it was on the final.
Calculating the value for hyperloglog was on the final. Calculators were not allowed, so we were expected to compute a*m^2*Z where Z = 1/sum_{j=0}^m 1/2^{c_j} and then round only at the last step? Honestly I don't know what that formula means, I just copied it from the slides right now. I left that question blank on the final because I didnt have the formula lol.
Oh also, imo, his slides are very unclear. just fyi. I relied on other school's notes the entire quarter tbh.
---Reddit information---
Before you take a class under this professor, be aware of this (seems that Rosario reported a student to the police regarding a discussion on reddit?) https://removeddit.com/r/Professors/comments/bzyco2/the_attendance_dilemma/ercowpa/?context=3
---Final distribution info---
For anyone looking for stats regarding the final, a 100/200 is a B, mean/median is 42%, high is 72%

So I signed up the account just for this professor, I got a bad grade in this class but don't get me wrong I get B+ for other cs courses as well but this grade is really not what i was expecting.
Projects: To be fair, the projects are interesting and I actually enjoy doing them. However the timeline RRR set up was ridiculous. For project 1A we have two weeks just to load the data, and for the same amount of time we are asked to do a much harder task.
Exam: Midterm is fair, we are tested on SQL queries and related stuff, fully expected. BUT THE FINAL, it was a disaster. For the final RRR tried really hard to make the questions in an eggert style but failed. We were not given any practice tests and the content was just randomly picked from the lecture slides... the average was around 84/200. And one of my friends just told me for the final he "expected" us to do better (maybe because he thought the questions were easy?) but as you can see from our average they weren't. So he just made the average of final a C. Well, I don't know what else to say.
Overall the class content is fine and interesting, but try not take it with RRR.