Computer Networks

I came in with network experience, so the first few weeks weren't terribly exciting or new for me. However, if you are new to networks, I think the information can help understand what's going on behind the scenes.

The TAs were available throughout the course. There was a lead TA for each project and they took the time and effort to make interesting problems to solve. I would recommend decent Python knowledge if you don't want to struggle mightily with the projects. Nothing dramatic, but understanding inheritance will go a long way.

The lectures were an abomination. The slides were fine and provided information. But the actual recorded video may have been recorded during a hostage situation. It was not motivating. And the professor popping up twice - once before the midterm and once before the final - to review quiz questions that aren't related to the exam didn't seem terribly useful. The least interested person in the class should not be the professor.

The exams were a minor annoyance. If you went through the slides you'll do fine overall.

Overall, there were interesting parts to the course.

I came into this course with no formal experience in computer networking or network engineering since I needed a class for registration and the waitlist for the other courses didn't have much momentum.

For me, I needed to play catch up in terms of assumed understanding of how subnet, NAT router, ports actually work to make sense of the modules.

Content:

• Pretty dense modules and you have to read carefully
• Kurose book and the youtube channel are helpful if I get confused by the professor's explanation. There were a few times where I needed to reference Kurose book for a clearer explanation

Projects:

• Required an intermediate level of knowledge with Python. Should not be too bad compared to other classes like DL from an "amount of coding perspective"

• If you leverage the resources provided and READ the specifications BEFORE diving into the code, you should be able to get full score on Gradescope

Community:

• The TAs are pretty available on Edstem. There are dedicated TAs for the programming projects, content Q/A and general office hours

• For the projects, the TAs provide chat sessions where you can ask questions about the projects. I found going to these really helpful in coming up with a simple, sufficient approach for the project implementation. Also, the chat sessions are very underutilized, so you should attend these when the projects come out to get the most support.

• The wider office hours from the head TA aren't really productive. It's just more of administriva and very limited questions are asked

• Professor Konte releases module summary videos which help to identify what are the main takeaways from the module. I didn't see too much of Professor Konte in the live setting so far

Some areas of improvement:

• Please update the documentation for how to setup mininet. Turns out the documentation provided in the Ed megathread was not accurate for Mac ARM processor devices. It would be nice to have a step by step video that's updated

• The head TA doesn't really seem to answer a lot of questions. There are times where I asked a legit question in office hours and the head TA "shoots down the question". If you're gonna be helpful, at least answer the question or point in the right direction (eg: linking the edstem thread). Don't just answer every question with "that was clarified earlier". Most of the program consists of working professionals that also have lives

• That being said, the project TAs and the content specific TAs are pretty good and clarify the doubts

• Lower the weight of individual exams and increase the weight of the projects

I learned quite a lot from this course, mainly from reading the textbook and doing the projects. The lectures are fairly boring for many of them, especially when half of them are just paraphrased passages from the textbook. I strongly recommend reading the textbook, it goes a bit more in-depth and helps you understand all of the required materials. The quizzes are easy as long as you have a good understanding of what you just learned. The assignments aren't that bad, I finished them in half a day each project. The exams are completely fair. The professor is mostly missing the entire semester which is unfortunate but, it is what it is. Overall, good course, but lectures could be improved.

CS6250 - CN - Is not a serious master's level course. The 4 coding assignments are trivial, and the quizzes' and exams' MC questions are taken nearly word-for-word from the lectures. Quizzes are open-book while exams have you install a spyware on your pc to ensure you are not opening other tabs (recommend just booting from a live USB).

Since they are all MC, obviously memorization of the theory is all the professor is teaching you. However, the level of rigour here is more like that of a high school class; it is all only a high level overview of the topic. "Lectures" are all just short webpages you read (no textbook) with a few of them having optional Kaltura narrations by the professor. There are "required" and "optional" research papers but you won't be tested on either - so why bother reading them? I honestly believe I could've learned all the content in a 2 hour YouTube video.

Lectures are definitely on the dull side as others have mentioned, but not necessarily poorly written. They still taught me a lot, and I felt good about the amount I learned in the course.

Each lesson's lectures took ~3 hours, and projects took anywhere between 5-8 hours. Some weeks there were multiple lectures assigned, but this might have been because I took it in the summer.

Regarding exams: They could be a little better written, but you don't need to go crazy and memorize everything in and out like some people would have you believe. I answered the questions in the study guides provided by instructors, and got 90%'s on both exams. Exams only count for 30% of your overall grade, and getting near 100% on the other 70% (projects, open-book quizzes) is straight-up easy if you put in the time needed. That's not to say you won't struggle on the projects - I was banging my head against the wall for the first project for a couple hours, but then I read some forum posts, thought about it, figured it out, and got a 100%. So outside of the exams, you shouldn't miss many points.

On top of all of that, you can get another 3% back from extra credit assignments. That means you only need to get 17% of your overall grade to come from the exams; 17/30 = you need a 56.6% average score on the exams overall. Anyone should be able to do this, even if you're not great at memorizing.

Point being, reviews that note the exams' focus on memorization make a point that has some truth, but it is nowhere near to the point where it should affect your grade unless you really mess up in other areas of the class.

Took this course over the Summer to catch my breath after AI (Fall 2024) and DL (Spring 2025) and before ML (Fall 2025) and IGA (planned for Spring 2026). For my purpose, the course almost worked.

The good:

1. Learned something new (e.g. how to collect and analyze BGP data, how DNS resolvers work, how firewalls work)
2. Indeed, the work load (aside from the exams) was not overwhelming. I've managed to take a family vacation (though had to spend some time studying during it), do some projects around the house and even watch a TV series :)

The bad:

1. As mentioned by almost everyone, the quality of the "lectures" is unacceptable
2. The closed book exams were very tough for me. I cannot memorize random stuff in such amounts. In our term, the exams were worth 30% combined one must study and memorize a lot for them, no way to score much using common sense or just by luck. More importantly, I see no educational use of such exam form - I scored high enough on both of them, but now, just a couple of weeks after the final, can hardly remember anything.

The final note: some reviewers compare this class to KBAI. I strongly disagree. There is no class (or at least I have not taken one) worse than KBAI. In CN, at least I learned something and it was not nearly as stressful and required nearly as much stupid work as KBAI. I really cannot understand why KBAI class even exists, it has no reason to.

Me: I had an undergrad in CS. I started in Fall '24. This was 3rd semester. I've taken:

• HCI. High A.
• KBAI. High A.
• AIES. High A.
• RAIT. High B. I was lazy. It's an easier A than KBAI.

I'm a full-time student.

CN Grade: A (90%)

It's a good course. I don't think any prior knowledge is required. Just know python and be intermediate at least. The course is very easy. Assignments are easy. Exams are MCQs, closed book and easy. There are weekly open book quizzes which are also very easy. I was very lazy over this semester. I did everything, literally, one night before. Every quiz and assignment. The lectures are not too necessary to do the assignment, but you should still watch them before starting the assignment. Alot of the time I didn't, and watched + did the assignment over night. I don't recommend it because sometimes I'd have to submit late with minor penalties (-10, - 5, depends how late you submit). I also studied for the exams one night before and did relatively well and above average.

Obviously, I don't recommend any of this. But this just goes to show that it's pretty easy. There's also an extra credit assignment that I didn't do which would've made things even easier.

All in all, decent class. I just took it for an easy A. It's not very relevant to me and as a result I found it pretty boring.

I've decided I don't really care about my GPA as long as it's not too bad. I'd rather take hard classes that'll teach me valuable things. If you're looking for an easy A, I'd recommend you take up my new thinking. Because from what everyone says, no one cares about GPA. Unless maybe you're going to pursue even higher education.

Courses I've taken in order of difficulty: AIES (1/5) < CN (1.5/5) < HCI (2/5) < RAIT (2.5/5) < KBAI (3/5).