Advanced Internet Computing Systems and Applications

https://www.ratemyprofessors.com/professor/1651694

The reviews on ratemyprofessors matches my experience in the class. Writing heavy (8 pager every week). Unclear rubric and flags you for AI even if you didn't use AI. Professor does not care and says Canvas' AI detector decision is final. Leading to a situation, where it is better to have grammar/spelling mistakes in your paper to have a lower chance of getting flagged. The grading criteria for assignments is extremely vague and the feedback you get from course staff is surface level.

got 91% in the end but it was quite a bit of work.

I have a non-CS background so I always wanted to take a course that gives me a survey of all modern internet application technologies in an academic flavour. this course gave me exactly that.

I spent 18 hours on average per assignment (either M assignment which is a design doc or P assignment which is paragraph response to questions based on lecture material. 6 hours per week going through the lecture videos and take notes. Questions were fair but the wordings may be confusing an unclear at times which may lead to points being deducted. TAs have very inconsistent marking standards. Sometimes I get a A and other times I get a high B because I get points deducted for being not detailed enough in my explanation. The 8-page limit made it very difficult to cover all question prompts efficiently in detail.

Exams (2 tests) were open-book and we were given 8 hours to complete the exams. The exams are not cumulative. Questions were fair but the wordings may be confusing an unclear at times.

I wanted to take a less intensive course this Summer and CS6675 fitted the bill because the coding project was optional for Summer 2025.

The course covered many major technologies used in web apps but it did not include content on building LLM-based web apps.

Instructor and TAs were nice. Instructional videos were helpful.

Overall, I thought this course was straightforward and workload was extremely predictable because of the repetitive assignment structure (either one methods or content 5 page write-up per week + 3 peer reviews, + 1 final larger scale project).

As a first course in OMSCS and coming from a non-traditional CS / engineering background, it was fun learning about different internet technologies at a broad brush every week. It was also good that workload was predictable which meant that I could allocate sufficient time between work and study. For weeks where I could not allocate as much time as I would like, it was also good that weekly assignments gave a lot of buffer to decide how much effort you wanted to put in without costing too many marks.

One course improvement I would note is that grading can be rather rigid at times and looks for "keyword spotting" of headers that exactly match the assignment description (which are not that clear in terminology to begin with) rather than understanding the logic of the write-up.

For instance, an evaluation plan was stated to be required for both the baseline and design refinements portion of the write-up. But because my design refinement portion did not have an equivalent evaluation plan even though the key components of my system did not change (and therefore an improved evaluation plan was not needed in my opinion), I was not allocated any marks at all because the evaluation plan was "missing".

But this might be because I'm also not formally trained in writing engineering specs - perhaps the course could include formal methods training in writing software spec sheets as part of the curriculum.

This course was a complete waste of time for someone working as an engineer already.

• Wildly inconsistent grading that seemed to favor a "by the rubric" approach rather than celebrating unique ideas. Oh, you didn't EXACTLY use 1.5 pages per response for a question? Goodbye a ton of points! Who cares if it's a succint response that covers a unique idea? This kind of grading HEAVILY incentivizes students to use systems like ChatGPT to complete their assignments, which was 1000% in rampant use the further the semester went on (as evidenced by the ridiculous amount of peer reviews needing to be done, but we will get to that later)
• The peer reviews need to go, or be HEAVILY reduced. At most, give a student one to do a week. Giving 3 peer reviews, again, incentivizes students to just use ChatGPT.
• Get rid of the "P" assignment entirely. The P assignments were a COMPLETE waste of time. I wish we could focus on...designing systems...instead of responding to the dumb questions that, again, incentivize students to just use systems like ChatGPT.

Honestly, screw this class. I'm pissed I wasted my time this summer and much rather would've done other things.

Spring 2024.

Background: Professional SWE (~3 YOE), but I have no professional experience in web dev and minimal academic CS background, so most of the topics were new to me.

Overview: For someone with my background, this was an informative class, though if you have prior background in this tech, then you may wish to skip it as it only covers the very basics. Though the topics and delivery can be a bit dry, I nevertheless learned a lot and appreciated the structure and pacing; if you took the old iteration of HCI (pre-Spring 2024), then the class is structured similarly, with M assignments, P assignments, and a final (individual) project. Like HCI, it is a no-code class, though you can certainly write code for your final project.

One of the biggest pros of the class is that all assignments except for exams and surveys are released at the beginning of the semester. Additionally, exams are open-book and open-note.

My biggest complaint has to do with how opaque the grading was, even though it was very lenient. I got a very comfortable A in the class, but the rubric is completely hidden, and how many points you lost depended more on which TA graded your assignments than on your performance. Additionally, as someone who didn't have any system design background, I struggled a bit on the open-ended design assignments and ended up winging most of them. Did I learn? Definitely--but it would've been great to have more guidance.

Assignments:

• P assignments (5x): Max 8 pages, due every other week, staggered with M assignments. In these assignments, you digest and regurgitate the lecture contents in your own words.

• M assignments (5x): Max 8 pages, due every other week, staggered with P assignments. These are open-ended "design anything based on what you learned in the covered modules."
• Exams (x2): Open-note, open-book, unproctored MC exams. You have an 8 hr window to complete the exams after beginning. These tended to be curved by 10-15%.
• Peer Feedbacks (3x/week): Get those participation points.
• Individual final project (max 20 pages): This is basically a longer M assignment, split up into a few check-ins at the end of the semester. It can be an extension of an M assignment, and yes, you may do an HCI-esque redesign as long as you discuss the system design changes you need to make to support your redesign. Some people coded for this project, but you can go entirely no-code as well.

I agree with previous reviews that if you don't like writing then do not take this course. Every week you have to essentially write a 6-8 page essay and only in the final few weeks do you have to focus on the project where coding is optional to help you write the 15-20 pages.

On the other hand, they are extremely lenient in marking. I peer-reviewed some extremely poorly written essays, but the averages were always 80+.

The interesting part was peer reviews. Some seemed like they used ChatGBT to write it and few students gave meaningful reviews that help you improve.

The TAs were also receptive to feedback. Students didn't like the brief one-sentence reviews from the TAs so the TAs started expanding which was nice to see.

The exams are open book and no HonorLock is used, but they are on top of the weekly assignments.

The topics were interesting, but the lectures had errors or were confusing at times. I feel like Data Warehousing and later concepts we didn't ingest properly due to the time constraints with the final project and final exam.

0 coding, 100% writing

First off, I will say - if you do not like writing - avoid this class.

Most of the other reviews here are pretty accurate. The synopsis and course syllabus would suggest that this is a system design oriented course that would help with FAANG interviews.

In reality it's just a bunch of busy work and you don't learn much. There is a 6-8 page paper due every weekend, lectures & readings to complete every week, and you have to review 3 other 6-8 page paper submissions every week for participation. Posting on Ed discussion is also important for the participation grade.

The assignment grading is very lenient. As long as you make some kind of effort and write something that makes it seem like you tried a little, you should get a 90+.

There are two exams - 1 midterm, 1 final. Both are open note, open book, open internet. I found some of the questions to be a little tricky but in general if you have a good understanding of the concepts, they are doable.

Many of the topics covered are fascinating and very relevant to modern software systems, but the execution of the course itself is poor. I did not find the lectures to be particularly engaging.

To be honest this just felt like a high school English class where you watch some lectures and read a book and then you are asked to answer questions in paragraph form.

As with most courses in the program, you get what you put in. If you have time on your hands and you are really motivated, yes, you can get something out of this class. But for most people, I would not recommend taking it.

If you don't mind writing, and you're looking for a less stressful course, this may be a good option.

Background: Non-Stem major but have been developing for a while. I honestly loved this class. It's modeled similar to HCI. During the summer it had the right balance of work. There have been numerous instances at my job where I've been able to converse intelligently with industry leaders on topics such blockchain, recommendation engines and even load balancers. The main reason is this class. Part of my job is also architecture, so the class was helpful from the point of view of practicing system design. I was lucky that on some of my peer reviews, the students were knowledgeable in the subject matter to challenge my ideas.