TL;DR

ECE344 is probably my favourite course from all of undergrad. I learned a ton (as a software engineer it was invaluable), and I had an amazing time while doing so. The labs are more than fair (they’re satisfying to do), and the current course coordinator (in the fall) Jon Eyolfson is the gold standard for teaching engineering/computer science courses (I don’t think I’ve ever seen a RateMyProf rating as high as unanimously positive as Eyolfsons; he literally puts all his content, lecture recordings, labs, midterm/exam and all on his website, it’s truly amazing).

What you learn

How operating systems (primarily Linux) work and are designed. Includes concepts like how an OS is able to run multiple applications (processes) at the same time, how applications can interact with OS, CPU scheduling, file systems, memory management, multithreading.

It’s cool stuff, you might’ve touched on this content in some way or shape on the past, but you’ll learn it all from first principles in ECE344 which is very nice because you get to build a bigger picture of how it all connects together.

What you do

In the labs you’ll write code (in C) that either interfaces with the OS or acts as part of an OS, so the overall practical aspect is lower level than the pure application level coding you’ll have done in APS105 Computer Fundamentals/ECE244 Programming Fundamentals/ECE297 Software Design and Communication but is higher level than the assembly you’d have written in ECE243 Computer Organization (which itself is one level higher than the Verilog written in ECE241 Digital Logic).

When to take it

Take it in the fall when Jon Eyolfson is teaching. Ding Yuan coordinates the course in the winter, and his labs are waaay harder and longer than Eyolfson’s labs in the fall. Not only that, but in the winter they’re done in partners (labs are individual in the fall), they build off of each other so missing the deadline or screwing up for one lab will seriously hinder you for the next lab (labs in the fall are more or less self-contained, although the types of code you’ll write in them will repeat), and if you’re doing interviews for software internships for your PEY co-op then you’ll be much better prepared having taken the course in the fall than in the winter (where you’ll most likely land a position, good or not, before you even finish the course). Until such time as the course coordination changes again (and believe me, people will very clearly let you know what to do then), take ECE344 in the fall, especially if you’re not a software guy, you do not to go through the pain of taking it in the winter when taking it in the fall is a so much better experience.

TL;DR

ECE345 is far from my favourite course, but it’s an important one if you’re going to be working in the software (primarily for interviews, but obviously it comes in handy on the job).

What you learn

Basic discrete math for algorithmic analysis (time complexity and space efficiency), data structures, algorithms (graphs, greedy, dynamic programming). Basically your typical data structures and algorithms course, so you can learn from any of the resources on the internet (shout out to Abdul Bari).

Note that you won’t actually be doing any coding in this course (other than the homework labs if you take it in the winter), rather you’ll be writing pseudocode to describe algorithms (focusing on semantics instead of syntax).

What you do

In the fall, you can expect biweekly homework assignments done in groups of 3. they’re not overly complex, usually a mix of simple problems you can solve directly using the course concepts and slightly harder problems where you’ll need to think for a bit in order to figure out what to do. other than that the course has practically no work, which is nice but does mean that it is easier to procrastinate on keeping up with the content (which you absolutely should be doing because the midterm and final are far from trivial, you need to know the course concepts and need to be actively practicing using them, solving those problems requires being able to recognize what concept you need to apply and how, which largely comes from said practice).

When to take it

Take it in fall for the better lectures, overall better organization, and to help you prepare for software interviews. if you don’t care about any of that and just want to take it to fulfill your kernel requirements then you’ll be fine in the winter since the assessments stay largely the same, but the course is definitely better experienced in the fall.

TL;DR

ECE302 course with low workload, great lectures, lenient marking, and really-good-to-know content that makes it a must-have for all engineers, regardless of discipline.

What you learn

You learn all about probability theory and how to apply it. If you took those Waterloo math contests as a kid in elementary and high school, then you’ll do seriously well because you start from the very basics concepts of probability (discrete probability, random variables, etc.) and gradually ramp up to more complex topics (joint, continuous, multi-variate probability distribution functions, etc.).

This course has one of the best transitions as it ramps up from first principles to the core concepts you learn later in the term. Lectures with Prof. Parham Aarabi (he coordinates the course in the Fall) were an absolute pleasure, I’m a serial lecture skipper and even I attended all his lectures by virtue of how engaging they were (the guy even put in a 5 min break through his 1h lectures, how many profs do you know who do that?). In addition, there was a large emphasis on practice (and rightfully so), which is why lectures were full of opportunities to actually solve practice problems involving the concepts we just learned about in that very lecture, it’s great.

What you do

On the topic of actual work to do, there was hardly any. Your mark on the course rests entirely on the two term tests worth 25% each (both really short) and the 50% final. And that’s it, there’s no other graded work. There are homework problems from the textbook assigned for practice (and tutorials are just an opportunity for you to get guided solutions to a selection of those assigned problems), but that’s it. It’s a fantastic course to take in a busy semester for that reason, and it has firmly cemented itself as one of my favorite courses in undergrad.

When to take it

Course content is largely identical in both fall and winter (there are some slight differences in what content is emphasized in later chapters, but it’s largely trivial with respect to what you need to study), so you’re fine with taking this whenever. Of course, if you’re interested in industries tangetially or directly related to networks (e.g. telecommunications, embedded networks, etc.) then taking this early on will come in handy for interviews

TL;DR

CSC343 is a great course for developing a solid foundation in relational databases, and I can highly recommend it. The content is interesting (if you learn SQL by yourself you will definitely not learn it to the same rigor as in this course), not too difficult, and it’s also fairly easy to get a high mark so it’s a nice GPA booster as well (ignore any older reviews on the internet calling the course “extremely bird”, it’s no longer that ridiculously easy but it’s still on the easier side of things compared to other courses). Lectures are recorded, the homework assignments are tedious but not overly difficult, and the midterm/final have been the exact same format for the past decade so it’s very easy to study for them.

What you learn

Relational algebra, querying and updating databases using SQL (PostgreSQL and psycopg2 in Python), database design (integrity constraints, normal forms), database system technology (querying processing, transaction management).

What you do

A few homework assignments with a mix of writing relational algebra and Python/PotgreSQL code. The midterm and final for CSC343 are the easiest to study for because the format has stayed the same for the past decade or so and as a result there’s a lot of practice material available (which means that studying for this course is almost trivial, excepting some slightly tricky concepts). There are some very good lecture notes by students that can be found by searching for CSC343 on GitHub and Google, I highly recommend using those because they’re good enough that other than those you only really need to write notes for how to solve some specific types of problems (some using code from the homework assignments) and you’ll be absolutely fine.

When to take it

The course is virtually identical in both fall and winter, so you have the freedom of selecting which semester is best for you to take it. Some co-op positions (e.g. fullstack) will value experience with databases so it can be a handy course to have for interviews. For other positions it won’t be of much value beyond being good to know.

TL;DR

I’ve taken this course but I’m getting really tired writing this post (these descriptions are not in chronological order) so I’ll keep it short. You learn about high level machine learning concepts here (so the different types of models, hyperparameter tuning, loss and activation functions, etc).

The midterm and exam used to reuse questions a lot from previous years, but one of the previous profs caught a lot of flack because they reused way too much from a previous midterm (such that it really pissed off some students and caused them to complain) so they’re of a totally new format as of last winter or so and as a result they can be quite brutal (the pre-curved midterm average this Fall was around 40%). Thankfully, there is a very healthy curve that brings marks back up to normal values, but this is something you should know (so you don’t get lulled into a false sense of security from doing past assessments).

There are some PyTorch labs in the course, done using Jupyter Notebooks running on Google Colab, but they’re fairly simple and don’t take long (you’ll spend a fair few minutes waiting for models to train, but they go by fast because the labs are short). The bulk of your time in this course, however, will be spent on the course project, where you work in a team of 4 people to build, train, and test a machine learning model of your own. And trust me, reader, this thing takes ages. If you’re just looking to get by in the course I recommend doing one of the listed project options, but you can also do your own thing with whatever idea your group proposes because having that project on your resume cna b very helpful for landing co-op positions. But yeah, you have to write a proposal and a final project report, and it takes a long time. Your experience in this course lives and dies based on how cooperative your group is, so I would recommend taking it with people you know (you’ll probably be fine taking with strangers, but just make sure you’re making consistent progress because the rushing the project is nothing something you want to be doing with 3-4 other courses in your schedule).

TL;DR

There have been a lot of good things said about this course on the internet, but in my opinion it’s just alright, it’s not bad but it’s also nothing exceptional. I would recommend taking it if you have the space and need a medium difficulty course, it’s not too difficult to get a good mark but it does require that you spend the time to figure out buffer overflows.

What you learn

How to identify security vulnerabilities and how they can be exploited (high level concepts as well as what you need to do in C code; this part is especially important for some of the “simpler” attacks like buffer overflows because they will 100% show up in some form in the midterm and final), how to defend against computer security attacks, cryptography and encryption (including over the web) with lots of acronyms you might have heard of but don’t really know what they do in depth (e.g. MAC, SSL, CORS, MITM, federated identity, etc.), blockchain.

What you do

Other than 4-5 labs that don’t take too long, this course has practically no workload other than studying, it’s really the first lab that takes the longest because you have to figure out how to use gdb to be able to see values on the stack (which is quite cool if you’ve never done it before) but you’re not taught how to do and if this do you have to figure out on your own. In addition, the first labs are all about executing buffer overflow attacks on some short code samples (in C) and they take some critical thinking to figure out (the solution is always really short but getting there takes a while). The concepts used in lab 1 however (buffer overflow, format string, double-free attacks) are the most important out of all the labs (the others focus on a bit of web and application security, so you’ll be working with Python and Javascript; the other labs are a lot simpler though, you’re not writing more than a couple lines of code for any lab). But if you can get that down then you’ll be totally fine, the midterm in Winter 2023 had a really high average and the final wasn’t that bad either (final course average was a B+).

When to take it

The course is virtually identical in both fall and winter, so you have the freedom of selecting which semester is best for you to take it. Some co-op positions (e.g. fullstack) will value experience with databases so it can be a handy course to have for interviews. For other positions it won’t be of much value beyond being good to know.

TL;DR

This course is one of the least useful out of all those offered in area 6 (and I’m including that new quantum computing course and ECE444 too), I don’t really recommend it because it’s a mishmash of a bunch of different concepts in computing like functional programming, Haskell, Prolog, and some OOP. The positions where this comes in handy are a small minority of all those out there, and especially so if you’re looking for co-op positions. Even if you’ve taken every single course in area 6. The only saving grace of this course (and why you might consider taking it) is that you can get a high mark in it fairly easily. Disclaimer though: no professor who’s taught the course has stayed beyond a single semester for like the past 3 years (dear god, I have stories of this one prof named Subhendu Bhadra), so teaching ability is very much a variable thing (although you can probably expect it to be substandard). If you really want to learn the content in this course, google the syllabus, look at the notes, and just search up the content on YouTube. And take a different course in undegrad instead.

TL;DR

The course was completely overhauled this year by Ashvin Goel after being practically identical for the past 15 years. You can find the new syllabus by googling the course, you’ll be working with Go instead of Java now and instead doing a course-long project with 4 milestones you’re simply working on labs.

This was one of the most time-consuming courses in it’s previous form, and knowing Goel’s penchant for making really time-consuming courses (like he did for ECE454 before Jon Eyolfson took over), I would be wary of taking this course in a busy semester.

What this course used to be like in the past

One of the most time-consuming courses in undergrad, the course primarily consisted of a semester-long project where you actually build a distributed database from scratch using Java in a team of 3.

Despite the project taking up practically 80% of your time in the course, it only consisted of 30-40% of your grade, and because it was so time-consuming it meant that you hardly had enough time to study for the midterm and exam (thankfully both were relatively straightforward, despite the midterm being an assessment on Quercus that you had to write in a computer lab, supervised and without any internet access except for access to course resources on Quercus). I can’t stress enough how painfully time-consuming this course was, especially at the start of it when you’re building the core foundation of your application.

Another great thing about this course is that it gives you a chance to use what you learned about threads and file systems/data structures from ECE344 and networks (socket programming) from ECE361 in a decently-sized application. You don’t get that from the biweekly labs in either course, so it’s nice to get that experience so you can talk about in your interviews.