There are not clear test errors, that points to another place in code where mistake is placed.

The course if good, but i believe the content should be more didactic.

Lectures are not easy to understand. It almost feels like they plucked various sections from a longer course.

Interesting course, but far too short. It ends up giving a quick introduction to some of the aspects of parallel programming. I enjoyed the assignments, apart from last one where the difficulty was more about the context of the example than any parallel programming aspects. That example, extended, could be a good capstone project, but I think it was not appropriate as a course assignment. I prefer courses that follow a book, and this course didn't.

Good course and good level on the assignments but needs a lot of improvement on the lectures, specially in the two last weeks.

Wrongly assumes that it'll be easier for students to jump from the conventional 'parallel' java paradigms and constructs to parallel streams.

Not a good job on explaining why. It'll require from students some efforts to create the right picture in their minds.

Exercises are great and partially make up for the above mentioned shortcomings.

Some example code is very bad and goes against all the 'good' things being taught by Martin in the first course of the specialization.

Too much imperative technics in this "functional" parallel programming course.

For now, the worst course in Specialization.

This course was okay, but I feel it could use some work. For starters:

-I feel some of the longer assignments should be split up into several weeks (the Barnes-Hut one for example, since there's a lot to implement and it took a while to understand conceptually).

-I feel that some parts of the assignment specs could be more clear. For example, in the K-means assignment, it was mentioned to use the "square distance" between the means to determine if the algorithm had converged or not and I feel it would be useful to provide a mathematical formula here (or at least give the name of the formula i.e., Pythagorean).

-I wish there was more focus on making the assignments practical instead of something that's more sophisticated. For example, it would've been nice to have an assignment for week 4 that allowed us to practice implementing a variety of parallelizable data structures and focus on making them as efficient as possible.

Good course. But I have a strong feeling that would be nice to have more focus on the practice. Some lecture not really much with the results of assignments. Good course anyway. Pavel

Content is good, less well polished than Odersky's courses though. Painful exercises with a lot of time spent on understanding the instructions instead of focusing on learning parallelism methods (especially the final assignment). Overall happy to have learnt from this, but clarity of assignments should be improved.

In my view, this course was nowhere near as good as the earlier courses by Prof Odersky. The materials and pacing seemed all over the place - in some cases excessive time was spent going through quite trivial concepts, while elsewhere the course seemed quite rushed when looking at much more complicated ideas.

Most of assignments are good. They show applications of parallel programming very well. Some lectures are booring? but important. For instance, lectures about assotiativity.

Not friendly with beginners in Scala. You will be presented with tons of Scala code with minimal detailing on algorithms. Obviously the lecture expects you to be quite familiar with Java and Scala. And the lecture is not very informative either (at least for green hands in Scala). I found that the best way that works for me is to pause at each slide, study the code, and then go on to see what the lecture has to say.

should contain more details

Good learning materials, however the assignments are designed for compute science students rather than professional software developers as the most important part of the assignments is not parallel code, but it is algorithms. Yes, there is parallel code in the code of the assignments, but it is mostly written and your task is to implement the algorithm using this existing parallel code, while in my opinion it should be the other way around - you should implement parallel code and algorithms code should be already provided for you.

I didn't like the last assignment. I didn't think there was much to learn there. I would've liked to implement more splitters and combiners. I did like the first 3 assignments very much. The lectures were pretty good, but I found myself loving some, and hating the others. I would've preferred more practical applications over theory. I think the lectures are un-necessarily math heavy.

Deep lectures, but exercises where more about complex problems than about parallel programing

Not enough Scala. Akka version was better :(

The course presents two core approaches to parallel programming - task and data parallelism. While these concepts are really useful, I expected more details on, for example, how to implement the _parallel_ construct used throughout the lectures. The materials were good and I can use some of the ideas straight after the lectures in my everyday coding, but the assignments were really slightly related to the topic.

Barnes Hut simulation assignment. I'd say it has nothing in common with parallel programming in Scala. It looks like the authors looked for some real-life example or use case for Combiners and this simulation model looked nice. But, really, does any developer not working in university would ever use it? It's completely impractical, the most of the time I spent reading the description of simulation model, but the coding itself was just playing with 2D coordinates and some basic recursion. I'd expect something more practical, e.g. implementing simplified form of Hadoop map-reduce or something like that so I can use it after the course. But Barnes-Hut is too specific and is pretty far from everyday problems developers encounter.

While some the material was useful, I was extremely disappointed with this course. The first two lectures were nearly intolerable to listen to, and the last assignment was very poorly designed and documented. I would not recommend this course to anyone unless it was significantly redesigned and updated.

For a while, that's the worst course of scala specialization. All the parallelism is just obvious division of a one big piece on several small pieses, and perfrom parrallel tasks on it. It's enogh to write merge sort with several threads to get the main idea, which is proposed here. You don't need to watch 4 weeks of coursera video and perform tasks to get it. Maybe I'm wrong, but I expected something more than trivial division. Besides that, I don't like the tasks. And I hate the last one week. Very bad description of the problem, it made me really suffer. I'd like much more the first 2 courses. Which I rated as 5 and 4 accordingly, but this one is fail. Besides that, I like the theory about parallel and concurent programming, and all the stuff about "iron". Hope the 4th one course will be better.

Excellent lectures, let down by the difficulty of the assignments.

While the tutors clearly know their stuff, I found the assignments really quite challenging and of limited educational value due to the mixture of new concepts introduced by the lectures along with new (to me) mathematics or CS that also needed to be grasped.

I'm an experienced professional programmer with the EPFL 'FP in Scala' and FRP courses under my belt, so I thought I knew what to expect but this was much more time consuming than I'd thought.

It's difficult to follow the content. The way it is organized could be greatly improved.

The approach in this course is incorrect, I think that will be a course with more concret examples, but is a course with a lot of teory that is not applied.