|
Course:Formal Semantics of Programming Languages (CSC-535-901) Quarter:Spring 03/04 Time: W 17:45 - 21:00 Location: Loop Campus |
James Riely PhD Associate Professor jriely@cs.depaul.edu Instructor homepage
|
Select a Page:
Summary 1 2 3 4 5 6 7 8 9
What are the major strengths and weaknesses of the instructor?
1. Strengths: Copious lecture slidesWeaknesses: Makes poor use of white board -- no pictures to illustrate concepts
2. Because the material is in Professor Riely's research area, beginner's problems weren't presented well.Following the text's presentation was not helpful--it would have been better to get a basis quickly and then move forward to more complex topics. We got bogged down in basic evaluation stuff and it took a while to get back to really making progress.
3. Strengths: He knows a great deal about the subject. He is very willing to explain things 5 times if necessary.
4. excited by the subject matter
5. He knows the material a little too well :)
7. I like the lessened emphasis on assignments, instead of under constant pressure to finish some homework every week, given more time to review material and prepare for midterm/final.
8. Knows the material very well. This is both a strength and a weakness. He can teach some of the most esoteric areas of the topic, but may not notice that as students we struggle to grasp it.
9. riely tends to go on tangents when talking. this makes it sometimes difficult to take notes during the lectures. i am not saying get rid of the tangents. he should just complete the entire thought that he is talking about before he pursues the extra stuff.
What aspects of this course were most beneficial to you?
1. Can't think of any
2. The project is providing the opportunity to research an interesting topic.
3. Understanding what properties make a language what it is and, according to my interests, how to judge the robustness of a language.
4. lambda calculus
5. I understand why certain languages evaluate statements in the order that they do.
6. The Class helped me to understand alot of under the surface qualities of Programming Languages.
7. This course forces you to think in a totally different manner towards programming languages.
8. Lecture. Prof Riely definitely improved on the coverage offered in the text.
What do you suggest to improve this course?
1. Style of lecturing could be clearer. Students tend to lose interest.
2. The textbook for the class does not present the material effectively. I think we could have covered the first five weeks or so of material much more quickly to get a basis to move forward into more interesting material. The fundamental concept of the class is really about types in programming languages, but we really didn't get to that until late in the class.
3. correct homework at the beggining of each session, and solve any question that could arise from those.Also, post the slides with a couple of days in advance. They help reading the text and allow us to print the material timely before going to class.
5. Spend more time on evaluating plain lambda expressions. Nobody I talked to cares about proofs or 'write this program' questions. I get lambda, but we never practiced enough that I didn't cringe whenever I saw it. It's like math. You don't spend a lot of time proving anything, you spend a lot of time executing equations until you can do them in your sleep. Then, maybe you prove it.Also, get a book that has more problems that don't suck. Every exercise in this book was a 'thinking problem'. I like to think, but I like to understand what I'm thinking about first. Some (20 to 40) basic, basic problems about each subject would be good.
6. It seems like there is almost too much information for one class, maybe not enough for two but possible break the class down differently.
7. No take home finals. Take home exams are always a bad idea, especially in a course like this where there's alot of material and its not just about memorize x y z and perform problem a b c.Take home tests are evil.Also, the final research paper is proving to be a difficult thing to write. Maybe next time you teach the course, give the students more info on what you want to paper to conclude. At least to me, its been difficult to write about a subject that already has research papers that do the exact thing, they spell out the semantics, give examples, reach conclusions and say what the point of this subject is. I'm left with very few things to add.
8. Have the homeworks correspond more closely with the type & difficulty of the questions on the exams. Use a different text: Pierce' text was very difficult to follow. It often jumped back and forth, alluding to material presented earlier without really placing it in context. If the course relies on OCAML (as was the case here), then time should be spent in lecture reviewing OCAML.
9. it would be helpful to have discussed the term paper earlier in the term so that we could have started it earlier. it also would have been helpful to have a written assignment, even if it was just a few lines indicating that we are supposed to write a 10-30 page paper comparing two or more ways of solving a problem in the field of programming language semantics.
Comment on the grading procedures and exams
1. Homework was not really a serious part of the final grade (10%), which for a 500-level class is reasonable, but I would have appreciated a system where a point was given for a question only if the question was answered correctly rather than a point for simply doing the work, even if it was wrong. At least that approach allows students to know what areas they need to study better.
3. I think it is ok. I am not sure that I like the binary grading system in that it does not provide good feedback.
5. Grading is fair.
7. Very fair, I like the homework grading, homework is just for practice...it shouldn't be a very high part of the overall grade. This is the only class so far where the professor realizes this.I have time to play around with problems, think about them and thus be better prepared for exams.
8. Grading was fair, but the exams were surprising. The homework assignments came from the text, and were very difficult to follow (which is a recurring theme with regard to Pierce's book), Prof Riely's exam was challenging, but if he'd been writing the hw ?s as well, then I think the course would've been better. As it was, there was a significant difference between the style of ? for the hws vs. the exams, which made it more difficult to handle.
9. there needs to be more feedback on the homework. a binary scale (based on whether something was turned in for the problem) by itself doesn't adequately help students know if they are doing the right thing for a particular problem. if riely wants to grade that way to make it easier, that is fine, but there should be at least a comment for each problem indicating whether the solution that was turned in (or portions of it) is correct or not.
Other comments?
1. Dr. Riely really understands this stuff, but his lecturing style made it tough for me to follow him. That may have something to do with this being his research area. Since it was hard to really absorb information during lecture, I had to (try to) teach myself the material which takes up a lot of time. The text book was difficult to follow -- I got really sick of flipping pages because the book doesn't do a good job of tying concepts together when extending semantics from previous chapters.
3. a very interesting course, but the material is very dry. :)
5. Give lots of easy problems and a couple hard problems for homework instead of the other way around.
7. You are the most relaxed professor I've ever had...keep up the good work.Too many professors put students to sleep.
8. CHANGE THE TEXT. Minimally, if Pierce is kept, then Prof Riely should take a more proactive approach in writing out his own hw questions and should probably develop his own set of presentation slides, instead of using those from Pierce. This way the important topics can be focused on and the other stuff left off, instead of skipping 30 slides here or there.