Proposal for New Curriculum
Task force: Walid, Corky, Alan, Joe, Dung, (Steven?)
Week |
Lecture |
Theme |
Content |
Lab |
1 |
1-3 |
Functional programming + types upto trees |
Booleans, records, unions, recipe, (auxiliary) functions, lists |
|
2 |
4-6 |
|
Complex list operations, natural numbers |
|
3 |
7-9 |
|
Trees, mutual recursion, GUI programming |
|
4 |
10-12 |
|
Abstraction in functions and types, first class functions |
|
5 |
13-15 |
|
Accumulators (list reverse), non-structural induction (parsing) |
|
6 |
16-18 |
|
|
|
7 |
19-21 |
Design patterns (transition to Java) |
Immutable list, composite, interpreter, helpers, visitor pattern |
|
8 |
22-24 |
|
Inner classes, GUI, MVC, generics, collections |
|
9 |
25-27 |
|
|
|
10 |
28-30 |
|
|
|
11 |
31-33 |
Programming with imperative features |
|
|
12 |
34-36 |
|
|
|
13 |
37-39 |
|
|
|
14 |
40-42 |
|
|
|
15 |
43-45 |
|
|
|
TODO:
- Can we postpone the effects of 210 to later in the course?
- Yes, but we have to make sure we replace the GUI fun stuff by something else
- Idea: let's provide them with a GUI harness that takes a functional program from them, and the functional program animates the GUI
- Specifying lab projects and weekly homework goals
- More explicit statements of motivation for different parts of the course
- Gain more experience with language levels (Walid)
Things for the second half:
Falling off the cliff:
- Mutual recursion
- Currying/uncurrying
- Graphs
- Backtracking
Ideas:
- Corky: Would like to use a boolean simplifier as a driving example
Concerns:
- Walid: The functional programming sounds like it will be boring.
- Alan: Motivate with 3 interesting problems from start?
- Joe: Feedback from students suggests we really need to emphasize why they are learning this material
- Walid: I've improved this aspect of the course a bit last semester. We need to continue and add to this direction. It would be good to spell some of this motivation in this proposal.
- Joe: Java issue:
- Language levels. People are split on that.
- We need to say something like "language levels let you isolate muscles, as in weight training"
- Need to introduce lab sessions aimed specifically at language newbies
- Walid: It would be good to have a teachers manual for a course like this, and to develop a community of teachers.
- What should we do for the final lab?
- Dung: Spending too much time on functional programming in Java might seem contrived to students. It doesn't seem to save us any time when we starting teaching OO design.
- Walid: The purpose of the transition is to make sure that students understand that the recipe approach evolves into the pattern-based approach in OO, and that patter-design is much easier in the functional setting. (In particular, what we want to preserve is not the understanding of lambda, but the recipe).
- Functional programming is just a backdrop to help us explain data-driven design and the recipe more clearly.
- Dung: We will need to explain how we express the datatype in Java, namely, using inheritance and composition. We really need to work on the details of how we explain this transition to students.
- Joe: Transition to Java requires caution:
- Syntactic hurdle
- Risk of students thinking "Rice research sucks"
- Syntax on a "need to know basis" tends to really annoy students
- Dung: The transition will be a risk, and it will be important to do this right for the course to succeed.
References
Reference materials
Corky's independent check:
Condensing the Comp 210 Syllabux
Lecture 1 Intro
Lecture 2 Old Lectures 2-3
Lecture 3 Old Lecture 4-6 excluding effects
Lecture 4 Old Lectures 7-8
Lecture 5 Old Lecture 9-10
Lecture 6 Trees (11)
Lecture 8 Local Definitions and Lexical Scope (18)
Lecture 9 Function Abstraction: Functions are Values (16-17)
Lecture 10 Lambda abstractions (Defining functions on the fly) (24)
Lecture 11 Accumulators and tail recursion (26-27)
Lecture 12 Generative recursion
...
|
Access Control: (Please don't edit)
- Set ALLOWTOPICCHANGE = Main.CSFacultyGroup