Question 1 @ 2024-10-01 18:35

Which of the following points is true about Typed Racket?

---

• A Any type can be a union of several types
• B Only function input types can be unions of several types
• C A type-system is kind of like a simple theorem prover for code
• D A type-system must be simple to avoid heavy runtime penalty
• E The Any type is at the top of the type hierarchy, and it represents knowing the most about a value.

Question 2 @ 2024-10-01 18:39

Consider the following WAE expression:

After one substitution step (i.e., what eval does in one step only), we get:

What should we see in the two ?A? and ?B? places respectively?

---

• xx x, x
• x5 x, 5
• 5x 5, x
• 55 5, 5

Question 3 @ 2024-10-01 18:42

In homework #3, we’ve seen tests like:

Is there any problem with this test?

---

• A

  No.

• B

  Impossible to tell from just an example, since it depends on how the test is implemented.

• C

  There’s a minor problem: it would be better to use (list 3 -3) in the output.

• D

  There’s a problem of testing a particular order of the results, which means that it’s over-testing an implementation detail.

• E

  There’s a possible problem when we consider that Racket should deal with floating point numbers, so using '(3.0 -3.0) is better.

Question 4 @ 2024-10-01 18:45

Given that the answer to the last question

is that this tests a particular order, what would be the best approach to fix it?

(Choose the best answer.)

---

• A Switch the implementation to use sets instead of lists.
• B Sort the results of sqrt.
• C Sort the results of eval.
• D Sort the results of run in the test.
• E Implement a new kind of a test that uses a two nested loops to compare each value on both sides to implement set equality.