Johannes Link

Hypothesis' and jqwik's generated are hacked in that only lists that sum to less than 256 are being generated in the first place. As for the the 16 bit two's...

> Hmm... for Hypothesis, this is only half the truth, isn't it? If I understand [this line](https://github.com/jlink/shrinking-challenge/blob/main/pbt-libraries/hypothesis/challenges/bound5.py#L12) correctly, the Hypothesis generator only generates lists of size 1, which IMO completely...

> > Also, jqwik currently normalizes this example to `[[], [], [], [], [-1, -32768]]` which, again if I understand the phrasing of the challenge alright, is a counter-example that...

I'd be happy to change the challenge in that way. @DRMacIver What do you think?

> 1. For starters, it's quite challenging to generate _both_ a random integer **x** and a list of integers that contains **x** at > least twice in order to falsify...

You can do it with a bit more verboseness: ```java Arbitrary strings = Arbitraries.strings().withCharRange('a', 'z').ofMinLength(1); Random random = new Random(42); JqwikSession.run(() -> { strings.generator(1000) .stream(random).limit(10) .map(Shrinkable::value) .forEach(System.out::println); }); ``` That...

I suggest two new overloaded methods: `JqwikSession.start(String randomSeed)` and `JqwikSession.run(String randomSeed, Runnable sessionCode)`. @gypsydave5 Would that fulfill your needs?

I added three new methods to `JqwikSession`: ```java /** * Returns the {@linkplain Random} instance associated with the current session. * @return a Random instance if a session is active,...

Feel free to reopen if any problems show up.

Those features have just been released in version 1.9.1: https://jqwik.net/release-notes.html#191