apint icon indicating copy to clipboard operation
apint copied to clipboard

Checklist for version 0.3.0

Open AaronKutch opened this issue 6 years ago • 6 comments

Here are the PRs that will need to be done before the crate is in a state that could be released.

  • [x] Rust edition 2018
  • [x] Rustfmt PR
  • [x] Fix constants from digit.rs usage
  • [x] Gather the standard ops trait implementations from around the library and put them in std_ops.rs
  • [ ] #41
  • [ ] Extract the shift operations and use the redesigned ShiftAmount from PR #36 to fix issue #28 and issue #23
  • [x] Replace the Bit construct to directly use booleans and figure out what functions to keep for issue #42
  • [x] Use NonZeroUsize (#39)
  • [ ] Use MaybeUninit
  • [ ] Fix big and little endian problems (#20, fix the from_vec_u64 TODOs)
  • [ ] Better constructors (#44) (not the arbitrary concatenation stuff, only constructor stuff)
  • [ ] Extract the multiplication and division code from arithmetic.rs into mul.rs and div.rs to use the simplified design from PR #36. This will also fix the remaining warnings and clippy warnings #48.
  • [ ] Extract the fuzz tester into fuzz.rs. #47 is unfortunately going to go unfixed until Rust fixes and stabilizes stuff related to RFC #2091.
  • [ ] Get MVP string serialization and deserialization working #25
  • [ ] Error PR?
  • [ ] Fix all of the TODOs besides performance related TODOs
  • [ ] Do any contributions.md stuff we want
  • [ ] Reorganize the entire crate's structure (#29) (not necessarily required, but would be nice)
  • [ ] Release 0.3.0

AaronKutch avatar Jul 23 '19 18:07 AaronKutch

I updated the checklist with everything I could think of. I have a week left before university starts again.

AaronKutch avatar Jan 07 '20 01:01 AaronKutch

Thank you very much for this nice list! Many important points! However, I am a bit unsure if the time plannings of one week are a bit optimistic. ^^ After the 0.3 release we should form a github organization so that we both are equal owners of the crate. What do you think? Would provide you with more power and rights.

Robbepop avatar Jan 07 '20 12:01 Robbepop

I'll stay as co-author for now, and maybe reconsider after 0.3.0. Would it be possible for you to review a PR every week, only skipping a week occasionally whenever work is too hard or something? In the past there was periods of months between reviews, but I estimate that there will be ~20 PRs in between now and where I want 0.3.0 to be.

AaronKutch avatar Jan 13 '20 00:01 AaronKutch

It mostly depends on the reviewability of the incoming PRs. If a PR is easy to review I will mostly do it right away. If it takes several hours on the other hand I am taking more time to start the review.

Robbepop avatar Jan 13 '20 10:01 Robbepop

This semester is by far the hardest semester I have ever had. I have had to work on weekends and even during this spring break to keep the school work from running me over. I think I am feeling some burn out right now. I still want to get one more PR in before end of semester stuff kicks in.

AaronKutch avatar Mar 11 '20 23:03 AaronKutch

Issue #42 is fixed now, lets keep that issue up a little longer in case we encounter a problem. I will do a PR for issue #41 to have BitWidth construction use TryFrom and other stuff. Then, I will have a PR to extract a new fuzz.rs using improved BitWidth ergonomics, testing, and feature flags. Then, you can get your GitHub Actions PR working.

AaronKutch avatar Jun 07 '20 21:06 AaronKutch