Paul Dicker

> Looks sane. Reviewing this is quite something, mostly because the huge indentation and lack of encapsulation of the old code. So this patch is surely a welcome improvement. Rustfmt...

> I hope this PR can land before it rots and needs to be redone from 0 to keep it a copy-paste. Hmmm, already has a conflict, I'll have to...

@djc Do you want to give this a second look?

Great, thank you for working on this! I had an experiment to replace all the serde boilerplate with a macro. https://github.com/pitdicker/chrono/tree/serde_experiment With that macro adding a new serialization format becomes...

> Here's my first shot at it. Can someone take a look and see if I'm on the right track? [chrisranderson#1](https://github.com/chrisranderson/chrono_fork/pull/1) I could comment on a few nits, but it...

Sorry for derailing. I've cleaned up the idea to use a macro and opened https://github.com/chronotope/chrono/pull/1396. That may take a while to get merged though, if at all. Feel free to...

For initialization the methods work much like the builder pattern, except that we don't need a seperate builder type. We have methods to initialize the individual components as `months()`, `days()`,...

Setting to draft for now. I have most of the parser ready, but need to clean up things a bit more before it is ready a second look.

> (Please avoid adding more stuff to this PR -- other than per review feedback.) Sorry. I added two more things, but minor I think: `#[must_use]` to a few methods,...

Started working on this again. @djc are you sure you don't want to have the parsing code split out from this PR so we can merge a minimal part first?