Survey surf

[FintanH]

There is a useful Rust API checklist. It would be great to see what applies to radicle-surf and list out the tasks that we could tackle on the checklist.

I'd expect the output of this issue an issue that contains a checklist of the things to implement. Each item on the checklist could be a new issue once someone is willing to tackle it.

This could be a great source for contributions :)

[FintanH]

Naming

• Check as, to, and into conversions - Relevant ticket: Ad-hoc conversions follow as_, to_, into_ conventions (C-CONV) #71

Interop

• Survey common traits - Relevant ticket: Common Traits #72
• Survey conversion traits - Relevant ticket: Implement Conversion Traits #73
• Should we implement serialisation? - Relevant ticket: De/Serialisation #74
• Thread safety of the library has come up before. git2 does not implement Send or Sync so what can we do here. - Relevant ticket: Threadability #75
• Error handling and error libraries should be used - Relevant ticket: Error Libraries and Practices #76

Documentation

• Better testing examples without unwraps/expects - Relevant ticket: Error Tests #77
• Error and panic documentation - Relevant ticket: Document Errors and Panics #78
• Survey hyperlinks - Relevant ticket: Link all the things! #79
• Survey Cargo.toml metadata - Relevant ticket: Cargo.toml Metadata #80
• Prep release notes - Relevant ticket: Release Notes #81
• Add Contributing guidelines - Relevant ticket: Contribution guidelines #33

Flexibility

• Survey ownership in functions - Relevant ticket: Ownership Rules #82
• Survey the functionality needed in functions and use generics + traits if appropriate - Relevant ticket: Generic Flexibility #83

[kim]

- Survey ownership in functions <https://rust-lang.github.io/api-guidelines/flexibility.html#caller-decides-where-to-copy-and-place-data-c-caller-control>

I hate that this is the opposite of what I thought was sensible.

[FintanH]

I hate that this is the opposite of what I thought was sensible.

I suppose it makes sense if you want to avoid clones 🤔 Like, if you pass in a reference you would hope that it's just the reference being used, but if the function takes owned data then you have to do the clone yourself and you know it's happening, rather than guessing.

Edit: Not avoid clones, but make them explicit

[cloudhead]

Edit: Not avoid clones, but make them explicit

Actually you do avoid clones in some cases when you can move. Imagine eg. where Self::func wants ownership:

let foo = Foo::new();
self.func(foo);
// I no longer need `foo` here, so I'm ok with `func` taking ownership. No cloning.

If func took a reference instead, it would need to clone foo.

[FintanH]

Actually you do avoid clones in some cases when you can move. Imagine eg. where Self::func wants ownership:

Sure, I meant in the case of you having &Foo and cloning in the function. You force the caller to decide if they're done owning by not calling clone (your case), or if they aren't then they have to clone :) I think we all agree 😄