Replies: 3 comments
-
+1 on the ownership issue. Multiple people are running into #13. Consumers of API shouldn't need to worry about your library's lifetimes |
Beta Was this translation helpful? Give feedback.
-
There are several crates for this thing, maybe Toasty can integrate or reference them: |
Beta Was this translation helpful? Give feedback.
-
Thanks for taking the time to write this up. Feedback is constructive at this point as Toasty's API is still very easy to change and iterate on.
Since plain-old structs are generated, you should be able to add impl User {
pub fn some_custom_selector(....) -> ... { ... }
pub fn has_billing(&self) -> bool { ... }
}
Definitely, and that is the goal. There was an issue w/ lifetimes earlier (referenced above) but that should be fixed.
I do want a really solid migration story. That said, I haven't yet really dug into it. I'm definietly open to suggestions on that front. |
Beta Was this translation helpful? Give feedback.
-
Hi @carllerche
This is an exciting project coming from tokio.
I've been building loco.rs which I hope fills in the "productivity" part for Rust (referring to that section in your blog post).
We're using SeaORM, which is stable and feature rich, and serves us well so far.
Here are some things I looked for when I evaluated SeaORM (against Diesel, sqlx-macros, etc.), I hope to offer (not to push) some insight for Toasty:
While there are many more bits and pieces (like relationships, lazy loading, etc -- which I'm sure you'll get right as do all ORMs eventually), I think these are the pivotal ones that can really affect an ORM's design.
Hope this help shape Toasty's direction for the better, and good luck!
Beta Was this translation helpful? Give feedback.
All reactions