วีดีโอ

Glad it helped 👍 ChatGPT isn't often much help with iced 0.13 so I thought I'd share 🙂

This is why I love the comments 👀👍

lol

Not sure yet, might revisit BDK and make something with iced + BDK, also thinking about doing some leetcode/codewars examples

Ah great. I'll check it out as the range scan was a feature I want to learn in due course. Thank you for the comment. Very much appreciated.

@@learning_rust Pleasure! Please feel free to raise any issue if you have questions :)

I'd love to, my laptop fan starts going mad when I compile it though - I definitely want to do a project with it, I may need to invest in a new computer soon.

@@learning_rust same lol

Ah yes, as opposed to just digitally sign? 👍

@@learning_rust yes. Let’s see if i can store it in a file somehow

well spotted, yes, that does work on my other laptop, I used this one for recording. I need to compare init.vim files, and get this one to do show method names.

In Rust it is quite common to modify the argument in-place as the '&' means it's a reference so it saves duplicating - but, if you prefer more functional programming style you can use "clone" but that can be more expensive.

Ah cool, thank you. std::ptr::write 👍

🤣👍

Ok, so return... -> impl instead? 'static dispatch' still learning - so all advice welcomed

Something like this? play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=5cb6d0dc5698d7ed189ea366d6fd4b1d

@@learning_rust Yeah, my understanding of the factory pattern is that you shouldn't worry about the exact implementation. That's where the factory comes in. But I usually use design patterns fairly loosely (when I use 'em) See my other remark on what I did with it. Your stuff is really inspiring to play with the concepts

@@learning_rust Yes, that was my interpretation too

Understood I'm trying to nail the theory, and then deviate/use more loosely exactly like you say, I'm going to revist "Rust in a Month of Lunches" page 244 (I'm not on any commission, but it's such a good book). Thanks for taking the time to comment 👍

Yeah, I think you can avoid using them *most* of the time!

UPDATE / SET... Something like this : UPDATE customer SET email = "newemail@example.com" WHERE id = "customer123"; and the you could leave name, phone number, address unchanged for instance?

@@learning_rust hi, thanks for reaching out. I already fixed this. I used ‘patch’ & “pathop”. I’m enjoying it, thanks.

Glad you like it!

tbh, the link to the code is in the description - that might be easier to unpick!

@@learning_rust the code isn’t the problem, it’s getting my head around the arc/mut thing

I studied the example in the rust docs and then typed it out over and over until I could do it from memory. Tbh, I probably knew how to type the code before I fully understood it! As long as you remember that you're cloning the reference, rather than the data that really helps. Also, you have multiple threads that have access to the data, but the mutex is making sure only one thread can edit the shared data and any given moment. I'm still learning this, it's not easy is it! ps. I still try and type this from memory maybe once a month, even if I've been doing something else : gist.github.com/rust-play/134777f5c866babddbd6464d82d04d14

Yes, not part of the standard library, as you may have noticed the standard library is deliberately kept slim, but there's a crate available : crates.io/crates/dashmap "DashMap is an implementation of a concurrent associative array/hashmap in Rust."

Yes, correct. They could be made global variables, that would eliminate the need to pass them as arguments to the get_user function, But..there's always a but isn' there? As I understand it, "using global variables in Rust is *generally discouraged* because it can lead to issues with thread safety + testing"

@@learning_rust thread safety: even when you pass it to the route handler, you have to deal with thread safety too, right (because there'll be multiple requests coming to the app at the same time)? testing: yes and no. Yes: it seems that if we use a global variable, we cannot mock it and pass the mock to the route handler. No: in this particular case, I prefer not to mock postgres. We can use testcontainers to spin up a postgres server when running tests and the PgPool will use that postgres in the docker container. That requires PgPool being able to connect to a postgres server specified in some system properties (system properties is a concept in Java, it's not environment vars, not sure if Rust has something similar).

@@avalagum7957 Yes, when you pass PgPool (a database connection pool) or a cache to a route handler, you do still need to handle thread safety. (Arc). + Yes, there is this crate : "testcontainers v0.21.1" - 'A library for integration-testing against docker containers from within Rust'. + No, there isn’t a direct equivalent to Java's "system properties," but I have read this elsewhere : "Rust's approach is more in line with Twelve-Factor App principles, which emphasize using environment variables for configuration, especially in the context of cloud-native and containerized applications."