James Kominick comments

Results 104 comments of


                                            James Kominick

Would using RwLock instead of Mutex make sense?

There's no reason other than that I haven't needed it and no one's asked for it yet. We could add an option to the macro `#[cached(rwlock)]` and have it use...

Would using RwLock instead of Mutex make sense?

Ah.. you know what... I forgot the reason a RwLock won't work is because of the signature ``` /// Attempt to retrieve a cached value fn cache_get(&mut self, k: &K)...

Would using RwLock instead of Mutex make sense?

Since a RwLock (correctly) prevents you from mutating on reads, the only scenario where you could use a RwLock is if the backing cache is a plain HashMap with no...

Would using RwLock instead of Mutex make sense?

Ah I see. I don’t think I’ll have time to update the macro this week, so in the meantime you could emulate what the macro would be doing (and have...

Would using RwLock instead of Mutex make sense?

Nice. Yeah, the `already_set` bit isn't required, but it would prevent multiple concurrent calls from calling `get_new_token` when the token is missing or expired. Without it there may be multiple...

Would using RwLock instead of Mutex make sense?

@adambezecny I'll take another look at this tonight and add a `#[once]` macro that implements the `RwLock` (only caching a single value) version we were chatting about a while ago

Would using RwLock instead of Mutex make sense?

@adambezecny A `#[once]` macro is now available in `0.26.0` https://github.com/jaemk/cached/commit/0e36dbd894867a635c94e433604bb8333976ed24

Would using RwLock instead of Mutex make sense?

@adambezecny `once` will always use a `RwLock`, but when `sync_writes = true` it will try to acquire the write-lock _before_ executing the function that returns the value to cache (and...

Would using RwLock instead of Mutex make sense?

Yes, correct!

Would using RwLock instead of Mutex make sense?

I think the difference might be because the macros expand to use `async_std`'s `async_mutex` and `async_rwlock` whereas the manual version you have is using `tokio::sync::RwLock`