Boscop
Boscop
I've been using FFTS until now, [The Fastest Fourier Transform in the South](https://github.com/linkotec/ffts) (it's faster than FFTW), specifically I've been using the 1d real function for real-to-complex FFT: https://github.com/linkotec/ffts/blob/2c8da4877588e288ff4cd550f14bec2dc7bf668c/include/ffts.h#L96 >...
Maybe it makes sense to take inspiration from FFTS? https://github.com/linkotec/ffts It's faster than FFTW. The `ffts_init_1d_real` function does exactly what we need: https://github.com/linkotec/ffts/blob/2c8da4877588e288ff4cd550f14bec2dc7bf668c/include/ffts.h#L96
It should work with this fork: https://github.com/Eh2406/preferences-rs/tree/patch-1
Nice. Btw, is this crate still being maintained? :)
I can't use this crate for my types which are using `#[derive(Serialize, Deserialize)]` from serde 1.0. I have to use serde 1.0 because of tarpc, diesel etc. Can you **please**...
Yes, please!
They are still separate crates, but share a [workspace](https://github.com/djc/askama/blob/master/Cargo.toml), but yes, I also recommend doing that.
It could also be done by calling `.to_string()` in the macro before constructing the `PursModule`. Then `PursModule` wouldn't need to track that lifetime.
Thanks! Sorry, I hadn't seen that :)
Hm, actually, no, because I can't use a diesel.toml file for this use case. There should be a way to specify `--only-tables` and `--except-tables` with diesel_cli_ext on the command line.