Rust let us abstract even file types (path to a fastq file, fasta file, annotations, etc) with no run time costs.
This eliminate many bugs at compile time.
You may say that we can get it in C too, and you will be correct. But in C we spend our time on herding pointers.
Research is given X money for N months (sort of), so we have time constraints on development time.
If we do bit wise work, the compiler tests our base types.
Not to mention multithreading just works.
Even big projects like BLAST had bugs that led to wrong results due C/CPP horrible multithreading.
We encountered two more tools that had similar bugs.
I think that if someone ever does a meta-studies of research code written in C it may get papers retracted.
We didn’t get similar run times with Haskell.
Rust let us abstract even file types (path to a fastq file, fasta file, annotations, etc) with no run time costs. This eliminate many bugs at compile time.
You may say that we can get it in C too, and you will be correct. But in C we spend our time on herding pointers. Research is given X money for N months (sort of), so we have time constraints on development time.
If we do bit wise work, the compiler tests our base types.
Not to mention multithreading just works. Even big projects like BLAST had bugs that led to wrong results due C/CPP horrible multithreading. We encountered two more tools that had similar bugs.
I think that if someone ever does a meta-studies of research code written in C it may get papers retracted.