I'll share some fun issues I encountered in the making of this.
When adding byte literals, the number `10` specifically did not work. Even though all other numbers did. Turns out; some component tries to be clever and automatically converts a file with just `\n` into `\r\n`, and as `10` is the ascii code for `\n` it would cause "memory" corruption.
At one point I got rate limited by the docker registry. This was because when using `docker:latest` it would check if there was an update for that image each time an allocation was made.
From Haskell/Lisp I've been taught that linked lists are the superior data structure for representing strings. Thus; strings in doiter spawn one docker container for each character. Unsurprisingly this affects performance quite heavily as `hello-world.dr` takes over a minute to complete.
> How good is the performance?
> Currently samples/hello-world.dr takes about a minute to run on my machine. With the state of the industry, this isn't too bad in comparison
Hey, atleast it's better than Java!
I'll share some fun issues I encountered in the making of this.
When adding byte literals, the number `10` specifically did not work. Even though all other numbers did. Turns out; some component tries to be clever and automatically converts a file with just `\n` into `\r\n`, and as `10` is the ascii code for `\n` it would cause "memory" corruption.
At one point I got rate limited by the docker registry. This was because when using `docker:latest` it would check if there was an update for that image each time an allocation was made.
From Haskell/Lisp I've been taught that linked lists are the superior data structure for representing strings. Thus; strings in doiter spawn one docker container for each character. Unsurprisingly this affects performance quite heavily as `hello-world.dr` takes over a minute to complete.