Name: Anonymous 2012-07-22 10:51
I mean, today I was trying to write a very simple parser for some very simple markup. The optimal code wouldn't need to traverse the markup more than once to parse it. It is that simple. Yet I couldn't find a simple way to do it in JavaScript. Almost all string methods mean that I'd need to traverse the string at least two or three times. Fine, fuck those methods. I resolved to traverse it "manually" character after character using a queue and a stack to keep track of things. Well apparently there aren't any decent data structures provided. Well fuck. I know I can probably find a decent third party library but I hate working in such an enviroment.
However, here's the thing. I still love Node.js and asynchronous programming. It's really ironic. JavaScript on a whole is a very inefficient POS but the basic premesis of Node.js is very elegant and simple.
So I resolved to do either of these.
1. Ditch Node.js altogether and use Erlang.
2. Keep Node.js as the final glue-code and use C or Haskell for anything other than the very basic stuff.
What would /prog/ suggest?
However, here's the thing. I still love Node.js and asynchronous programming. It's really ironic. JavaScript on a whole is a very inefficient POS but the basic premesis of Node.js is very elegant and simple.
So I resolved to do either of these.
1. Ditch Node.js altogether and use Erlang.
2. Keep Node.js as the final glue-code and use C or Haskell for anything other than the very basic stuff.
What would /prog/ suggest?