Sjoerd Visscher
JSON is a UNICODE format, like any modern format is today. ByteStrings are not going to work.
Well, neither is String as used in the code I responded to. I'm not intimately familiar with JSON, but I believe ByteStrings would work on UTF-8 input, and both ByteString and String would fail on UTF-16 and UTF-32.
If everybody starts yelling "ByteString" every time String performance is an issue, I don't see how Haskell is ever going to be a "real world programming language".
Insisting on linked lists of 32-bit characters isn't going to help, either. I'm also looking forward to a fast, robust, and complete UniCode support, but the OP asked about performance. -k -- If I haven't seen further, it is by standing in the footprints of giants