|  0.10 | BSD | 2009-09-03 07:11 | Select | 
| 
  - fixed a bug that treated null as mapping values as invalid (reported by Philipp Hancke)                                                                           \r\n- fixed a bug that parsed some escape sequences wrong (reported by Philipp Hancke)\r\n
 | 
    |  0.9 | BSD | 2009-06-16 08:32 | Select | 
| 
  - fixed a bug that could cause a return value\r\n      to be truncated on 64bit machines. This happened\r\n      only in case of mal-formatted json and could lead\r\n      to wrong error message\r\n    - integer/float parsing now supports higher precision\r\n      types on architectures that support it (i.e. where\r\n      INT_TYPE/FLOAT_TYPE is bigger).\r\n    - fixed a bug in the json rendering that did not error\r\n      when trying to encode strings with characters outside\r\n      of unicode. instead it produces malformed json.\r\n    - working configure-checks for the availability of\r\n      wchar
 | 
    |  0.8 | BSD | 2008-10-02 06:33 | Select | 
| 
  - fixed a bug in the string escpape function                                                                                                                       that potentially produced wrong json. shame                                                                                                                      on me
 | 
    |  0.7 | BSD | 2008-09-25 23:27 | Select | 
| 
  - added parse_utf8()                                                                                                                                               \r\n- added render_utf8()                                                                                                                                              \r\n- rewrote the render-function inside the cmod                                                                                                                      \r\n- the parser is more strict about unicode now,                                                                                                                     \r\n  hence its not possible to send anything but                                                                                                                      \r\n  unicode characters in strings.
 | 
    |  0.6 | LGPL | 2008-02-22 08:14 | Select | 
| 
  - uses swscanf for number parsing in case sizeof(p_wchar2) == sizeof(wchar_t)                                                              instead of copying in to a char* and parsing using sscanf.
 | 
    |  0.5 | LGPL | 2007-12-01 22:48 | Select | 
| 
  fixed a memory leak
 | 
    |  0.4 | LGPL | 2007-04-21 17:06 | Select | 
| 
  - changed the array and mapping parser from using                                                                                 \r\n  the pike stack to collect values (stupid me) 
 | 
    |  0.3 | LGPL | 2007-04-16 08:20 | Select | 
| 
  - parser wasnt strict enough\r\n- fixed some output formating
 | 
    |  0.2 | LGPL | 2007-03-21 10:16 | Select | 
| 
  - added render()\r\n- added validate()\r\n- added extra types for null, false and true.\r\n- some documentation (hopefully)
 | 
    |  0.1 | LGPL | 2007-03-13 06:20 | Select | 
| 
  initial version
 |