-------------------------------
2004-10-11 : Release 2.9.0-1.19
-------------------------------

The C++ Perfometer 2.9.0-1.19 Release is now available on the C++ Perfometer sites:
* http://sourceforge.net/projects/cpp-perfometer/
* http://alexvn.freeservers.com/s1/perfometer.html
* http://www.planet-source-code.com/vb/scripts/ShowCode.asp?txtCodeId=3...

Note. Release number N1-N2 consists of :
* N1 : Version of C++ Perfometer itself; current version is 2.9.0
* N2 : Version of Testsuites; current version is 1.19

1. Perfometer changes from 2.8.0 to 2.9.0
   --------------------------------------
   The Perfometer has been updated to be compiled with the following compilers :
   * GNU gcc/gpp 3.4.1 (Djgpp).

2. Testsuites changes from 1.18 to 1.19
   ------------------------------------
   2.1. Updated testsuite "Access to an element", file t_access.cpp;
   2.2. Updated testsuite "Comparing function objects to function pointers", file t_trd4.cpp.

   List of all testsuites
   ----------------------
   - access to an element
   - argument passing
   - argument passing (built-in types)
   - argument passing (containers-1)
   - argument passing (containers-2)
   - argument passing (containers-3)
   - calling functions : templated vs. non-templated
   - container performance (Stepanov & Stroustrup)
   - copying files : input to output
   - dynamic_cast vs. switch
   - endl vs. "\n", '\n' [! stdout/stderr > tmp]
   - exception handling vs. if-statement
   - find method vs. find algorithm
   - for-loop vs. for_each
   - for_each vs. transform
   - iterators
   - multiple inheritance vs. single inheritance
   - primitive operators
   - pthreads
   - reading contents from file into one string
   - recursion vs. iteration vs. accumulate
   - str-functions vs. mem-functions (C)
   - string initialization (C vs. STL)
   - string length (C vs. STL)
   - string processing (C vs. STL)
   - stub tested functions
   - tech report : D1 - overhead of class operations
   - tech report : D3 - the Stepanov abstraction penalty benchmark
   - tech report : D4 - comparing function objects to function pointers
   - tech report : D5 - measuring the cost of synchronized I/O
   - virtual vs. ordinary methods

3. Size of sources
   The C++ Perfometer 2.9.0-1.19 contains 42 program files (*.h, *.cpp)

      --- Summary Report (numbers) ---
      =====================================
      |               :   Lines :   Chars |
      |-----------------------------------|
      | Code Only     :   11451 :  268610 |
      | Code+Comment  :     252 :       - |
      | Comment Only  :    3823 :  135714 |
      | Empty         :    4232 :   67418 |
      |...................................|
      | * Total       :   19758 :  471742 |
      =====================================

      --- Summary Report (percents) ---
      =====================================
      |               :   Lines :   Chars |
      |-----------------------------------|
      | Code Only     :  58.0 % :  56.9 % |
      | Code+Comment  :   1.3 % :       - |
      | Comment Only  :  19.3 % :  28.8 % |
      | Empty         :  21.4 % :  14.3 % |
      |...................................|
      | * Total       :   100 % :   100 % |
      =====================================

   Note. Size of the Perfometer sources has been counted with using cncc - "Counter of C/C++ code"
   http://groups.google.com/groups?selm=bvtt1p%2410rbd7%241%40ID-79865.n...

--
   ==========================
   Alex Vinokur
     email: alex DOT vinokur AT gmail DOT com
     http://up.to/alexvn
   ==========================