___ ___ _ _ ___ ___ ___ ___ _ ___ _ _  ______  
_______________            _________ |_, |== |\| |== |-< |-|  |  | [_] |\| __/_/_______   
_____  ___/__(_)_____________  ____/____________ ___________________________ _______  /__
____  /_ ___  /__  ___/  _ \  /    _  __ \_  __ `__ \  __ \  __ \_  __ \  _ \_  __ \  __/
___  __/ __  / _  /   /  __/ /___  / /_/ /  / / / / / /_/ / /_/ /  / / /  __/  / / / /_  
  /_/     /_/  /_/    \___/\____/  \____//_/ /_/ /_/_/ __/\____//_/ /_/\___//_/ /_/\__/  

                  (c) Lada 'Ray' Lostak (c) Orcave (c) 1996-2004

Legend:

    [date]
    : coment (release, ...)
    % major feature adding
    + feature adding
    > improvment   
    = major/important bugfix
    - bugfix
    ! interface changed - adding only, minor version was updated
    * interface changed - function names, parametres, major version update -> source recompile required 

    additionals:

    [WIN]       - windoze only related


[2005-03-20]
% Started Kju
> 'source tree' dance (FcXml remaded back to Xixit :)

[2005-03-10]
+ real unicode support (UTF8 UTF16 UTF32)
+ new codepages
+ converts & direct upgrading datas from unicode.org

[2004-10-02]
> Xixit moved into FC, renamed to FcXml

[2003-11-16]
> text stream divided into 'stream' and memory format text functions
> general string functions (e.g. fcsmkansi etc.) renamed to 'str*' and coresponding to 'fcs/ans'


[2003-10-21]
% added buffer lengths to ALL string and filesystem functions/interfaces
> return values of string functions returns 'error code' not 'destination string'
- various local buffer overflows accross whole libraries due adding buffer lengths
- various local buffer overflows in fcsPrintEx/fcsPrintExArg

[2003-06-18]
+ PE loader now returns section memory pointer, if module loaded (caller have to care about releasing library by itself)
= while attaching virtual stream to physical stream with related seek root, current position was changed to 2*start
  instead start (0 from vstream p.o.v.)
- FC resource locker uses 'size in file' instead of RAW size for resource section
- [WIN] file stream doesn't call properly Clone (incorect system path flag used)
- string list after commit read overwrites 'loaded' strings (strAdd required properly sets LastFreeIds)


[2003-03-04]
- some 'detach' style function changed error code to 'OK' state (removed, they only return)
- fixed error code remebering (sma error can be remeberd more times to one slot, even it is negative)
- while attaching to NULL datas arrayManager doesn't properly initialize alloc step (-> bad memory operations after
  writing new items)

[2003-02-23]
! added features to fcStrCommandLine
 
[2003-02-07]
! added strPush/strPop at fcStrStringList
* changed fcDump* dumping objects

[2002-11-06]
= asmI32_InterlockedCompareAndExchange changes EBX register, even it have __cdecl calling convention. this
  causes various access violation or other fall downs at totally different parts (like EnumWindows ansi
  wrapper for example)

[2002-11-01]
% memory heap is thread-safe
% console output is thread safe
= thread manager was isnstable during releasing, due low level thread releasing troubles
- [WIN] low level 'get current thread' doesn't work before initialization was done and cause problems with call
  ansi versions of API's
- self tests-heap moves 'realoc' blocks after first realloc to 'alloc/free' only

[2002-10-07]
= some race conditions in sync objects (fully remaded, including 'sysdep' part), speed increasing

[2002-09-30]
> performance increase with event/semaphore/critical_section
= mutex sync object doesn't relocate owning to wait thread
! fcThreadList::thlTest - test if given thread present in wait queue

[2002-09-25]
% sync objects were storing stack-related pointers to lists - it causes hang up sometimes
+ new/delete operator debug pair test now shows real source file/line, not 'FcMemory.h'
+ object queue with priority - fcMemObjQueue
- Mutex mtxRelease doesn't call 'Release' to itselfs and causes component leacks
! fcThreadList allows storing 'user datas' not only pointer
* fcThreadList

[2002-09-24]
+ added memory tests
= internal thread 'OpenId' remaded to thread table index, because 'get current thread' doesn't work properly
= mutex sncWaitFor keeps guard in 'wait' state -> deadlocks
= all sync objects shoudl fall down if immediateely after 'wait' was called its 'release' - they do not
  addref/release themself before resuming waiting threads
= on single-cpu machines thread never starts at some conditions
- attach to current thread fall down at some conditions (mainly OpenId realted)
- typos in fcStrConFile methods
- language mistakes

[2002-08-12]
: started history track finally
+ created history fix/feature log
- ANSI console output render under NT fixed (national chars/frames)