handles as wineserver handles
- console input handle object is no longer waitable (input record
  synchronisation is now implemented as a simple semaphore), and removed
  FD_TYPE_CONSOLE from fd types in wineserver
- console handles now always have their two lower bit set so one can
  distinguish a console handle from a kernel object handle
- implemented some undocumented kernel32 console related APIs
  (CloseConsoleHandle, GetConsoleInputWaitHandle, OpenConsoleW,
  VerifyConsoleIoHandle, DuplicateConsoleHandle)
- allowed a few kernel32 APIs to take console pseudo-handles
  (FlushFileBuffer, GetFileType, WaitFor*Object*)
- simplified the console inheritance at process creation
- in console tests, no longer create a console if one already exists

and IMAGE_SCN_CNT_INITIALIZED_DATA flags set together.

IMAGE_SCN_CNT_UNINITIALIZED_DATA and IMAGE_SCN_CNT_INITIALIZED_DATA
are set.

- fixes to the http protocol

Shearman).

compilation issues.

Fix asynchronous event locking.
Make asynchronous event handling support DPMI.

main() {return sizeof int;}'). Use 'sizeof(type)' instead.

- added some console related tests to process creation
- added a few more test cases to the ntdll environment tests

  the other one in environment variable expansion)
- the process parameters, when passed thru wineserver, are now fully
  handled in ntdll, they are stored in the RTL_USER_PROCESS_PARAMETERS
  structure.
- later on in kernel32 loading sequence, those parameters are copied
  into STARTUPINFO shadow structures
- later modification to those paramters are now reflected to the
  RTL_USER_PROCESS_PARAMETERS structure (and STARTUPINFO is kept
  untouched) (for example, StdHandle setting) (Win 2k behaves like this)
- ENVDB has been removed
- command line inheritance (from unix command line) is now purely in ntdll
- all kernel32 environment functions now rely on their ntdll counterparts
- goodies: input/output handle inheritance while asking for a detached
  console is better handled; a few more kernel32 environment tests now
  pass ; silenced a valgrind warning in process creation

  IDirectMusicLoader8Impl::LoadObjectFromFile
- don't use IDirectMusicSegment*Impl when IDirectMusicSegment*8Impl
  can be used instead (avoid code duplication)

- more traces

_thiscall functions.
Fixed some bugs, and added a few missing functions.

  between cabextract.c and fdi.c
- reminders to eliminate global variables (for multithread
  compatibility)
- remove struct fdi_cab: due to the nature of the FDI API, we cannot
  preload all the cabinets; this appears to obviate the need for struct
  fdi_cab
- "oppress" (that is, do not process) partial files which were
  continuations from another cabinet
- more than one partial file can exist in a single cabinet (how!?) --
  so move the partial file notification (and "oppression" that goes with
  it) into the loop that iterates through files

retrieval of their interfaces via GetObjectInPath.
Fix and clean dsound and dmusics objects allocation when initializing
the performance object.
Some stubs improvements.

  faster on my NVIDIA hardware)
- small optimizations

  widechar classes.
- fix a stupid bug in callback handling caught by Lionel.
- add some forgotten IID on class factory (better if we want to active
  the code)
- many cleanups on COM Macros (don't need A and W variants as they are
  the same)

patch from Roderick Colenbrander.

- GetDC does not lock in ReadOnly mode as it leads to problems with
  some D3D optimizations
- added a fast Back => Front / Front => Back Blt fast path

- Make it more closely resemble the native control
- Add focus rectangle
- Misc fixes and cleanups