- separation of some resources that only existed in English
- lesser adjustments in menus

  ntdll counterparts
- {Enter|Leave}CriticalSection is now a pure forward from kernel32 to
  ntdll (we now longer can use {Enter|Leave}CriticalSection in kernel32)
- replaced a few kernel32 heap management calls from ntdll, with
  RtlHeap* equivalents

- remove warning
- preparations to use multiple fdi_decomp_state structures in a linkedlist,
  which will be used to implement split cabinets, if all goes according
  to plan -- this is somewhat analogous to struct cabinet in cabextract.

- Fixed bugs in AVISaveOptions dialog handling.
- Fixed bug in AVIFILE_LoadFile -- now check for more than
  MAX_AVISTREAMS streams in file.
- Implemented AVIMakeFileFromStreams and AVISaveVA method.
- Added IAVIEditStream interface.
- Implemented EditStream* methods.
- Added stubs for AVISaveVW, CreateEditableStream.
- Added stubs for clipboard handling.

event, so that we get correct behavior in synchronization handling.

Change DPMI asynchronous event handling.
Always use alternate stack in DPMI relays and check for pending events
after original stack has been restored.

- faster Blt to the framebuffer using texture engine

proper size, and were partially visible.

- added stubs for GetSaveFileNamePreview{A,W}
- removed invalid VIDC.MRLD line from system.ini

  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