Create/Wait NamedPipeA ?

The Ghost In The Machine ewill at
Wed Apr 25 12:31:13 CDT 2001

In, Dan Kegel
<dank at>
on Tue, 24 Apr 2001 16:25:56 -0700
<3AE60B84.91D68F96 at>:
>Martin Kroeker wrote:
>> Is anybody working on implementing CreateNamedPipeA/WaitNamedPipeA
>> (KERNEL32.168 , KERNEL32.725) or can anybody comment on how much
>> needs to be done to implement these ? Some applications seem to
>> use this to communicate with a license manager process - even
>> freely available product catalogs that apparently create a pro forma
>> license file during the installation process.
>I don't know, but from what it says at
>that call isn't available under Windows 95.  Do the apps
>you're talking about run only under NT, then?  Are any of
>them freely downloadable (doesn't hurt to ask :-)?

Just out of curiosity, would WinE have to implement CreateNamedPipeW
as well?  (Unicode pipe names -- ow, my brain hurts. :-) )

I'll also note that my copy of Borland C++ (4.51) has a declaration
for CreateNamedPipeA and CreateNamedPipeW -- and it looks like it's
not guardbanded against any specific Windows OS (how could I tell?).
The declarations are in bc45/include/winbase.h.

The most current Wine I have (a CVS update I did at the time of the last
release) has a CreateNamedPipeA implementation, a FIXME/stub, in
dlls/kernel/sync.c.  A CreateNamedPipeW implementation is just below it,
also a FIXME/stub.

Now I've got my curiosity up. :-)  What the hell *is* an NT named
pipe?  (I don't think it's quite the same as a *nix 'mkfifo()' call.
Last I looked at NT named pipes, it was more like a pipe cluster...)

Guess I'll have to reboot into That Other OS and fiddle with it.... :-/

>- Dan

ewill at -- insert random misquote here
EAC code #191       8d:06h:57m actually running Linux.
                    Linux.  When Microsoft isn't enough anymore.

More information about the wine-users mailing list