Great, for the most part...
I'll have to download the DX8 headers and give them a try...
John
----- Original Message -----
From: "Brian S. Julin" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Saturday, June 30, 2001 4:32 PM
Subject: DirectX cygwin build
>
> OK, here are the results of trying to build current CVS on
> cygwin with DirectX 8.0a (i built and installed perl-5.6.1
> and autogen works.)
>
> 1) It was necessary to both link c:/mssdk/include to the cygwin
> root directory AND to specify --with-extra-includes=/mssdk/include,
> either/or would not do the trick.
>
> 2) For some reason I had to cd to libgg and make install before
> building the rest of libgii
>
> 3) input-tcp had to be disabled, due to an undefined symbol
>
> 4) There are compile errors in input-directx due to some code
> in directx.h:
>
> In file included from /usr/include/w32api/windows.h:110,
> from dxinput.h:30,
> from input.c:29:
> /usr/include/w32api/winnls.h:411: warning: redefinition of `LONG_PTR'
> /mssdk/include/basetsd.h:94: warning: `LONG_PTR' previously declared here
> In file included from dxinput.h:33,
> from input.c:29:
> /mssdk/include/dinput.h:1285: redefinition of `DIDEVICEINSTANCE'
> /mssdk/include/dinput.h:1275: `DIDEVICEINSTANCE' previously declared here
> /mssdk/include/dinput.h:1288: redefinition of `LPCDIDEVICEINSTANCE'
> /mssdk/include/dinput.h:1286: `LPCDIDEVICEINSTANCE' previously declared
here
> make[3]: *** [input.lo] Error 1
>
> The code being:
>
> 1279 typedef const DIDEVICEINSTANCEA *LPCDIDEVICEINSTANCEA;
> 1280 typedef const DIDEVICEINSTANCEW *LPCDIDEVICEINSTANCEW;
> 1281 #ifdef UNICODE
> 1282 typedef DIDEVICEINSTANCEW DIDEVICEINSTANCE;
> 1283 typedef LPCDIDEVICEINSTANCEW LPCDIDEVICEINSTANCE;
> 1284 #else
> 1285 typedef DIDEVICEINSTANCEA DIDEVICEINSTANCE;
> 1286 typedef LPCDIDEVICEINSTANCEA LPCDIDEVICEINSTANCE;
> 1287 #endif // UNICODE
> 1288 typedef const DIDEVICEINSTANCE *LPCDIDEVICEINSTANCE;
>
> 5) For some reason input-linux-mouse is built.
>
> 6) LibGGI built fine also using 1) and 3) configure options
> (of course, cannot test until input-directx also builds.)
> Except wrap.c tried to build despite nonpresence of IPC.
>
> --
> Brian
>
