Mail Archives: cygwin/1998/07/19/23:55:46
Douglas Steele <Douglas AT dsteele DOT demon DOT co DOT uk> writes:
>
> In message <Pine DOT SUN DOT 3 DOT 93 DOT 980717095438 DOT 27332A-100000 AT modi DOT xraylith DOT wisc.
> edu>, Mumit Khan <khan AT xraylith DOT wisc DOT edu> writes
> >Douglas, why not just use the cygwin32 version? All you need is windres
> >and cygwinb19.dll, right? (of course I haven't personally tested this,
> >so don't know for sure).
> Tried it.
> First checked windres worked OK on full cygwin32 installation.
> Then replaced Mingw32 windres with cygwin32 windres and cygwinb19.dll
> ... same problem, rc file dumped to screen.
> As far as I can make out, its the Mingw32 cpp.exe thats the common
> factor here.
I just checked out whether I could cross-compile the R-0.61.3 sources
from Linux, and lo and behold:
$ PATH=~/mingw-cross/i386-mingw32/bin/:$PATH make
windres -i wincons.rc -o res.o
windres:
/home/pd/mingw-cross/lib/gcc-lib/i386-mingw32/egcs-2.90.27/include/stdarg.h:70:
parse error
cpp: output pipe has been closed
make: *** [res.o] Error 1
Apparently, windres balks at typedef constructions in stuff included
via <windows.h>. The problem could well be just that w95 is not
handling the death of the receiving end of a pipe gracefully, but the
real solution is to prevent it from dying in the first place. If you
include <Windows32/Defines.h> instead of <windows.h>, all seems to be
well [as proven by my standard torture test: It starts up and runs
plot(1:10)...].
--
O__ ---- Peter Dalgaard Blegdamsvej 3
c/ /'_ --- Dept. of Biostatistics 2200 Cph. N
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~~~~~~~~~ - (p DOT dalgaard AT biostat DOT ku DOT dk) FAX: (+45) 35327907
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".
- Raw text -