Re: [kln2-devel] Problems installing...using make

[Marek W. <mar...@o2...>]

Mo,

On Monday 16 February 2004 21:53, Mo wrote:
> Not sure if this is a bug or if there is something I can do
> help ease the installation process.=20

This part comes from CVoiceControl. Yep, it is written in quite a bad=20
C. And generally C++ is much more strict than C. I changed a few=20
places in the source, but I stopped as soon as things started to work=20
for me. :(

Anyway, the places which ended as errors in your case, on my machine=20
just write out warnings. One day, yes, the source will have to be=20
fixed. But for now, I think it would be easier, to let your compiler=20
not to be so touchy.

Try '--disable-strict' option of configure.

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D

If this does not help, I'm not sure what to do. The command that in my=20
case successfully compiles the voicerecognition.cpp file is

if g++ -DHAVE_CONFIG_H -I. -I. -I.. -I/usr/include/kde=20
-I/usr/lib/qt3-gcc2.96/include -I/usr/X11R6/include   -
DQT_THREAD_SUPPORT  -D_REENTRANT  -Wnon-virtual-dtor -Wno-long-long=20
-Wundef -Wall -pedantic -W -Wpointer-arith
 -Wmissing-prototypes -Wwrite-strings -ansi -D_XOPEN_SOURCE=3D500=20
-D_BSD_SOURCE -Wcast-align -Wconversion -Wchar
-subscripts -O2 -fno-exceptions -fno-check-new -fno-common  -MT=20
voice_recognition.o -MD -MP -MF ".deps/voice_r
ecognition.Tpo" \
  -c -o voice_recognition.o `test -f 'voice_recognition.cpp' || echo=20
'./'`voice_recognition.cpp; \
then mv ".deps/voice_recognition.Tpo" ".deps/voice_recognition.Po"; \
else rm -f ".deps/voice_recognition.Tpo"; exit 1; \
fi

How is it different (except from libs/includes locations) from yours?

For example: there should NOT be '-pedantic-errors' or '-Werror'=20
options.

~Marek

PS. About the lines with errors:
voice_recognition.cpp:2467: error: ISO C++ forbids variable-size array=20
` prefetch_buf'
Actually, Mo, I think you have an old version of the source. Get the=20
new one from cvs (it is not much different though). Anyway, in the=20
new version it is line 2566. The problem is fixed by adding 'const'=20
to line 2564; it should look like:
const int  prefetch_N    =3D 5;

2612 and 2637:"`void *' in pointer arithmetic": we have an void* v1=20
pointer and an integer v2 counting how many bytes farther the memcpy=20
should copy some data; the warning is caused by 'v1+v2'. How to move=20
from a void* pointer by few bytes in a decent way? Anybody?

Anyway, there are LOTS of other warnings in other parts coming from=20
CVoiceControl, so it will need some work to fix them all. For now, I=20
hope --disable-strict will work.
M.

--=20

     \/   /|\  Marek Wieckowski
  #####  | | |
 =3D . . =3D  \|/  Institute of Theoretical Physics
    U      |   Warsaw University
  / ~ \___ |
 <|   |    |
  > . <    |   http://www.fuw.edu.pl/~wiecko
 <<___>>   |   http://klearnnotes2.sourceforge.net