gnupg tries to build with non-existent libusb

[David Shaw]

On Sat, Jan 01, 2005 at 10:25:33PM -0800, Mike Castle wrote:
> Strange goings on.  Tried this with 1.4.0, 1.9.13 and 1.9.14.  Results all
> the same.
> 
> if i486-linux-gcc -DHAVE_CONFIG_H -I. -I../../gnupg-1.9.14/scd -I..  -I../../gnupg-1.9.14/intl -I../../gnupg-1.9.14/common -DLOCALEDIR=\"/usr/share/locale\" -DGNUPG_BINDIR="\"/usr/bin\"" -DGNUPG_LIBEXECDIR="\"/usr/libexec\"" -DGNUPG_LIBDIR="\"/usr/lib/gnupg\"" -DGNUPG_DATADIR="\"/usr/share/gnupg\""           -I/usr/include -g -O2 -Wall -MT ccid-driver.o -MD -MP -MF ".deps/ccid-driver.Tpo" -c -o ccid-driver.o ../../gnupg-1.9.14/scd/ccid-driver.c; \
> then mv -f ".deps/ccid-driver.Tpo" ".deps/ccid-driver.Po"; else rm -f ".deps/ccid-driver.Tpo"; exit 1; fi
> ../../gnupg-1.9.14/scd/ccid-driver.c:85:17: usb.h: No such file or directory
> 
> Hmm... I'd installed libusb at sometime in the past, and removed the header
> files, but forgot the remove the libraries.
> 
> Configure should probably check for a working usb.h in addition to libusb
> (guess this might qualify for a situation where someone forgot include
> libusb-devel or whatever on their system).

It shouldn't.  The people who package the xxx-devel packages generally
expect this sort of problem and distribute e.g. libusb-0.1.so.4
rather than libusb.a or libusb.so.

Nevertheless, I'm going to improve the libusb check anyway.  It
certainly doesn't hurt to do a test compile before enabling libusb.

David