David Stacey writes:
Same here. I maintain a local mirror and use my own setup.ini files
(created with genini); these do not have 'sig' files. Directory
structure is as follows:
cygwin-2015-10-16
/cygwin
/x86
/x86_64
/cygwinports
/noarch
/x86
/x86_64
The 'setup.ini' files are located in the 'x86' and 'x86_64'
directories; setup is pointed at the 'cygwin-2015-10-16'
directory.
If that is the complete directory structure at those two levels then I
don't see how it can't work. I'm doing exactly the same, save for the
exact directory names. I'm using the "-mX" switches when invoking
setup. If you have an x86 or x86_64 directory under cygwin-2015-10-16
however, then setup would never look any further (you can't nest mirror
directories).
Setup does essentially this (and finds the following setup files on my
system when it gets started in the top-level mirror directory, provided
that $arch="x86"):
$ find -mindepth 1 -maxdepth 1 -name $arch | \
xargs -I: sh -c 'for f in :/setup.{xz,bz2,ini}; \
do [ -e $f ] && { echo $f ; break; } done' && \
find -mindepth 2 -maxdepth 2 -name $arch | \
xargs -I: sh -c 'for f in :/setup.{xz,bz2,ini}; \
do [ -e $f ] && { echo $f ; break; } done'
./cygport/x86/setup.bz2
./cygwin/x86/setup.bz2
./maint/x86/setup.xz
./patch/x86/setup.xz
./perl/x86/setup.ini
What's the result you're getting?