- all uses of dgettext() and friends are already protected by
appropriate ifdef's, no need to provide dummy defines here
- setlocale() and <locale.h> are required by C89, C99 and POSIX .. assume
its there and if not, one can disable the whole thing with --disable-nls
- just two places where needed, dont pollute system.h needlessly
- include depending on HAVE_MMAP instead of separately checking for
sys/mman.h, if sys/mman.h doesn't exist or is broken HAVE_MMAP wont be set
- the AC_HEADER_TIME check is unnecessary for any remotely recent systems,
and the HAVE_SYS_TIME_H conditional in system.h is just broken as we
dont even check for <sys/time.h> header in configure
- dont include from system.h, the time.h and sys/time.h get already
included through our public headers where necessary
- vprintf(), vsnprintf() and snprintf() are in C99, no point checking
especially as we dont have fallbacks for them
- no point testing for inline capability as we dont try to work around
- whole lotta unnecessary wankery here (what have I been thinking)
- avoid using the non-standard strndup(), we checked for it in configure
but didn't provide an alternative so checking was just bogus
- move the includes out of system.h, not commonly needed
- <poll.h> is conditional as we actually provide a fallback through select(),
but for <sys/select.h> missing there's no fallback so doesn't make
much sense to test for (and both poll.h and sys/select.h are posix anyway...)
- only the lua posix extension "uses" this by providing putenv()
to Lua, make it conditional and return error if not supported by
the underlying operating system
- print chanced from statement to a function in python 3, for our
purposes sys.stdout.write() is sufficient and avoids dealing with
the incompatibilities between Python 2.x and 3.x
- suggested by David Malcolm / Ville Skyttä
- what we support now in payloads is XZ, not the former LZMA format, rename
user- and header-visible parts to match this
- rpmlib(PayloadIsXz) dependency versioned 5.2-1 to avoid unnecessary
incompatibility with what rpm5 has and what Mandriva is already using
- only provide the rpmlib dependency if actually built with xz support
- rpmFileIsCompressed() attempts to guess the magicless old lzma format
by file suffix, so it can handle both right now
- only needed by the testsuite
- optimally testsuite should skip parts needing fakechroot if it's not
there but at least this'll leave a trace into rpmtests.log...
- the previous implementation was pretty busted up, cap_size() is the
size of external representation, not sizeof(*cap_t) as pointed out by
Andrew Morgan
- check for cap_compare() in configure and use it if available (it's
a Linux-specific extension to the posix draft and only very recent
libcap has it)
- if cap_compare() isn't available, use internal variant that does it
the hard way by converting the capabilities to external presentation
and comparing those
- if built with --with-acl, check that files and directories don't have
any extra acl's set
- for now, any acl beyond the regular unix permission set is reported as
file mode difference as the acl's cannot have been set by rpm itself
- patch from Andreas Gruenbacher, modified to use libacl instead of raw
xattrs for portability, BUT atm this uses non-portable acl_equiv_mode()
Linux libacl extension, the posix draft doesn't seem to have much in
the way of comparing entries :-/
- add minimal bits and pieces to check for capabilities in files on verify
- for now, any capability set is a verification failure as the capability
cannot have been set by rpm itself
- patch from Andreas Gruenbacher, modified to use libcap instead of raw
xattrs for portability
- build-time generated rpmints.h causes issues with distcheck and is
not really worth the trouble
- we have all sorts of legacy junk to deal with, sticking them all into
rpmints.h would be just wrong...
- don't build internal copy of Lua
- move 3rd party extensions (posix and rexlib) to toplevel luaext/
directory, built by default (unless --without-lua specified)
- auto*foo checks for external Lua
- minimal tweaks to lposix.c and rpmlua.c to get them build with Lua 5.1
- tarball down to almost 1/4 of earlier size
- check if internal db is present in configure, point to INSTALL if not
- distcheck needs --with-external-db now to have any chance at working
- leaving to hg for now for easy reversal if getting second thoughts ;)
HAVE_LOCALE_H is used by system.h, ensure it is defined properly
(the issue only occurs when compiling without __OPTIMIZE__ (ie -O2)
otherwise libintl.h do include locale.h)
Patch from Pascal Rigaux
Florian Festi