| Age | Commit message (Collapse) | Author |
|---|
|
explicitly as we don't use any functionality of it.
|
|
|
|
the default value as intended
|
|
|
|
breaking anything this time
|
|
compiling of universal binaries
|
|
|
|
incarnations
|
|
|
|
|
|
-Add: Unix man file gets now installed if necessary
|
|
|
|
./configre --help instead of static strings
|
|
as expected
|
|
entry files
|
|
introduced options.
|
|
|
|
builds without asserts
|
|
without asserts
|
|
declarations in GCC 2.95, so only use if for GCC >= 3.
|
|
declarations.
|
|
|
|
a string, also use DEBUG instead of fprintf in one case.
|
|
|
|
As gcc is not necessary for FreeBSD compiles we 'just' use g++ as that support -dumpmachine.
|
|
(much) better chance of actually getting something useful out of OSX stacktraces.
|
|
remove -Wno-empty-body from the configure script
|
|
|
|
Thanks to TrueLight for finding the solution to this one
Note: if the shared-dir setting is already broken due to this then you have to manually run configure again
|
|
the search paths (based on patch by pv2b)
It is possible to set this to something else setting shared-dir with config
This works on other OSes as well but it's off by default (and not really tested)
|
|
in config.lib
This should increase readability and conflict solving
|
|
move the debugdumpcommand stuff out of the network 'area').
|
|
-Codechange: shuffle params a bit, sort params in save_params_array in the same order as the table above
|
|
-Codechange: silence a few warnings of ICC
|
|
default settings
-Change: [OSX] configure will no longer enable static by default
Turns out that Leopard can't handle static builds without the user installing extra libs
The tradeoff is that now people might have to enable static manually if moving the binary around
|
|
|
|
at compile time.
|
|
functionality.
|
|
building universal binaries
10.5 support needs this and it will not break 10.3 support
|
|
and handle broken iconv declaration in OSX 10.5.
|
|
never show them for this compiler
|
|
|
|
add --disable-unicode in configure help
|
|
Mercurial/HG instead of norev000 (when using a Mercurial/HG clone that is).
|
|
|
|
allowed value, making reconfigures to fail (special commit for Bjarni, now he should fix things too :p)
|
|
assumption that signed integers do not overflow, whereas it is very likely that some of the legacy code does so. It furthermore breaks with the assumed standard (the behaviour is not actually defined in the standard) about the behaviour when signed integers wrap.
|
|
configure parameters to save a long time ago. Now it is only causing warnings.
|
|
|
|
actual behaviour.
|
