well, I don't get the difference, we have a bunch of standards, my limited understanding is that now these .org's have joined forces, and everything gets messy, at least for me.
for instance, APE has lot's of headers with the feature test macro _POSIX_SOURCE, which has been superseded by _POSIX_C_SOURCE, which should be the same as _XOPEN_SOURCE == 600, which in turn is compliant with ISO C! so, I really don't get this dance, anyways there is a contrib package in my sources dir "posix-man" with the "posix"/xopen/ieee man pages, sections 0 and 3. http://www.opengroup.org/onlinepubs/009695399/functions/xsh_chap02_02.html On Feb 4, 2008 11:26 PM, Pietro Gagliardi <[EMAIL PROTECTED]> wrote: > Some other reasons: > - Some systems (read: Linux) do not have pthreads > - What if I said I'm running GCC on Microsoft Xenix? Is that POSIX- > compliant? > - Curses is not POSIX. It's Single Unix Spec, though. > - C99 is still new and although it's in POSIX, not many systems have > it (Plan 9 doesn't have complete C99) > > > On Feb 4, 2008, at 9:23 PM, David Arnold wrote: > > > On 03/02/2008, at 8:29 PM, Lyndon Nerenberg wrote: > > > >> Autoconf is nothing but a stinking rotten corpse that lives only > >> because the cult of GNU adherents cannot (no, refuse to) grok the > >> concept of POSIX. > > > > the problem with POSIX is that it doesn't specify enough. > > > > for instance, if you have to write some code to list the network > > interfaces on a (*nix) machine, you have some that provide a > > specific function to do so (getifaddrs), some where you should use > > SIOCGIFCONF, another where SIOCGLIFCONF is better and one where > > your best bet is to hope the /proc filesystem is mounted and read > > from that. > > > > POSIX doesn't help for things like this. and autoconf, for all its > > failings, does. > > > > > > > > d > > > > -- Federico G. Benavento
