RPM Community Forums

Mailing List Message of <popt-devel>

POPT 2.0: C89 or C99?

From: Jeff Johnson <n3npq@mac.com>
Date: Thu 17 Jun 2010 - 18:38:18 CEST
Message-id: <235B4247-2A09-4E82-8EDC-C6EC17EFA4D7@mac.com>
There are various tedious "portability" issues in need
of sorting out, particularly if I'm gonna start chasing
more aggressive functionality like pattern matching
and mutexes/threads/C++/SWIG "stuff" in POPT 2.0.

The names of integers is always the silly start of
the discussion. Up till now I've largely targeted
C89 (and haphazardly verified conformance if/when
issues were reported).

I personally don't care whether I type
	unsigned long long
or
	uint64_t
(I prefer "uint64_t" if that matters. My fetishes don't matter ...)

What I do care is "consistency" in POPT, and clearly stated goals.

So I'm gonna assume
	#include <stdint.h>
(with typedef retrofits for the C99 challenged because that's easy
to fix with AutoFu if needed)

OTOH, I will continue adding declarations in the usual places unlike C99
(because that's a much more invasive change that cannot be fixed with AutoFu).

I also see little use for other C99 arcana like bool etc and so will avoid
using.

The above is just an statement of de facto coding practice in POPT, but
now clearly and explicitly stated.

73 de Jeff
Received on Thu Jun 17 18:38:37 2010
Driven by Jeff Johnson and the RPM project team.
Hosted by OpenPKG and Ralf S. Engelschall.
Powered by FreeBSD and OpenPKG.