Mail Archives: djgpp/2015/06/06/07:28:09
On 06/06/2015 10:07 AM, Eli Zaretskii (eliz AT gnu DOT org) wrote:
>> Date: Sat, 06 Jun 2015 08:06:53 +0300
>> From: "Andris Pavenis (andris DOT pavenis AT iki DOT fi)" <djgpp AT delorie DOT com>
>>
>> As the result I would vote for removal of excluding additional macros when __STRICT_ANSI__ is defined.
> What about -pedantic -- do we still want it to flag any symbols that
> are not in the standard?
>
>
They do not get flagged dor Linux or Mingw (as far as I checked with Linux to Mingw
cross-compiler). Do we really need to be more pedantic? (especially with our limited resources)
Change which I committed to trunk fixes current problems with C++ by defining macros when use of
corresponding C++ standard is being requested. We have been living with absence of these macros for
C when __STRICT_ANSI__ is defined for many years.
So I could commit same change also for v2.05 and build a release (lets name it release candidate
for now as actual release would require rearranging files between djgpp/beta, djgpp/current and
djgpp/deleted)
After that:
1) we could leave things as they are
2) if we can recognize that standard permits defining these constants when __STRICT_ANSI__ is
defined then we could remove excluding #define statements
I would prefer variant 2 as it
a) would make maintenance easier for little resources we have
b) it would make life easier for users (I mean users who are developing own software using DJGPP).
c) it would make our implementation slightly nearer to ones of Linux and Mingw as currently only
DJGPP hides these macros when __STRICT_ANSI__ is defined
Andris
- Raw text -