Daniel Berlin wrote:
> On Mon, 2005-04-11 at 09:47 -0400, Michael Sweet wrote:
>
>>Max Bowsher wrote:
>>
>>>NormW wrote:
>>>
>>>
>>>>Greetings All,
>>>>Recent 'builds' of SVN 1.2-rc1 for NetWare fail with the following
>>>>'errors'. The first one (fs_fs.c) has been 'updated' in trunk, but I
>>>>note the attached log reference to a GCCism (Revision 13954), which I
>>>>take to understand in a negative light.
>>>>As the list below shows, there are further instances of that construct
>>>>still extant in the files noted, if they also need to be attended.
>>>
>>>
>>>Ah, thanks!
>>>
>>>For reference, we aim to be compilable on any compiler supporting plain
>>>C89 - though an awful lot of people do use GCC, so GCCisms occasionally
>>>creep in.
>>>Fixing now.
>>
>>FWIW, I added the GCC "-ansi" option to the ESP Ghostscript configure
>>script in our latest fork of GPL Ghostscript 8.15 to catch this sort
>>of thing; these three lines are all that are needed:
>>
>> if test -n "$GCC"; then
>> CFLAGS="-ansi $CFLAGS"
>> fi
>>
>
>
> The other, more clear option, is -std=<standard name>, where standard
> name is gnu, c99, c89, etc.
>
> -ansi is current equivalent to -std=c89, but i don't believe it will
> stay that way forever :)
Well, I don't see the ANSI folks falling over themselves to adopt
C99 (which is an ISO standard), but it does look like even old GCC
2.95 supports -std=foo, so I'd agree that -std=c89 is the better
option to add... :)
--
______________________________________________________________________
Michael Sweet, Easy Software Products mike at easysw dot com
Internet Printing and Document Software http://www.easysw.com
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Apr 11 16:23:19 2005