[petsc-dev] http://ftp.mcs.anl.gov/pub/petsc/nightlylogs/archive/2013/12/2 0/examples_master_arch-linux-c89_thrash.log

[Barry Smith]

On Dec 21, 2013, at 10:47 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> Barry Smith <bsmith at mcs.anl.gov> writes:
> 
>>  So a known problem that has been hanging around for a long time breaking our tests but never fixed.
> 
> Yes, unfortunately.  The right answer depends on what we want -std=c89
> to do.  If we intend to use feature test macros to enable POSIX, the
> test snippet should do so and then we'll use

   This sounds reasonable. So PETSC_DESIRE_FEATURE_TEST_MACROS just needs to be added in the PetscRand_Rand48() file and the python missing test augmented? Then make it so.

>  If we are trying to produce an
> artificially-limited environment that can only be found on very obscure
> and obsolete operating systems, then maybe we should have a
> --with-feature-test-macros=0 option.

    The intention is that -std=c89 is used to insure that young bucking broncos don’t use every subtle feature of c99 all over the source code. It is not intended to cripple us.

   Barry

> 
>>   Hmm, then maybe it should be removed rom the list 
>> 
>>   functions = ['access', '_access', 'clock', 'drand48', 'getcwd', '_getcwd', 'getdomainname', 'gethostname’,    
>> 
>>   it is confusing to have it tested in two places, I saw this one and just assumed it is the only place.
> 
> Indeed.
> 
>> request-assigned: petsc-dev missing drand48() prototype with -std=c89
> 
> Is this a tagging experiment of yours?