FreeTDS Development Group

["James K. Lowden" <jklowden AT freetds.org>]

ZIGLIO, Frediano, VF-IT wrote:
> I don't like all that code just to fix a problem for a single system but
> this problem has to be fixed... and problem reside on the system header
> so you are just writing a workaround...

Agreed.  In general it's better to test for the feature than for the OS,
but there's another general rule: it's better to fix bugs upstream than to
work around them, and workarounds don't have to be elegant.  Or general. 
;-)

With that in mind, let me try to save you some work, or at least offer a
better alternative.  Please read 

http://h50236.www5.hp.com/solutions/copartner/IA64/reference/Porting_Tuning/portguideipf.pdf
and http://www.docs.hp.com/en/B9106-90013/xopen_networking.7.html.  

It looks like defining _XOPEN_SOURCE might do the trick.  Certainly HP
knows about the conflict.  

I suggest the OP go back to HP and see if there's some way to define his
environment such that FreeTDS compiles without your patch.  If he's
willing to do that but needs a clearer test case, would you be willing to
create one?  

Regards, 

--jkl