On Feb 26, 2008, at 7:22 AM, Arkadiusz Miskiewicz wrote:

>
> 4.4.9 is working fine so count this one as regreesion.
>

I'll count as an RFE. The flaw that was fixed has been there for ever.

(aside) It would be easy to have it both ways if the getopt(3)  
POSIXLY_CORRECT envvar
was a viable mechanism was acutally a viable way to control getopt(3)  
behavior.
Setting/removing an envvar to control library functionality is the  
only means known
atm to control getopt(3) behavior.

73 de Jeff