>Furthermore, gethostent_r() in the IRS implementation is incompatible
>with existing implementations like Solaris or glibc.
>
>The current position of gethostent_r() within an enumeration on
>Solaris and glibc is a process-wide property which is shared by all
>threads.
>But current position of gethostent_r() in the IRS library is a
>thread-specific property, and it is not shared by other threads.
>
>Because applications which uses gethostent_r() usually assume
>Solaris/glibc behavior, the IRS implementation is not only useless,
>but also dangerous because it provides different semantics with
>same function name.

	which standard defines the behavior?  if there's no standard the
	programmmer who assumed solaris/glibc behavior is the one to be
	punished.  you can't blame ISC (nor solaris/glibc).

itojun