On Wed, Jan 30, 2002 at 12:17:31PM -0800, B. James Phillippe spamblock wrote:
> Hello,
> 
> A co-worker of mine recently noticed that NetBSD does not define ENOTSUP
> (though it does define EOPNOTSUPP).  According to a quick net search, it
> appears that ENOTSUP and EOPNOTSUPP are actually two different error codes
> (Single UNIX Specifiction v2)*.  I'm curious to know what the BSD take on
> this is, since otherwise NetBSD seems quite conformant to the spec.

While I have no comment on this particular issue, I will note that while
we use SUS as a guideline in some areas of the system interface, our
documentation on the subject is pretty explicit in stating that we don't
generally consider SUS to be an example of a "reasonable standard".

In my opinion, SUS is a worse grab-bag than SVR4: every feature any Unix
vendor ever implemented, including many/most of the harebrained, broken
ones, is in there, often specified in terms of the worst, most unreasonable
behaviour or *any* system offered by any of the organizations that
participated in writing the spec.  An obvious example is that SUS 
explicitly allows PTY allocation to fail in programs that happen to have
handlers for SIGCHLD -- why?  Because someone's broken System V 
implementation couldn't manage to use waitpid, and unconditionally zorched
the existing SIGCHLD hander instead when it went to run an external pt_chown
program.  Of course, the people writing the SUS didn't dare offend *any*
of their corporate masters -- the spec is just littered with examples of
the worst kind of standard-writing of this type.

-- 
 Thor Lancelot Simon	                                      tls@rek.tjls.com
   But as he knew no bad language, he had called him all the names of common
 objects that he could think of, and had screamed: "You lamp!  You towel!  You
 plate!" and so on.              --Sigmund Freud