On Wed, 10 Feb 2016 19:26:17 +0100, Peter Hunkeler wrote:

>> ?That doesn?t apply to ?true?, though, right??
>> ?Of course not, use some common sense.?
>
>That would require the knowledge of /bin/true to be common sense, which I 
>doubt. I like the idea of help being available even for what might look like 
>an obvious command to some.
> 
Agreed.  And "IEFBR14" is far less "common sense" than "true".  So, where
does IBM document IEFBR14.  Utilities?  No, those are largely "IEB" prefix.
A brief search turns up several documents that mention use of IEFBR14, but
none that officially specify it.  It shouldn't default to "That's common 
knowledge,"
as an IBM employee has here attempted to justify absence of documentation
of another z/OS facility.

I note that in one GNU Linux, /bin/true respects "--help" and "--version";
builtin true does not.  This violates POSIX, which requires the behavior
to be the same.  (But GNU Linux doesn't claim to be POSIX.)

On Wed, 10 Feb 2016 10:20:58 -0700, Jack J. Woehr wrote:
>>
>> http://www.theregister.co.uk/2016/02/10/line_break_ep2/?page=2
>
>If you read down to the bottom, you'll understand why the cognoscenti (such as 
>I, snerk snerk) live on OpenBSD, not Linux.
>
Maybe.  In fact, that's a call back to the shell builtin, which:

o May involve extra overhead of fork()/exec().

o Should (POSIX requires) running the user's $ENV script, which
  may have side effects that violate the specification of the function.
  Many (I count 17) z/OS utilities are implemented as such callbacks.
  My $ENV (sometimes, for testing) issues messages.  That's contrary to
  the specification of "true".  One more thing to PMR when I'm in a bad
  mood.  IIRC, OS X has a tweak that performs the callback without
  invoking the $ENV script.

(I use ":" instead of "true")

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to