James Carlson wrote:
> John Sonnenschein writes:
> > I've tried the normal channel with no bites, seems people don't like
> > doing CR ;)
> >
> > Anyhow, could I get a code review/sanity check on 6733971 from someone ?
> >
> > code at:
> > http://cr.opensolaris.org/~error404/6733971/
> 
> Instead of introducing spaces -- which I think makes the code just a
> little more obscure -- why not get rid of the superfluous parenthesis?

Good question... the answer is that we do not know exactly _why_ this
stuff was originally added with subshells. It seems to be done
intentionally but noone knows _why_.

> In other words, instead of this:
> 
>         @( $(LHEAD) $(LINT.c) $< $(LTAIL) )
> 
> just do this:
> 
>         @$(LHEAD) $(LINT.c) $< $(LTAIL)
> 
> That way, there's no question for future maintainers, and both forms
> appear to be equivalent for ksh93.
> 
> Are all these subshells doing us any good?

As said above I tried to get an answer and so far the investigation was
for /dev/null. My gut feeling says we should keep them.

Options so far are:
1. Keep the patch in the current state and forward the investigation to
the ToDo list of the OS/Net Janitor project
2. Keep the patch and ignore the extra brackets (they cause an extra
|fork()| for the original Bourne shell but ksh93 does not |fork()| for
subshells which will make this efectively a no-op)
3. Make a new patch and remove the extra brackets

I would prefer [1] in this case since [3] would turn this patch IMO into
a high-risk patch and [2] is IMO just reckless.

----

Bye,
Roland

-- 
  __ .  . __
 (o.\ \/ /.o) roland.mainz at nrubsig.org
  \__\/\/__/  MPEG specialist, C&&JAVA&&Sun&&Unix programmer
  /O /==\ O\  TEL +49 641 3992797
 (;O/ \/ \O;)

Reply via email to