On Mon, Jun 16, 2008 at 03:24:26PM -0400, James Carlson wrote:
> Ceri Davies writes:
> > On Mon, Jun 16, 2008 at 10:04:42AM -0400, James Carlson wrote:
> >
> > > The existing /etc/rc*.d stuff is there so that software that *must*
> > > run on older versions of Solaris (either because it was originally
> > > written there and never updated for S10 or newer, or because it's
> > > still supported on those old versions) can still run.
> > >
> > > Since that's the purpose, the constraints of S9 and older here are
> > > quite important for the discussion.
> > >
> > > And on S9, if you put "#!/bin/ksh" at the top of your rc script, it
> > > won't work. Never has, never will.
> >
> > Hmm. I agree that changing it probably isn't a good idea, but there's a
> > subtle difference on S10 as Nico pointed out explicitly: the only way to
> > get a binary or non-/bin/sh script to be executed correctly on S10 is to
> > call it S99foo.sh, which is just plain weird.
>
> Not only weird, but an undocumented side-effect of the implementation.
> Where does anything say that ".sh" is even supported?
>
> > I think that at the very
> > least this deserves a NOTE in a (which?) manpage, even if it's just a
> > statement that everything is expected to be /bin/sh scripts.
>
> /etc/init.d/README has this documentation.
>
> > > Note that these supposed rc scripts could never have been run on S10
> > > or anything older, because they won't work. So, they must have either
> > > never been tested on Solaris at all (in which case it's hard to see
> > > how they're supported), or they're specifically designed for
> > > OpenSolaris.
> >
> > Who says they're supported? Numerous vendors will quite happily give
> > you an rc script that runs on Linux and expect you to take it from
> > there. In my case, my backup guys had used the exact same script on my
> > Solaris machines as they do on other folk's Linux boxes (#!/bin/bash,
> > natch).
>
> In that case, what you're doing is called "porting." Yes, even if all
> that you need to do is rewire the rc script because it's not out-of-
> the-box ready for Solaris, I claim that this is porting.
>
> The most trivial but valid port is probably this:
>
> cat >/etc/init.d/mything <<EOF
> #!/sbin/sh
*bzzt!* /bin/sh, thank you, very much.
> /path/to/mything/rc-script $*
> EOF
> chmod +x /etc/init.d/mything
> ln -s ../init.d/mything /etc/rc3.d/S99mything
> ln -s ../init.d/mything /etc/rcS.d/K00mything
>
> ... but there are probably better ports that can be done. I don't
> think I agree that rewriting the legacy support for rc scripts is a
> necessary or desirable component of that. It's not clear that it'd
> even work right, as it obviously assumes that all of the bits install
> the same way on Solaris as on these other systems (the paths are the
> same), and may well make many other assumptions.
Not given "even if all that you need to do is rewire the rc script".
You're flapping between extremes.
Ceri
--
That must be wonderful! I don't understand it at all.
-- Moliere
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/smf-discuss/attachments/20080616/429b9e2c/attachment.bin>
