Thread: pgsql/src/bin/initdb Makefile initdb.sh
CVSROOT: /home/projects/pgsql/cvsroot
Module name: pgsql
Changes by: momjian@hub.org 01/05/08 12:28:46
Modified files:
src/bin/initdb : Makefile initdb.sh
Log message:
Allow SHELL in Makefile.* to control initdb.
Bruce Momjian - CVS <momjian@hub.org> writes:
> Allow SHELL in Makefile.* to control initdb.
Waste of time, Bruce. By my count, our distribution contains sixty-five
other shell scripts with hard-wired "#! /bin/sh", most of which are not
being passed through any autoconfiguration step. Foremost on the list
is configure itself. It's going to be rather difficult to autoconfigure
anything if /bin/sh doesn't work.
I would suggest reverting this change, since there's no point in having
initdb work differently from all our other shellscripts. I agree with
Doug McNaught that any system where /bin/sh isn't a Bourne shell is
not Unix.
regards, tom lane
> Bruce Momjian - CVS <momjian@hub.org> writes: > > Allow SHELL in Makefile.* to control initdb. > > Waste of time, Bruce. By my count, our distribution contains sixty-five > other shell scripts with hard-wired "#! /bin/sh", most of which are not > being passed through any autoconfiguration step. Foremost on the list > is configure itself. It's going to be rather difficult to autoconfigure > anything if /bin/sh doesn't work. > > I would suggest reverting this change, since there's no point in having > initdb work differently from all our other shellscripts. I agree with > Doug McNaught that any system where /bin/sh isn't a Bourne shell is > not Unix. Another issue is that initdb is already going through parameter substitution, so it wasn't a big deal to fix it. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
> Bruce Momjian - CVS <momjian@hub.org> writes:
> > Allow SHELL in Makefile.* to control initdb.
>
> Waste of time, Bruce. By my count, our distribution contains sixty-five
> other shell scripts with hard-wired "#! /bin/sh", most of which are not
> being passed through any autoconfiguration step. Foremost on the list
> is configure itself. It's going to be rather difficult to autoconfigure
> anything if /bin/sh doesn't work.
>
> I would suggest reverting this change, since there's no point in having
> initdb work differently from all our other shellscripts. I agree with
> Doug McNaught that any system where /bin/sh isn't a Bourne shell is
> not Unix.
I understand, but why is SHELL set in Makefile.ultrix, and how did they
get far enough for initdb? Someone must have played with this long ago.
My guess is that he somehow he overrode SHELL in configure, perhaps by
setting CONFIG_SHELL:
SHELL=${CONFIG_SHELL-/bin/sh}
If I back out the patch, I would remove the SHELL from Makefile.ultrix.
If initdb is Ultrix's only problem, seems we are OK keeping it. Any
other problems and I would discard the patch.
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026
Bruce Momjian <pgman@candle.pha.pa.us> writes:
> If initdb is Ultrix's only problem, seems we are OK keeping it.
We have no reason to think that. The complainant hadn't gotten past
initdb, so how would he know what other problems there are?
regards, tom lane
> Bruce Momjian <pgman@candle.pha.pa.us> writes: > > If initdb is Ultrix's only problem, seems we are OK keeping it. > > We have no reason to think that. The complainant hadn't gotten past > initdb, so how would he know what other problems there are? Let's see what they report back. I sent them the patch. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026