Re: [Gnumed-devel] boostrap on openSUSE

[Karsten Hilbert]

> On Sunday, November 27, 2011 05:36:09 PM Karsten Hilbert wrote:
> > > > > >OSError: [Errno 17] File exists
> > > > > >Bootstrapping "gnumed_v16" did not finish successfully. Aborting.
> > > > > 
> > > > > Seems not log file is created.
> > > > > 
> > > > > Removing the faulty symlink actually starts the boostrap process.
> > > 
> > > It cannot create and therefore write to /var/log/gnumed/server
> > > 
> > > When manually creating that directory it works
> > 
> > This seems to indicate that the RPM needs improvement.
> > 
> 
> Why would it indicate that ?
> 
> It indicates that gm_boostrap wants to write in that directory and does
> not 
> check if it is there ?
> 
> It could create that directory if it was missing

The bootstrap-latest.sh script does not assume a log directory. It
uses ./ unless GM_LOG_BASE is set:

if test ! -n "${GM_LOG_BASE}" ; then
        GM_LOG_BASE="."
fi ;

Hence GM_LOG_BASE must be getting set elsewhere, namely to
/var/log/gnumed/server/ which, all in all, seems sensible.

I would think it reasonable for the script to assume that a
directory passed to it would exist, no ?

Unless I am mistaken Debian policy would also support not blindly
creating directories all over the place (as instructed, that is).

AFAIR I had an exchange with Andreas about the Debian packages
and I seem to remember some agreement that the log file should NOT
by default be generated in ./ but rather somewhere known in case of
an absent GM_LOG_BASE.

Karsten
-- 
NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie!               
Jetzt informieren: http://www.gmx.net/de/go/freephone