On Sat, Mar 06, 2004, Ralf S. Engelschall wrote:

> On Sat, Mar 06, 2004, Matthias Kurz wrote:
> 
> > > On Sat, Mar 06, 2004, Matthias Kurz wrote:
> > >
> > > > I think it is not a good idea that amanda forces the install of xinetd.
> > >
> > > What is the alternative you suggest?
> >
> > Hmmm, this question baffles me. Where is the problem with the native
> > inetd e.g. under Solaris ?
> 
> The problem is that for using the native inetd(8) the OpenPKG "amanda"
> package would have to edit the system global /etc/inetd.conf file. This
> is both a violation of our policy that we do not touch anything outside
> our instance and does not allow deployments into multiple instances,
> too. Additionally, not all inetd(8) allow listening on particular IPs
> (they often always do wildcard binds only). By using our own xinetd
> all those problems can be easily workarounded.

I recognized that there is no "portable" way to add entries to "the"
inetd.conf, until now. But i thought that this is not a great issue. I
missed the fact that some inetd do not support binding to particular IP
addresses - maybe because i do not use inetd for my networked applications
anyway.
Uhmmm, i know the inetd is not worthless - it would be a waste of resources
when every app would install for "direct" access.

> > Or can the native inetd and xinetd coexist as long as there are no duplicate
> > entries in their config files ?
> 
> Yes, they can coexist as long as they do not listen on conflicting IP/port
> combinations.

Cool. That's just a very nice solution, isn't it ? So, why are you asking
for alternatives ;-)

> > I never wanted to replace the native inetd,
> > so i never tried it. And i can not try it currently, because there the make
> > aborts (under Solaris 8/x86) with
> >
> > ld: fatal: file ../libs/lib/libportable.a: unknown file type
> > ld: fatal: library -lportable: not found
> > ld: fatal: File processing errors. No output written to xinetd
> > collect2: ld returned 1 exit status
> > /opkg/home/bin/gcc -O2 -pipe  -I../libs/include  ./itox.c -o itox -L../libs/lib 
> > -lsio -lmisc -lxlog -lportable -lstr -lpset -lm -lkstat -lresolv -lsocket -lnsl
> > ld: fatal: file ../libs/lib/libportable.a: unknown file type
> > ld: fatal: library -lportable: not found
> > ld: fatal: File processing errors. No output written to itox
> > collect2: ld returned 1 exit status
> > make[1]: *** [itox] Error 1
> > make: *** [makeprog] Error 2
> > error: Bad exit status from /projects/tmp/opkg/rpm-tmp.1098 (%build)
> >
> > and i have no time to look into it. Perhaps this evening...
> 
> Ah, perhaps a parallel build issue: try "%{l_mflags}" if we have
> "%{l_mflags -O}" in the package...

Ridiright. The error happened on a dual CPU coachwakki. Though, i looked
at the .spec-file, and there the only "-O" is in the compilation part,
not where make is called. I know that i try to pass my problem - i'll
look at it tomorrow, and when nobody solved it, i will...


   (mk)

-- 
Matthias Kurz; Fuldastr. 3; D-28199 Bremen; VOICE +49 421 53 600 47
  >> Im prämotorischen Cortex kann jeder ein Held sein. (bdw) <<
______________________________________________________________________
The OpenPKG Project                                    www.openpkg.org
Developer Communication List                   [EMAIL PROTECTED]

Reply via email to