Hi all,

sorry if this has been answered already, i did check on the list archives but couldn't quite figure out my problem.

I have a general maildroprc that has among other things this:

SHELL="/bin/sh" TIMESTAMP=`date "+%b %d %H:%M:%S"` logfile "log.txt" log "=== new run on $TIMESTAMP ==="

and a user-included maildroprc from there that looks kinda like this:

SMS=1 if ( $SMS == 1 ) { exception { cc "| /path/to/some/prog $USER" xfilter "subjadd [SMS]" } log "EXITCODE=$EXITCODE, TEMPFAIL=$EX_TEMPFAIL" echo "EXITCODE=$EXITCODE, TEMPFAIL=$EX_TEMPFAIL" } to $DEFAULT

"prog" takes usermail and the message, works on it some and then sends out an sms via an sms gateway. the internals of prog seem ok as the program works fine on the command line ("cat mailfile | prog us...@server.tld") but I am unable to run it from within maildrop:

Feb 26 15:33:14 - BEGIN maildrop processing for us...@server.tld === == User will.test has own mailfilter, including /var/spool/mail/virtual_mailboxes/.mailfilters/us...@server.tld ... inside user's mailfilter file... Date: Mon Feb 26 15:33:14 2007 From: Fred Jounters <fj...@server.tld> Subj: dada !Err: | /path/to/some/prog user\@server.tld (918)

EXITCODE = 4294967295, TEMPFAIL =

Now from another mailinglist I found out that 4294967295 == 0xffffffff == (unsigned int) -1

Question: What does this mean? / Where to look for the error?

Whenever my prog runs well, it returns "1", else it might return something like "60", "59", etc.

Does the prog always have to return TRUE/1 for mailfilter to keep on going "ok"?

Btw: the prog is a perl program with executable bits set and a correct shebang, no switches (wcT) so that it doesn't output anything else.

Any ideas? Thanks for looking into this, Fred