Re: [Nmh-workers] problems with slocal

2009-07-27 Thread Jerrad Pierce
>From strace write(1, "!/var/spool/mail/belg4mit\n", 26) = 26 open("/var/spool/mail/belg4mit", O_RDWR|O_CREAT|O_NONBLOCK, 0600) = 5 write(1, "!/var/spool/mail/belg4mit.lock\n", 31) = 31 uname({sys="Linux", node="vm10.myvpshost.com", ...}) = 0 umask(022) = 077 open("/va

Re: [Nmh-workers] problems with slocal

2009-07-27 Thread Jerrad Pierce
>You might try telling slocal to be verbose, in which case it might >say what was going wrong, or alternatively run under a debugger and >see why status is being set to -1... This has been suggested before, but it's not the most end-user friendly route. # slocal -debug vec[0]: "default" vec[1]: "-

Re: [Nmh-workers] problems with slocal

2009-07-27 Thread Peter Maydell
Jerrad Pierce wrote: >Strike that, reverse it. As long as -user *is not supplied*, presumably >due to these lines of localmail: > >443 /* last resort - deliver to standard mail spool */ >444 #ifdef SLOCAL_MBOX >445 return usr_file (fd, mbox, MBOX_FORMAT); >446 #else >447

Re: [Nmh-workers] problems with slocal

2009-07-27 Thread Jerrad Pierce
>At least on my box, as long as you give slocal -user when it's invoked, Strike that, reverse it. As long as -user *is not supplied*, presumably due to these lines of localmail: 443 /* last resort - deliver to standard mail spool */ 444 #ifdef SLOCAL_MBOX 445 return usr_file (f

Re: [Nmh-workers] problems with slocal

2009-07-27 Thread Jerrad Pierce
After doing without notifications, and typing 'from' (aliased msgchk) every 30 seconds, I finally decided to tackle this issue, and it turned out to be *almost* pretty simple. At least on my box, as long as you give slocal -user when it's invoked, any undelivered messages fall through .maildelivery