Re: Trouble starting postfix

From: Magnus Bäck (no email)
Date: Mon Aug 01 2005 - 11:53:18 EDT

  • Next message: Scott Merrill: "After Queue BCC"

    On Monday, August 01, 2005 at 11:50 CEST,
         Tony Earnshaw <> wrote:

    > Having looked at the newaliases-style alias map, I understand that
    > it's differently formatted from other maps.

    Yes, local alias files in Postfix have this particular format, as
    documented in aliases(5). There is no reason to refer to this format as
    "newaliases-style". It simply is the official local alias file format
    for indexed maps. The fact that a database built with postmap(1) works
    is probably more of an implementation sideeffect rather than a behaviour
    that can be relied on.

    > I've never used such a formatted map for aliases; apart from a brief
    > excursion into a mailman alias saga, which I left for Dax Kelson's
    > postfix-to-mailman.py, I've never used /etc/aliases.

    That certainly explains your confusion.

    > All my alias maps and virtual alias maps are in LDAP, which uses
    > standard table lookups and "just works". If I do use an ordinarily
    > formatted hash map for /etc/aliases, that works too. In fact, I can
    > do more with such a map than with a newaliases-formatted map.

    How so?

    > Neither are virtual alias maps, as detailed in VIRTUAL_README,
    > formatted in the newaliases (or as you put it "old unix style
    > aliases files) style.

    No, of course they're not. The format of virtual alias files is
    documented in virtual(5).

    -- 
    Magnus Bäck
    

  • Next message: Scott Merrill: "After Queue BCC"





    Hosted Email Solutions

    Invaluement Anti-Spam DNSBLs



    Powered By FreeBSD   Powered By FreeBSD