Quoting from upstream.
Julian Field writes:
> This is the ancient old problem the Postfix authors (Wietse in
> particular) have with MailScanner. I dared to "think outside the box"
> and not do things the way Wietse intended them to be done. Every way
> that Wietse didn't think of is wrong. They keep whinging. There are
> thousands of people using Postfix and MailScanner who do not have any
> problems at all. I fail to see how using internal data structures that
> they refuse to publish automatically means that data will be lost. It
> would be easier all round if they would just get over it and publish
> their formats, like all the other MTAs have.
> The checks I do on the files in the "hold" queue are these:
> ($nlinks, $tmpdate) = (stat($file1))[3,9]; # 9 = mtime
> next if -z _;
> next unless -f _;
> next unless -R _;
> next if $nlinks>1; # Catch files being moved into "hold"
> as I think this covers all the eventualities, watching for files
> appearing in the "hold" queue.
Bastian, do you have other points which could rectify the severity of
this report?
Matthias
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
