Hi qmailers,
I found a problem with the databytes check.
As it's explained in man pages, qmail-smtpd checks the message size
AFTER writes it on disk. In addition to performance questions, this
behaviour involves that CR+LF characters are changed by LF characters
BEFORE the size check.
Howeve
On Tue, 31 Aug 1999 [EMAIL PROTECTED] wrote:
> Hi qmailers,
>
> I found a problem with the databytes check.
>
> As it's explained in man pages, qmail-smtpd checks the message size
> AFTER writes it on disk. In addition to performance questions, this
> behaviour involves that CR+LF characters
>
> It doesn't say that. It says it 'counts bytes AS stored on
> disk', not
> 'counts bytes AFTER they're stored on disk'. There's a difference. A
> quick look at the sources will confirm this.
>
Vince,
Thanks by your correction.
However, I verified that the messages stored on disk (i
On Tue, 31 Aug 1999 [EMAIL PROTECTED] wrote:
> >
> > It doesn't say that. It says it 'counts bytes AS stored on
> > disk', not
> > 'counts bytes AFTER they're stored on disk'. There's a difference. A
> > quick look at the sources will confirm this.
> >
>
> Vince,
>
> Thanks by your cor
On Tue, 31 Aug 1999 07:30:23 -0400 (EDT), [EMAIL PROTECTED]
wrote:
>Moreover, the value recorded in qmail logs is exactly the data message
>size without the CR+LF number.
If this is important to you, modify qmail-send by subtracting 1 from
bytestooverflow (and check it as in put()) in blast()
[EMAIL PROTECTED] wrote:
>Since the difference between size meanings depends on each message's
>line number, it's very difficult to fit the qmail's size limit with
>others MTA's.
So what? who cares if your max message limit counts LF's and mine
counts CRLF's?
-Dave