Matt Sergeant wrote:
So I'm about to apply Hano's fix for this - if you still object we can
always back out the patch. The point being that per-object hooks never
really worked, and this way is faster.
Go ahead, I've been off the net for the last 5 days (installing a new OS on a
new hard driv
On Mon, 31 Mar 2008, John Peacock wrote:
but I don't know what Matt was trying to fix. As Ask surmised, the whole
point was to have the object hooks, not a package hooks, be used, since the
hooks registered are attributes of the object, not the class.
Attached is a diff that fixes both tls a
Matt Sergeant wrote:
But there's a patch in this thread to fix that, right?
Not a complete one, no. I'm at $WORK now, and I can't take a look at it
Changing %hooks to an "our" makes that work though.
Yes, but my mind automatically recoils at using package globals like
that. YMMV ;-)
I
Matt Sergeant wrote:
Because otherwise hooks are re-loaded on every connection (register()
needs re-run on every mail) which is a huge performance detriment.
Then we need to rethink that architecture; would it be acceptable to
return immediately if hooks are already defined (instead of just wa
On 31-Mar-08, at 11:04 AM, John Peacock wrote:
It was because AUTH might have two roundtrips with the client (or
not) depending on the AUTH method used. The first AUTH command is
clearly defined (since it starts with "AUTH" ;-), but the client
response to the server challenge has no obvious
On 31-Mar-08, at 10:12 AM, John Peacock wrote:
Matt Sergeant wrote:
Because otherwise hooks are re-loaded on every connection (register
() needs re-run on every mail) which is a huge performance detriment.
Then we need to rethink that architecture; would it be acceptable
to return immediatel
On 31-Mar-08, at 7:34 AM, John Peacock wrote:
Attached is a diff that fixes both tls and auth, by backing out 814
and tweaking a couple of other lines. I won't commit it until Matt
pipes up why he tried to make hooks a package global...
Because otherwise hooks are re-loaded on every connect
Hanno Hecker wrote:
On Sun, 30 Mar 2008 21:51:26 -0700
Ask Bjørn Hansen <[EMAIL PROTECTED]> wrote:
On Mar 30, 2008, at 4:39 PM, Markus Ullmann wrote:
After tracing it down to rev 814-818, a friend of mine came up with
this patch and it seems to make it work for me again. Maybe someone
with
On Sun, 30 Mar 2008 21:51:26 -0700
Ask Bjørn Hansen <[EMAIL PROTECTED]> wrote:
>
> On Mar 30, 2008, at 4:39 PM, Markus Ullmann wrote:
>
> > After tracing it down to rev 814-818, a friend of mine came up with
> > this patch and it seems to make it work for me again. Maybe someone
> > with dee
On Mar 30, 2008, at 4:39 PM, Markus Ullmann wrote:
After tracing it down to rev 814-818, a friend of mine came up with
this patch and it seems to make it work for me again. Maybe someone
with deep qpsmtpd knowledge could take a look at it
I think the patch breaks a feature of being able t
>
> After tracing it down to rev 814-818, a friend of mine came up with
this
> patch and it seems to make it work for me again. Maybe someone with
deep
> qpsmtpd knowledge could take a look at it
>
> Greetz
> -Jokey
>
I would test this path ASAP. Auth is indeed broken : nothing is
advertised as
11 matches
Mail list logo
