Re: Cyrus to USENET gateway

[Florian Hars]

Ken Murchison wrote:
Why not just have the newsserver feed directly to Cyrus?
Things aren't what they used to be. Typical providers will offer their 
clients NNRP access with NEWNEWS disabled (if they offer news at all), 
there is no talk of "feeding".

Trying to get suck  to 
cooperate with cyrus would probably be easier that reimplementing suck 
within cyrus.

Yours, Florian.

Re: skiplist vs. Berkeley

[Florian Hars]

Dmitry Alyabyev wrote:
Which the advantages are in skiplist comparing with BerkeleyDB ?
It doesn't show seen messages as new tree times a day.

Yours, Florian.

Re: SSL/TLS

[Florian Hars]

Lee Hoffman wrote:
> This is VERY weird!!! When I telnet into the mailserver on 993:
> No commands works, yet it says that its connected! '. logout' does
> nothing, '. starttls' does nothing etc...

The "Hacker Test" that floated around the net several years ago hat the 
two questions:

+++ Can you whistle 300 Baud?
+++    1200?

If you have to answer "No" to these questions you should probably also 
refrain from trying to do telnet connects to a SSL service. Not only 
because the math is quite hard to do in your head, but even more so 
because you cannot enter all the necessary octets with a normal 
keyboard. You would have whistle them (in 8,n,1) into a phone connected 
to a modem.

But seriously: 993 is the imaps port and it is the whole point of imaps 
that plaintext connections don't work.

Yours, Florian Hars.

Re: trash-bin

[Florian Hars]

[EMAIL PROTECTED] wrote:
> Is there an easy way to implement a trash-bin in imapd ?
> Its not that I would need one but my users demand it. (you know they
> once had one in outlook for their pop-accounts )

I created a folder named "Trash" in the users INBOX and activated the
altnamespace, and at least Mozilla uses this folder as its Trash bin
without any further configuration on the server or the client after
selecting the option that Mozilla should move deleted messages to
the trash bin.

Somtimes the obvious works even if you use a computer :-).

Yours, Florian Hars.

sieveshell segfaults on sasl configuration error

[Florian Hars]

Something like this might make life easier for admins who still
have problems understanding what the first S in SASL is supposed to
be referring to:

--- perl/sieve/managesieve/managesieve.c.2.1.4  Thu May 16 17:06:30 2002
+++ perl/sieve/managesieve/managesieve.cFri May 17 08:46:02 2002
@@ -281,6 +281,12 @@
 
   mechlist=read_capability(obj);
 
+  if (mechlist == NULL) {
+ /* Some useful error message complaining about missing SASL
+  * methods.
+  */
+  }
+
   mlist = (char*) xstrdup(mechlist);
   if(!mlist) XSRETURN_UNDEF;

Yours, Florian Hars.

What ever happened to deliver -m ?

[Florian Hars]

I am trying to use deliver (2.1.4) from procmail. According to the
man-page and several HOWTO-documents on the web describing solutions
with earlier versions of cyrus, "deliver -a foo -m user.foo.folder"
shoud deliver to the folder "folder" of user foo, as should
"deliver -a foo -m folder foo". Now in my case, the last sytax delivers
to foo's INBOX, while the first one fails because "+user.foo.folder:
Mailbox does not exist". 

Is this an intended, but not yet documented change in the semantics of
the -m option, an off-by-one error in the code that handles the
expansion of the '+' chars that newer versions of deliver happen to 
introduce into mailbox names, or a configuration error on my part? 

Yours, Florian Hars