Friday 28 April 2006 17:51 skrev John Jetmore:
On Fri, 28 Apr 2006, W B Hacker wrote:
...and had delivered the messages into:
~/{domain}/{user}/Maildir/Maildir_create = yes/new
heh, never done it with exim, but I have to admit to creating a folder
along the line of ~/mail/* ^From:[EMAIL
Derek Listmail Acct wrote:
Is there a way to make this router just fall thru like .forward
didn't exist if $home doesn't exist?
I think the require_files precondition should do what you need.
http://exim.inode.at/exim-html-4.61/doc/html/spec_html/ch15.html#id2607819
@Philip (or Nigel?): Would
On Fri, 28 Apr 2006, Daniel Tiefnig wrote:
As it is now one has to review the source code of the website
to find the corresponding id tags...
A bit off-topic, but Pederick's web developer toolbar for
Mozilla/Seamonkey/Firefox (a useful extension *not only* for web
developers, IMHO) has a
Philip Hazel wrote:
Lots of things start failing when path names get longer than 256
characters (or perhaps 1024). I think 256 is a POSIX limit.
How does Exim handle that?
It could end up in running into some file system limit (like no more
free inodes though space left on the device) after
On Fri, 28 Apr 2006, listrcv wrote:
Lots of things start failing when path names get longer than 256
characters (or perhaps 1024). I think 256 is a POSIX limit.
How does Exim handle that?
Without searching the code, I can't be sure, but I don't think it does
any checks itself. However,
Michael Haardt wrote:
On Thu, Apr 27, 2006 at 05:14:27PM +0200, listrcv wrote:
The sieve filtering of Cyrus just falls back to delivering into the
inbox for deliveries into non-existing subfolders.
You do not need to change Exim to get that behaviour. Just have the file
transport check
Philip Hazel wrote:
It's the sysadmin who creates the string expansions and presumably
controls the contents of lookups. Or am I misunderstanding what you are
saying?
I'm not sure --- Exim offers so much flexibility that an admin setting
these things up may either be unaware of possible
listrcv wrote:
Philip Hazel wrote:
It's the sysadmin who creates the string expansions and presumably
controls the contents of lookups. Or am I misunderstanding what you are
saying?
I'm not sure --- Exim offers so much flexibility that an admin setting
these things up may either be
On 4/28/06 8:51 AM, John Jetmore [EMAIL PROTECTED] wrote:
On Fri, 28 Apr 2006, W B Hacker wrote:
...and had delivered the messages into:
~/{domain}/{user}/Maildir/Maildir_create = yes/new
heh, never done it with exim, but I have to admit to creating a folder
along the line of ~/mail/*
On Wed, Apr 26, 2006 at 08:34:50PM +0100, Philip Hazel wrote:
On Wed, 26 Apr 2006, Michael Haardt wrote:
I am just afraid that a maildirfolder file might get created where it
should not. As a result, Exim would not calculate the quota of a mailbox,
but the quota of all mailboxes. After
On Thu, 27 Apr 2006, Michael Haardt wrote:
Indeed it makes sense to create the parent maildir, too, in case it is
missing.
Unless prevented by the create_directory option, Exim will always create
all the directories it needs. The code I have committed does not create
the parent
Philip Hazel wrote:
Unless prevented by the create_directory option, Exim will always create
all the directories it needs.
Delivery will be deferred if anything Exim is trying to
create cannot be created.
Exim would create lots of directories when in some errorneous setup a
string
On Thu, 27 Apr 2006, listrcv wrote:
Exim would create lots of directories when in some errorneous setup a
string expansion (or whatever) unexpectedly yields a subdirectory that
is maybe a thousand or a twentythousand levels deep.
Lots of things start failing when path names get longer than
On Thu, Apr 27, 2006 at 01:58:35PM +0100, Philip Hazel wrote:
On Thu, 27 Apr 2006, Michael Haardt wrote:
Indeed it makes sense to create the parent maildir, too, in case it is
missing.
Unless prevented by the create_directory option, Exim will always create
all the directories it
Jakob Hirsch wrote:
Quoting Daniel Tiefnig:
Sadfully this is not the case with every IMAP installation out
there. (Cyrus e.g. may use / instead of . to delimit
subfolders.)
This may be true for the frontend (IMAP), but surely not for the
backend, which is the only thing that matters to
On Wed, 26 Apr 2006, Daniel Tiefnig wrote:
Philip Hazel wrote:
This bug has occurred because I never considered the possibility of
delivering into anything other than the new folder; I hadn't thought
of direct delivery into subfolders.
So, do you think we could get an expanded boolean
Philip Hazel wrote:
Consider the case when the transport ends up with
directory = some string expansion
where the expansion sometimes yields /some/maildir (for a non-folder
delivery) and sometimes yields /some/maildir/.subfolder - depending
on the local part or local part affix, or on
Daniel Tiefnig wrote:
directory = ${lookup mysql{SELECT maildir ...}}\
${if def:h_X-Spam:{/.Junk}{}}
subfolder_delivery: ${if def:h_X-Spam:{yes}{no}}
Another possibility would of course be a subdirectory option, that
causes delivery into a subfolder, if it doesn't expand to the
On Wed, 26 Apr 2006, Daniel Tiefnig wrote:
Consider my simple example, extended by subfolder_delivery:
directory = ${lookup mysql{SELECT maildir ...}}\
${if def:h_X-Spam:{/.Junk}{}}
subfolder_delivery: ${if def:h_X-Spam:{yes}{no}}
Documentation should be explicit about that,
Philip Hazel wrote:
After my last post I thought of a better idea: use a pattern. So you
could have
subfolder_delivery_regex = /\.Junk$
But if my maildirs are assembled like /somewhere/$domain/$local_part
and one of my users decides to create the mailbox [EMAIL PROTECTED]? :o)
Or if I
On Wed, 26 Apr 2006, Daniel Tiefnig wrote:
directory = ${lookup mysql{SELECT maildir ...}}
subdirectoy = ${if def:h_X-Spam:{/.Junk}{}}
Still clumsy, though.
Yes, and it would mean you couldn't lookup the whole name (directory
plus subdirectory) from a database or file. I think that's too
On Wed, Apr 26, 2006 at 11:59:22AM +0100, Philip Hazel wrote:
I'm not quite sure what to call the option: perhaps something like
maildirfolder_create_regex would be more explicit. It certainly needs to
include maildir because it wouldn't apply to other delivery types. So
a (better) pattern
On Wed, 26 Apr 2006, Michael Haardt wrote:
I like that solution and think it will work fine, but please add a check
if the parent directory contains cur, new and tmp directories.
Why? Is it mandatory to have those directories? This would prevent
somebody having *only* subfolders and no main
Philip Hazel wrote:
Why? Is it mandatory to have those directories? This would prevent
somebody having *only* subfolders and no main folder.
By my read of the maildir++ spec, it is, in fact, mandatory. However,
that's no reason for Exim to require it; be liberal in what you accept
and all
Marc Sherman wrote:
Philip Hazel wrote:
Why? Is it mandatory to have those directories? This would prevent
somebody having *only* subfolders and no main folder.
Is this really possible? (And if: what for?)
By my read of the maildir++ spec, it is, in fact, mandatory. However,
that's no
On Wed, Apr 26, 2006 at 09:10:41AM -0400, Marc Sherman wrote:
Philip Hazel wrote:
Why? Is it mandatory to have those directories? This would prevent
somebody having *only* subfolders and no main folder.
By my read of the maildir++ spec, it is, in fact, mandatory. However,
that's no
On Wed, 26 Apr 2006, Michael Haardt wrote:
I am just afraid that a maildirfolder file might get created where it
should not. As a result, Exim would not calculate the quota of a mailbox,
but the quota of all mailboxes. After consuming considerable resources,
the mail would most likely
On Wed, 19 Apr 2006, Derek Listmail Acct wrote:
Exim is noticing the 'maildirfolder' file in the .Trash folder and then
incrementing the maildirsize file in the parent directory.
The main problem is that by incrementing the maildirsize file when
delivering to the .Trash folder the
Derek Listmail Acct wrote:
Subfolders aren't stored any differently than other folders, they
just have the period in them more than once.
Sadfully this is not the case with every IMAP installation out there.
(Cyrus e.g. may use / instead of . to delimit subfolders.)
How can nonexistent
Sadfully this is not the case with every IMAP installation out there.
(Cyrus e.g. may use / instead of . to delimit subfolders.)
This isn't Maildir++ compliant behavior though is it?
--
## List details at http://www.exim.org/mailman/listinfo/exim-users
## Exim details at
Quoting Daniel Tiefnig:
Sadfully this is not the case with every IMAP installation out there.
(Cyrus e.g. may use / instead of . to delimit subfolders.)
This may be true for the frontend (IMAP), but surely not for the
backend, which is the only thing that matters to Exim. Cyrus uses some
On Thu, April 20, 2006 2:51 am, Daniel Tiefnig wrote:
Derek Listmail Acct wrote:
Exim is noticing the 'maildirfolder' file in the .Trash folder and
then incrementing the maildirsize file in the parent directory.
But if exim is the one to create the IMAP Folder, it will not create
this
Derek Listmail Acct wrote:
Exim is noticing the 'maildirfolder' file in the .Trash folder and
then incrementing the maildirsize file in the parent directory.
But if exim is the one to create the IMAP Folder, it will not create
this maildir file, because it's not aware it's delivering into a
On Tue, 18 Apr 2006, Derek Listmail Acct wrote:
maildir_quota_directory_regex = ^(?:cur|new|\.(?!Trash).*)$
I copied this from the documentation, which says this should cause the
Trash folder to be exempt from quotas.
It excludes the .Trash folder, not the Trash folder.
--
Philip Hazel
On Wed, April 19, 2006 2:26 am, Philip Hazel wrote:
On Tue, 18 Apr 2006, Derek Listmail Acct wrote:
maildir_quota_directory_regex = ^(?:cur|new|\.(?!Trash).*)$
I copied this from the documentation, which says this should cause the
Trash folder to be exempt from quotas.
It excludes the
On Wed, 19 Apr 2006, Derek Listmail Acct wrote:
On Wed, April 19, 2006 2:26 am, Philip Hazel wrote:
On Tue, 18 Apr 2006, Derek Listmail Acct wrote:
maildir_quota_directory_regex = ^(?:cur|new|\.(?!Trash).*)$
I copied this from the documentation, which says this should cause the
On Wed, April 19, 2006 2:54 am, Philip Hazel wrote:
On Wed, 19 Apr 2006, Derek Listmail Acct wrote:
On Wed, April 19, 2006 2:26 am, Philip Hazel wrote:
On Tue, 18 Apr 2006, Derek Listmail Acct wrote:
maildir_quota_directory_regex = ^(?:cur|new|\.(?!Trash).*)$
I copied this from the
Derek Listmail Acct wrote:
On Wed, April 19, 2006 2:54 am, Philip Hazel wrote:
On Wed, 19 Apr 2006, Derek Listmail Acct wrote:
*trim*
It excludes the .Trash folder, not the Trash folder.
The Trash folder on the file system begins with a period. ie,
Maildir/.Trash
Have a look at the
On Wed, 19 Apr 2006, Derek Listmail Acct wrote:
Here is the debug output (or the part I think your interested in)... I
don't see anything wrong, but I don't really know what I'm looking for
either...
ensuring maildir directories exist in
Philip Hazel wrote:
How is this happening? I thought that maildir deliveries always went
into the cur directory? Oh, it's going into .Trash/cur? Aarrgghh!!
What is the configuration of your transport?
Are there any maildir experts reading this who would like to comment? I
am not a
Marc Sherman wrote:
Philip Hazel wrote:
How is this happening? I thought that maildir deliveries always went
into the cur directory? Oh, it's going into .Trash/cur? Aarrgghh!!
What is the configuration of your transport?
Are there any maildir experts reading this who would like to comment?
Philip Hazel wrote:
How is this happening? I thought that maildir deliveries always went
into the cur directory? Oh, it's going into .Trash/cur? Aarrgghh!!
What is the configuration of your transport?
Are there any maildir experts reading this who would like to comment? I
am not a maildir
All,
I'm delivering a message to the Trash folder in a Maildir with quotas
enabled via an exim filter. I have the following set on the transport
doing the delivery, but the maildirsize file is being incremented when the
message is delivered.
maildir_quota_directory_regex =
43 matches
Mail list logo
