On 23.2.2007, at 1.09, Robert Schetterer wrote:

...

• Trash plugin was more or less broken with multiple namespaces and with multiple trash mailboxes

Hi Timo, did you change something with trash plugin, which seemed to me
still be broken in 1.rc23?

Yes.

and i missed the download url in your mail *g

Yea, forgot it this time.

Timo Sirainen schrieb:

On 23.2.2007, at 1.09, Robert Schetterer wrote:

...

...

- Trash plugin was more or less broken with multiple namespaces and
  with multiple trash mailboxes

Hi Timo, did you change something with trash plugin, which seemed to me still be broken in 1.rc23?

Yes.

...

and i missed the download url in your mail *g

Yea, forgot it this time.

Hi Timo,

i am sorry that

trash plugin lda seems not to work in 1.rc24 as well :ignore=Trash in quota with sql

incomming mail is simply bounced with over quota, and i cant delete messages to trash if this would go over quota

i have nothing special in the logs about the missfunction, what should i do to give you more verbose info for checking the bug?

this are confs

auth_verbose = yes mail_debug = yes pop3_uidl_format = %08Xu%08Xv

base_dir = /var/run/dovecot/ protocols = imap imaps pop3 pop3s listen = * shutdown_clients = yes log_path = /var/log/dovecot info_log_path = /var/log/dovecot.info log_timestamp = "%b %d %H:%M:%S " disable_plaintext_auth = no ssl_disable = no login_chroot = no

#maildir_copy_with_hardlinks = yes

ssl_cert_file = /etc/postfix/mailserver.cert ssl_key_file = /etc/postfix/mailserver.key

# Login processes login_dir = /var/run/dovecot/login login_process_per_connection = no login_processes_count = 5

# Set limit for MySQL lookup processes auth_worker_max_count = 50

login_greeting = Welcome login_log_format_elements = user=<%u> method=%m rip=%r lip=%l %c login_log_format = %$: %s default_mail_env = maildir:/usr/local/virtual/%h first_valid_uid = 1001

protocol imap { mail_plugins = quota imap_quota trash } protocol pop3 { mail_plugins = quota } plugin { quota = maildir:ignore=Trash trash = /etc/dovecot/dovecot-trash.conf }

mail_extra_groups = mail

protocol lda { mail_plugins = cmusieve quota trash log_path = /var/log/dovecot-deliver.log info_log_path = /var/log/dovecot-deliver.log postmaster_address = postmaster@darkchatroom.de auth default { mechanisms = plain

passdb sql { args = /etc/dovecot/dovecot-sql.conf }

         userdb sql {
         args = /etc/dovecot/dovecot-sql.conf
         }

         userdb prefetch {
         }

         socket listen {

         master {
         path = /var/run/dovecot/auth-master
         mode = 0660
         user = vmail
         group =vmail
         }

         client {
         path = /var/spool/postfix/private/auth
         mode = 0660
         user = postfix
         group = mail
         }

	}
         }
         }

sql-conf

driver = mysql connect = dbname=postfix user=root host=127.0.0.1 password=... default_pass_scheme = PLAIN password_query = SELECT password FROM mailbox WHERE username = '%u' AND active = '1' user_query = SELECT concat('/usr/local/virtual/', maildir) AS home, concat('maildir:/usr/local/virtual/', maildir) AS mail, 1001 AS uid, 1001 AS gid, concat('dirsize:storage=' , ROUND( mailbox.quota / 1024 ),':ignore=Trash') AS quota FROM mailbox WHERE username = '%u' AND active = '1'

-- Mit freundlichen Gruessen Best Regards

Robert Schetterer

https://www.schetterer.org Munich/Bavaria/Germany

Charles Marcus schrieb:

...

this are confs

dovecot -n output is desired when showing dovecot config - it shows precisely what dovecot is using, in case somehow you miss something important or dovecot is somehow using a different config file than you think it is using...

:)

sorry charles , but i know what iam using , trust me but here it comes

dovecot -n # /etc/dovecot/dovecot.conf base_dir: /var/run/dovecot/ log_path: /var/log/dovecot info_log_path: /var/log/dovecot.info protocols: imap imaps pop3 pop3s ssl_cert_file: /etc/postfix/mailserver.cert ssl_key_file: /etc/postfix/mailserver.key disable_plaintext_auth: no login_dir: /var/run/dovecot/login login_executable(default): /usr/lib/dovecot/imap-login login_executable(imap): /usr/lib/dovecot/imap-login login_executable(pop3): /usr/lib/dovecot/pop3-login login_greeting: Welcome login_process_per_connection: no login_chroot: no login_processes_count: 5 first_valid_uid: 1001 mail_extra_groups: mail default_mail_env: maildir:/usr/local/virtual/%h mail_location: maildir:/usr/local/virtual/%h mail_debug: yes mail_executable(default): /usr/lib/dovecot/imap mail_executable(imap): /usr/lib/dovecot/imap mail_executable(pop3): /usr/lib/dovecot/pop3 mail_plugins(default): quota imap_quota trash mail_plugins(imap): quota imap_quota trash mail_plugins(pop3): quota mail_plugin_dir(default): /usr/lib64/dovecot/modules/imap mail_plugin_dir(imap): /usr/lib64/dovecot/modules/imap mail_plugin_dir(pop3): /usr/lib64/dovecot/modules/pop3 pop3_uidl_format: %08Xu%08Xv auth default: verbose: yes worker_max_count: 50 passdb: driver: sql args: /etc/dovecot/dovecot-sql.conf userdb: driver: sql args: /etc/dovecot/dovecot-sql.conf userdb: driver: prefetch socket: type: listen client: path: /var/spool/postfix/private/auth mode: 432 user: postfix group: mail master: path: /var/run/dovecot/auth-master mode: 432 user: vmail group: vmail plugin: quota: maildir:ignore=Trash trash: /etc/dovecot/dovecot-trash.conf

so charles i cant see any difference to my prior mail

-- Mit freundlichen Gruessen Best Regards

Robert Schetterer

https://www.schetterer.org Munich/Bavaria/Germany

Charles Marcus schrieb:

...

...

...

this are confs

...

...

dovecot -n output is desired when showing dovecot config - it shows precisely what dovecot is using, in case somehow you miss something important or dovecot is somehow using a different config file than you think it is using...

:)

...

sorry charles , but i know what iam using , trust me

The point is, unless you are personal friends with the person you are asking (and if you are personl friends with Timo, and *he* trusts you to 'know what you are using', then I would say you shouldn't have asked on the list), trust is irrelevant.

Trust me... ;)

...

but here it comes

dovecot -n

<snip>

...

so charles i cant see any difference to my prior mail

The difference is, unless you copied/pasted badly, this output can be 'trusted' as the exact settings that dovecot is actually *using*, as opposed to a manually edited version from a .conf file that may or *may* *not* be the .conf file that dovecot is reading at startup. What if you had done something silly, and dovecot was actually using a different .conf file than you thought? It happens...

Anyway, my apologies if you were offended, as none was intended...

:)

Hi Charles, i agree, your right with posting dovecot -n is the right way ( standart way ) i simply mail that much to tec list that i sometimes miss this *g , cause mostly posted this sometime before, but my meaning is that this should not be a reason not to help each other finding bugs, last time i had to post equals confs in another list for 5 times, cause some people missed threads before *g, that was very boring

Mit freundlichen Gruessen Best Regards

Robert Schetterer

https://www.schetterer.org Munich/Bavaria/Germany

Timo Sirainen schrieb:

On Fri, 2007-02-23 at 01:16 +0100, Robert Schetterer wrote:

...

user_query = SELECT concat('/usr/local/virtual/', maildir) AS home, concat('maildir:/usr/local/virtual/', maildir) AS mail, 1001 AS uid, 1001 AS gid, concat('dirsize:storage=' , ROUND( mailbox.quota / 1024 ),':ignore=Trash') AS quota FROM mailbox WHERE username = '%u' AND active = '1'

You have dirsize here and not maildir..

Hi Timo, youre right, this fixed all problems

with dovecot-trash.conf and ignore=Trash

now everything works as described it was my fault ( big big shame on me ) as setting maildir is written in the example of the wiki

you re a hero !!!

-- Mit freundlichen Gruessen Best Regards

Robert Schetterer

https://www.schetterer.org Munich/Bavaria/Germany

Marc Perkel schrieb:

Seems to be the best release yet. System load levels are at a record low.

How's the docs coming? Definitely close to ready to be 1.0

Can people please stop prodding Timo to a 1.0 release?

It's outright annoying and offensive if many people gather in the lobby and shout for 1.0, of which many will just see the outside (how it behaves) but not the inside (the code itself).

It's 1.0 if it's done, and I hope not a day sooner.

On 23/02/2007 08:03, Bill Landry wrote:

Matthias Andree wrote the following on 2/22/2007 11:54 PM -0800:

...

It's 1.0 if it's done, and I hope not a day sooner.

For what it's worth, I couldn't agree more. Please trust that Timo will release Dovecot 1.0 when HE believes it's ready, and hopefully not before.

I agree.

Too often I have run supposed release ready software that should never have been released from beta or release candidate, and suffered the consequences because of it. Doing that also sets up the developer with a bad reputation for release buggy code, and thus everyone waits for someone else to try out new releases first before upgrading themselves.

Which is why almost nobody in business will touch Windows Vista until at least SP1, same as happened with XP.

Cheers,

John.

El Viernes, 23 de Febrero de 2007 09:04, Jay Chandler escribió:

I understand what you're saying, but realize that some of us have pointy-haired bosses who refuse to consider ANYTHING until it's hit "version 1.0."  Personally I think it's a ridiculous and arbitrary standard, but unfortunately, they sign my paychecks, so I ultimately have to keep them happy.

And so this should be the main reason for dovecot 1.0 release?

-- Joseba Torre. CIDIR Bizkaia.

Jay Chandler wrote:

Dovecot personally, but again, there are manglement pressures afoot...

Your company has a hierarchy of over-paid manglers too :)

(Sorry, made me laugh).

Dick

On Friday February 23, 2007 at 03:24:29 (AM) Jay Chandler wrote:

[snip]

I've also had to deal with the "Twenty something release candidates?
What, can't they get it right by now?" comments. I'm a big fan of Dovecot personally, but again, there are manglement pressures afoot...

Personally, it makes no difference to me what naming convention you use to describe a software realise; however, there has been chatter on some of the boards that I frequent regarding this abuse of "RC" releases. Apparently, no one has ever heard of any software ever approaching this astronomical level of 'RC' releases. It is becoming something of a joke. I did see a reference on the MailScanner boards recommending not using Dovecot simply for this reason.

It might be prudent at this point to do what Wietse just did with Postfix-2.4 and implement a code freeze. No further code is being added to the product and it is scheduled for release on or about April 1.

Just my 2¢.

-- Gerard

Bumper sticker:

To err is human, to forgive divine. 
Neither is Marine Corps policy

On Fri, 2007-02-23 at 13:04 +0100, Matthias Andree wrote:

Timo's call, particularly the part about April 1 I'd personally call too tight a deadline. The influx of changes has been orders of magnitude above Postfix's.

Also if you want 1.0 to be regarded as a stable release avoid releasing it on April the 1st ;-)

ciao

Luca

Jay Chandler schrieb:

Matthias Andree wrote:

...

It's outright annoying and offensive if many people gather in the lobby and shout for 1.0, of which many will just see the outside (how it behaves) but not the inside (the code itself).

It's 1.0 if it's done, and I hope not a day sooner.

I understand what you're saying, but realize that some of us have pointy-haired bosses who refuse to consider ANYTHING until it's hit "version 1.0." Personally I think it's a ridiculous and arbitrary standard, but unfortunately, they sign my paychecks, so I ultimately have to keep them happy.

Understood - but for those p. h. bosses in particular, a solid 1.0 is actually an advantage IMO: they don't shift their acceptance minimum version if 1.0 actually work - and slapping a 1.0 tag onto the software in whatever shape is in will just make PHB state "never install anything before 1.2"...

A real boss should state "use what works best whatever the name" and let the technical staff make the decision and not care about the reasons as long as the tech staff make reasonable decisions...

On Fri 23 Feb 2007 at 12:04AM, Jay Chandler wrote:

Matthias Andree wrote:

...

It's outright annoying and offensive if many people gather in the lobby and shout for 1.0, of which many will just see the outside (how it behaves) but not the inside (the code itself).

It's 1.0 if it's done, and I hope not a day sooner.

I understand what you're saying, but realize that some of us have pointy-haired bosses who refuse to consider ANYTHING until it's hit "version 1.0." Personally I think it's a ridiculous and arbitrary

Consider yourself lucky. A lot of pointy hairs won't consider anything until it's version 2.0.

    -dp

-- Daniel Price - Solaris Kernel Engineering - dp@eng.sun.com - blogs.sun.com/dp

Dan Price schrieb:

On Fri 23 Feb 2007 at 12:04AM, Jay Chandler wrote:

...

Matthias Andree wrote:

...

It's outright annoying and offensive if many people gather in the lobby and shout for 1.0, of which many will just see the outside (how it behaves) but not the inside (the code itself).

It's 1.0 if it's done, and I hope not a day sooner.

I understand what you're saying, but realize that some of us have pointy-haired bosses who refuse to consider ANYTHING until it's hit "version 1.0." Personally I think it's a ridiculous and arbitrary

Consider yourself lucky. A lot of pointy hairs won't consider anything until it's version 2.0.

And why's that? Most likely because they've all too often seen some 0.3 version marketed as 1.0 for funding or marketing reasons that was not ready to market at its time -- too tight deadlines promised, problem complexity underestimated -- reasons enough.

And what's the problem with PHB meddling in the affairs of his tech staff anyways? If PHB has the time to be involved in such discussions rather than handing on a set of fit criteria to the CTO, he's got too little work, with all due consequences 8-)

I'm always at a loss when end users who have rarely, if ever, seen the internal code, i. e. looked behind the scenes, make any statements to code quality, 1.0 readiness and everything.

On Fri, Feb 23, 2007 at 08:54:56AM +0100, Matthias Andree wrote:

Marc Perkel schrieb:

...

Seems to be the best release yet. System load levels are at a record low.

How's the docs coming? Definitely close to ready to be 1.0

Can people please stop prodding Timo to a 1.0 release?

It's outright annoying and offensive if many people gather in the lobby and shout for 1.0, of which many will just see the outside (how it behaves) but not the inside (the code itself).

Agreed. From the reports on this list, Dovecot still has a quite a few bugs in it. Every day I am seeing new patches for various crashes, etc.

My sincere appreciation goes to the authors, as it does for any open-source project, but this software is not release quality IMO.

-- Dean Brooks dean@iglou.com

On Fri, 23 Feb 2007, Marc Perkel wrote:

That's your opinion. Software is never "done" and therefore 1.0 is arbetrary. I disagree.

Come back if you know how to disable HTML and spell properly and lay out your phrases consistently, then I'll reconsider if I can take you serious.

It is perfectly possible to collect a set of requirements and release 1.0 when the requirements are met, and add further requirements for later versions as required/found adequate or whatever.

At any rate, it's impolite -- to put it mildly -- to pressure someone whom you haven't contracted to do anything for you, and all the more if your arguments are so weak.

-- Matthias Andree