Re: [Mailman-Users] Admin interface not recording changes (not the FAQ)

2009-09-25 Thread Mark Sapiro
Matthew Miller wrote:

>We have mailman up and running successfully, in as far as it forwards
>e-mails to the list and shows the interfaces for /listinfo and /admindb.
>However, when I go into the web interface to configure list settings, it
>accepts the input but doesn't record the change when I click Submit.  We're
>running OS X Server v10.5.8, using mailman v2.1.9.
>
>Before you leap on the "redirect" default e-mail, I have read the FAQ.  I
>_did_ have a redirect going, because our root site URL was redirecting to
>the webmail interface for the server. I have corrected that, and restarted
>both the webserver and mailman.  I'm now going directly to the site, but the
>behavior hasn't changed. (I'm sure of this, because I removed the redirect
>first to find out why it was there - the admin interface was failing because
>the root of the site was redirected to /webmail and the webadmin just
>redirect mailman back, which of course didn't work well.  That has been
>resolved so there is no root site redirection any longer, and thus the
>mailman redirect isn't needed either and was removed as well.)  I don't
>_think_ any of the other situations referred to in the FAQ are involved, but
>I'm not a UNIX/Linux guru, so I may well have missed something due to
>ignorance.


View the source of the admin page in your browser. Near the top, there
is a . What is the action URL? Is it
what you expect, i.e. the same host and path as the page itself?

-- 
Mark Sapiro The highway is for gamblers,
San Francisco Bay Area, Californiabetter use your sense - B. Dylan

--
Mailman-Users mailing list
Mailman-Users@python.org
http://mail.python.org/mailman/listinfo/mailman-users
Mailman FAQ: http://wiki.list.org/x/AgA3
Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
Unsubscribe: 
http://mail.python.org/mailman/options/mailman-users/archive%40jab.org

Security Policy: http://wiki.list.org/x/QIA9


Re: [Mailman-Users] bouce processing problem

2009-09-25 Thread Mark Sapiro
Dimitrios Karapiperis 

>verp enabled lsit -> test
>maximum member bounce score 2.0
>The number of days after which a member's bounce information is 
>discarded, if no new bounces have been received in the interim  7
>How many /Your Membership Is Disabled/ warnings disabled member should 
>get before their address is removed from the mailing list. 0
>The number of days between sending the /Your Membership Is Disabled/ 
>warnings  0
>
>
>non-verp list -> test12
>maximum member bounce score 2.0
>The number of days after which a member's bounce information is 
>discarded, if no new bounces have been received in the interim  7
>How many /Your Membership Is Disabled/ warnings a disabled member should 
>get before their address is removed from the mailing list. 1
>The number of days between sending the /Your Membership Is Disabled/ 
>warnings  0


This is inconsistent. This says you are going to disable delivery to
the member and send the member one warning notice, but then you are
going to remove the member from the list after 0 days, effectively
giving the member no time to respond to the warning if it is received.

>Logs
>Do you mean /var/log/mailman/bounce?

Yes.


>It catches the bounce but it doesn't increase the bounce score for the 
>e-mail address in question.
>
>Sep 25 18:43:54 2009 (5993) test: la...@domain.gr already scored a 
>bounce for date 25-Sep-2009
>Sep 25 18:43:55 2009 (5775)  processing 
>1 queued   bounces
>Sep 25 18:43:55 2009 (5775) test: c...@domain.gr current bounce 
>score:2.0
>Sep 25 18:43:55 2009 (5775)  processing 
>1 queued  bounces
>Sep 25 18:43:55 2009 (5775) test: c...@domain.gr already scored a bounce 
>for date 25-Sep-2009
>Sep 25 18:58:55 2009 (22975)  
>processing 1 queued bounces
>Sep 25 18:58:55 2009 (22975) test12: la...@domain.gr current bounce 
>score:  2.0
>Sep 25 18:58:55 2009 (22975)  
>processing 1 queued bounces
>Sep 25 18:58:55 2009 (22975) test12: la...@domain.gr already scored a 
>bounce for date 25-Sep-2009


Are these all the messages in that portion of the log? There should be
a "disabling" entry after the 

Sep 25 18:58:55 2009 (22975) test12: la...@domain.gr current bounce
score:  2.0

since the threshold for that list is 2.0.

What is in Mailman's error log (/var/log/mailman/error)?

If no exception occurred in bounce processing, there should at a
minimum be a "sending probe" log message or a "disabling" log message
following the message where the score reaches the threshold. If an
exception occurred, it should be logged in the error log.

The only other possibility I see is that somehow the
bounce_score_threshold for these lists is set to the string "2.0"
rather than the floating point value 2.0. Then bounce processing will
not see that 2.0 is >= "2.0".

-- 
Mark Sapiro The highway is for gamblers,
San Francisco Bay Area, Californiabetter use your sense - B. Dylan

--
Mailman-Users mailing list
Mailman-Users@python.org
http://mail.python.org/mailman/listinfo/mailman-users
Mailman FAQ: http://wiki.list.org/x/AgA3
Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
Unsubscribe: 
http://mail.python.org/mailman/options/mailman-users/archive%40jab.org

Security Policy: http://wiki.list.org/x/QIA9


Re: [Mailman-Users] digests without 'message #n..."

2009-09-25 Thread Mark Sapiro
Rob Lingelbach wrote:
>
>on further investigation, it appears Apple's Mail.app is behaving  
>unexpectedly
>with MIME format digests, when compared to mail(1), mutt(1), and the  
>web-based
>Squirrelmail.   Mail.app doesn't present separators in any predictable  
>fashion, and
>includes huge amounts of white space, with no individually-viewable  
>message option,
>as far as I can tell.


Apple's Mail.app is designed to allow you to compose a plain text
message and to drag and drop almost anything into it. Unlike other
MUAs which do this by creating a multipart/related message with a
text/html part that references other 'attached' parts by Content-ID,
Mail.app just creates a multipart/mixed message with a bunch of
text/plain parts interspersed with image/jpeg, etc. parts, so when it
receives a multipart message, it essentially just concatenates all the
part contents inline.

This works well for mail it created, but not for standards compliant
mail, and it doesn't always work well when it's messages are viewed
with another MUA.




>Incidentally Squirrelmail, my last-resort reader, presents the  
>messages quite nicely
>as a formatted list of attachments.  I have a feeling mutt(1) could do  
>the same thing
>with a bit of tweaking.


I use the 'v' command in mutt to see the attachment list. Then I can
select an individual message (attachment), open it, read it and reply
to it as if I had received the individual mail.

-- 
Mark Sapiro The highway is for gamblers,
San Francisco Bay Area, Californiabetter use your sense - B. Dylan

--
Mailman-Users mailing list
Mailman-Users@python.org
http://mail.python.org/mailman/listinfo/mailman-users
Mailman FAQ: http://wiki.list.org/x/AgA3
Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
Unsubscribe: 
http://mail.python.org/mailman/options/mailman-users/archive%40jab.org

Security Policy: http://wiki.list.org/x/QIA9


Re: [Mailman-Users] digests without 'message #n..."

2009-09-25 Thread Rob Lingelbach


On Sep 24, 2009, at 5:13 PM, Mark Sapiro wrote:


Plain format digests have messages inline separated by the RFC 1153
separator line (30 hyphens) followed by selected headers and the
scrubbed message body.

MIME format digests have each message as a separate MIME message/ 
rfc822

part with partial headers and the MIME message body.

The "Message: n" line is aadded by digest processing as a message
header. Which headers appear in plain and MIME format digests is
controled by the settings MIME_DIGEST_KEEP_HEADERS and
PLAIN_DIGEST_KEEP_HEADERS, both of which include Message: by default.

However, what you actually see in the MIME format digest depends  
almost

entirely on the MUA used to view it.


thank you for the detailed reply.

on further investigation, it appears Apple's Mail.app is behaving  
unexpectedly
with MIME format digests, when compared to mail(1), mutt(1), and the  
web-based
Squirrelmail.   Mail.app doesn't present separators in any predictable  
fashion, and
includes huge amounts of white space, with no individually-viewable  
message option,

as far as I can tell.

Incidentally Squirrelmail, my last-resort reader, presents the  
messages quite nicely
as a formatted list of attachments.  I have a feeling mutt(1) could do  
the same thing

with a bit of tweaking.

--  
Rob Lingelbach

r...@colorist.org










--
Mailman-Users mailing list
Mailman-Users@python.org
http://mail.python.org/mailman/listinfo/mailman-users
Mailman FAQ: http://wiki.list.org/x/AgA3
Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
Unsubscribe: 
http://mail.python.org/mailman/options/mailman-users/archive%40jab.org

Security Policy: http://wiki.list.org/x/QIA9


Re: [Mailman-Users] bouce processing problem

2009-09-25 Thread Mark Sapiro
Dimitrios Karapiperis wrote:
>
>Every day I create a bounce and the result is the same. It scored 1.0 
>and never increases again.
>I test it with VEPR and with the usual sending way.


What is in Mailman's bounce log?

What are the list's bounce processing settings?


>what is the last notice thing?
>
>last notice date: (1970, 1, 2)

That is the date that the last "disabled warning" was sent. Since no
notice has been sent yet, it shows the 'initialized' date which is
intended to be the first day of the epoch that contained a midnight -
1970-January-2.

-- 
Mark Sapiro The highway is for gamblers,
San Francisco Bay Area, Californiabetter use your sense - B. Dylan

--
Mailman-Users mailing list
Mailman-Users@python.org
http://mail.python.org/mailman/listinfo/mailman-users
Mailman FAQ: http://wiki.list.org/x/AgA3
Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/
Unsubscribe: 
http://mail.python.org/mailman/options/mailman-users/archive%40jab.org

Security Policy: http://wiki.list.org/x/QIA9