From take at nerd.fi  Thu Oct  1 21:20:59 2015
From: take at nerd.fi (Take)
Date: Thu, 01 Oct 2015 22:20:59 +0300
Subject: [Mailman-Developers] Problems with mailman3 installation via bundler
Message-ID: <560D879B.9080506@nerd.fi>

Hello,

I already posted details from my experiences with mailman3 to 
users-lists where Mark Sapiro suggested that I should post my questions 
to dev-list instead.

So, long story short, I installed mailman3 via bundler and can access 
both postorius and hyperkitty via web browser. Additionally I can both 
join and leave the list, access help via email etc, so installation 
seems to be in order.

However, I'm in a situation where posts I send don't end up on mailing 
list and replies end up shunted. On the log file I get this:

Oct 01 22:03:21 2015 (12225) SHUNTING: 
1443726201.2465158+c8d546daa806696842da9fd07724ed6738b30854
Oct 01 22:03:21 2015 (12225) Uncaught runner exception:
Oct 01 22:03:21 2015 (12225) Traceback (most recent call last):
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/core/runner.py", 
line 165, in _one_iteration
     self._process_one_file(msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/core/runner.py", 
line 258, in _process_one_file
     keepqueued = self._dispose(mlist, msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/runners/incoming.py", 
line 63, in _dispose
     process(mlist, msg, msgdata, start_chain)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/core/chains.py", 
line 86, in process
     link.function(mlist, msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/chains/hold.py", 
line 182, in _process
     autorespond_to_sender(mlist, msg.sender, language)):
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/chains/hold.py", 
line 83, in autorespond_to_sender
     address = user_manager.create_address(sender)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/database/transaction.py", 
line 92, in wrapper
     return function(args[0], config.db.store, *args[1:], **kws)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/model/usermanager.py", 
line 110, in create_address
     address = Address(email, display_name)
   File "<string>", line 4, in __init__
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/sqlalchemy/orm/state.py", 
line 306, in _initialize_instance
     manager.dispatch.init_failure(self, args, kwargs)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/sqlalchemy/util/langhelpers.py", 
line 60, in __exit__
     compat.reraise(exc_type, exc_value, exc_tb)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/sqlalchemy/util/compat.py", 
line 182, in reraise
     raise value
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/sqlalchemy/orm/state.py", 
line 303, in _initialize_instance
     return manager.original_init(*mixed[1:], **kwargs)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/model/address.py", 
line 59, in __init__
     getUtility(IEmailValidator).validate(email)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/email/validate.py", 
line 65, in validate
     raise InvalidEmailAddressError(email)
mailman.interfaces.address.InvalidEmailAddressError


On validate.py I added logging to see what seems to be the problem with 
address, to validate-function:

from distutils import log
log.error("Validating address -> " + email)
if not self.is_valid(email):
...

It appears that variable email is empty at that point, for some reason. 
I tried to search the problem via google but came up empty. At this 
point I'd be willing to say that I have encountered a bug, but I'm not 
competent enough with python to actually verify the situation.

-- 
Take

From take at nerd.fi  Fri Oct  2 00:05:40 2015
From: take at nerd.fi (Take)
Date: Fri, 02 Oct 2015 01:05:40 +0300
Subject: [Mailman-Developers] Problems with mailman3 installation via
 bundler
In-Reply-To: <560D879B.9080506@nerd.fi>
References: <560D879B.9080506@nerd.fi>
Message-ID: <560DAE34.4030501@nerd.fi>


On 10/01/2015 10:20 PM, Take wrote:
> On validate.py I added logging to see what seems to be the problem with
> address, to validate-function:
>
> from distutils import log
> log.error("Validating address -> " + email)
> if not self.is_valid(email):
> ...
>
> It appears that variable email is empty at that point, for some reason.

This is somewhat strange. I tried to dig deeper and learn from the 
system, so I just assigned valid address to email variable just to see 
what would happen.

Shunted messages were bounced messages from postfix, caused by 
misconfiguration and loop in mail transport, but that's not the strange 
part. What's strange that I still have one shunted message and log 
suggests encoding issues:


Oct 02 00:42:38 2015 (16593) Traceback (most recent call last):
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/core/runner.py", 
line 165, in _one_iteration
     self._process_one_file(msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/core/runner.py", 
line 258, in _process_one_file
     keepqueued = self._dispose(mlist, msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/runners/outgoing.py", 
line 97, in _dispose
     self._func(mlist, msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/mta/deliver.py", 
line 90, in deliver
     refused = agent.deliver(mlist, msg, msgdata)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/mta/bulk.py", 
line 105, in deliver
     mlist, msg, msgdata, recipients)
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/mta/base.py", 
line 73, in _deliver_to_recipients
     sender, recipients, msg.as_string())
   File 
"/home/mailman/mailman-bundler/venv-3.4/lib/python3.4/site-packages/mailman/mta/connection.py", 
line 87, in sendmail
     results = self._connection.sendmail(envsender, recipients, msgtext)
   File "/usr/lib/python3.4/smtplib.py", line 765, in sendmail
     msg = _fix_eols(msg).encode('ascii')
UnicodeEncodeError: 'ascii' codec can't encode character '\xe4' in 
position 1460: ordinal not in range(128)
Oct 02 00:42:38 2015 (16593) SHUNTING: 
1443735758.2455976+005fdccae22be4461400beae5aa37465af396554

Somehow I'd expect that those were thing of the past, but apparently 
they aren't. There's some mentions about similar problems on 
smtp-libraries with python on the internet, but, as mentioned, I'm not 
that good with python that I could say anything certain.

-- 
Take

From fsantiago at garbage-juice.com  Tue Oct  6 22:00:06 2015
From: fsantiago at garbage-juice.com (fsantiago at garbage-juice.com)
Date: Tue, 06 Oct 2015 16:00:06 -0400
Subject: [Mailman-Developers] issues with MM3 / Apache mod_wsgi errors
Message-ID: <13ff7df14e034b4676348ddad8ed5708@garbage-juice.com>

Hello all,

Mark Sapiro helped me with some basic config issues earlier but now i'm 
stuck here; my web browser shows me:

"Service Unavailable

The server is temporarily unable to service your request due to 
maintenance downtime or capacity problems. Please try again later."


my server's apache postorius error log shows me:

[Tue Oct 06 15:52:34.961249 2015] [:error] [pid 11691] (13)Permission 
denied: [client <IP>:42590] mod_wsgi (pid=11691): Unable to connect to 
WSGI daemon process 'hyperkitty' on '/run/httpd/wsgi.11685.0.1.sock' 
after multiple attempts.

this occurs when i try to access:
/hyperkitty/

if i try /postorius/, i see same browser result and similar apache log 
result:

[Tue Oct 06 15:58:53.355159 2015] [:error] [pid 11693] (13)Permission 
denied: [client <IP>:42654] mod_wsgi (pid=11693): Unable to connect to 
WSGI daemon process 'postorius' on '/run/httpd/wsgi.11685.0.2.sock' 
after multiple attempts.


MM3 components installed from hyperkitty-el.repo and using apache 2.4.6 
on centos 7. python 2.7 and 3.4 installed.

What's missing here? Thanks all.


- Fabe S.



From simon.hanna at serve-me.info  Wed Oct  7 02:53:48 2015
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Wed, 7 Oct 2015 02:53:48 +0200
Subject: [Mailman-Developers] issues with MM3 / Apache mod_wsgi errors
In-Reply-To: <13ff7df14e034b4676348ddad8ed5708@garbage-juice.com>
References: <13ff7df14e034b4676348ddad8ed5708@garbage-juice.com>
Message-ID: <20151007005348.GA21991@turing>

 
> my server's apache postorius error log shows me:
> 
> [Tue Oct 06 15:52:34.961249 2015] [:error] [pid 11691] (13)Permission
> denied: [client <IP>:42590] mod_wsgi (pid=11691): Unable to connect to WSGI
> daemon process 'hyperkitty' on '/run/httpd/wsgi.11685.0.1.sock' after
> multiple attempts.
 
> [Tue Oct 06 15:58:53.355159 2015] [:error] [pid 11693] (13)Permission
> denied: [client <IP>:42654] mod_wsgi (pid=11693): Unable to connect to WSGI
> daemon process 'postorius' on '/run/httpd/wsgi.11685.0.2.sock' after
> multiple attempts.
> 
> 
> MM3 components installed from hyperkitty-el.repo and using apache 2.4.6 on
> centos 7. python 2.7 and 3.4 installed.

Are you sure the sockets exist? If they exist, what are their permissions?
Do they belong to the same user as apache is using?

You might want to look at the wsgi logs as well to see if the python application is the culprit.

You can check out my deployment instructions [1] using nginx and uwsgi.
It might help you with debugging the issue


[1]: https://gitlab.com/thelinuxguy/mailman-suite
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20151007/ebf8b3e5/attachment.sig>

From fsantiago at deviltracks.net  Wed Oct  7 03:21:58 2015
From: fsantiago at deviltracks.net (=?utf-8?q?fsantiago=40deviltracks=2Enet?=)
Date: Tue, 06 Oct 2015 21:21:58 -0400
Subject: [Mailman-Developers] issues with MM3 / Apache mod_wsgi errors
In-Reply-To: <20151007005348.GA21991@turing>
References: <13ff7df14e034b4676348ddad8ed5708@garbage-juice.com>
 <20151007005348.GA21991@turing>
Message-ID: <66F85E9C-050F-4E3B-98D5-54695175002A@deviltracks.net>

Excellent. Thank you. I'll give this a shot. 

Sincerely,

Fabian Santiago

Sent from my iPhone

> On Oct 6, 2015, at 8:53 PM, Simon Hanna <simon.hanna at serve-me.info> wrote:
> 
> 
>> my server's apache postorius error log shows me:
>> 
>> [Tue Oct 06 15:52:34.961249 2015] [:error] [pid 11691] (13)Permission
>> denied: [client <IP>:42590] mod_wsgi (pid=11691): Unable to connect to WSGI
>> daemon process 'hyperkitty' on '/run/httpd/wsgi.11685.0.1.sock' after
>> multiple attempts.
> 
>> [Tue Oct 06 15:58:53.355159 2015] [:error] [pid 11693] (13)Permission
>> denied: [client <IP>:42654] mod_wsgi (pid=11693): Unable to connect to WSGI
>> daemon process 'postorius' on '/run/httpd/wsgi.11685.0.2.sock' after
>> multiple attempts.
>> 
>> 
>> MM3 components installed from hyperkitty-el.repo and using apache 2.4.6 on
>> centos 7. python 2.7 and 3.4 installed.
> 
> Are you sure the sockets exist? If they exist, what are their permissions?
> Do they belong to the same user as apache is using?
> 
> You might want to look at the wsgi logs as well to see if the python application is the culprit.
> 
> You can check out my deployment instructions [1] using nginx and uwsgi.
> It might help you with debugging the issue
> 
> 
> [1]: https://gitlab.com/thelinuxguy/mailman-suite
> 
> _______________________________________________
> Mailman-Developers mailing list
> Mailman-Developers at python.org
> https://mail.python.org/mailman/listinfo/mailman-developers
> Mailman FAQ: http://wiki.list.org/x/AgA3
> Searchable Archives: http://www.mail-archive.com/mailman-developers%40python.org/
> Unsubscribe: https://mail.python.org/mailman/options/mailman-developers/fsantiago%40deviltracks.net
> 
> Security Policy: http://wiki.list.org/x/QIA9



From fmouse at fmp.com  Wed Oct  7 05:49:19 2015
From: fmouse at fmp.com (Lindsay Haisley)
Date: Tue, 06 Oct 2015 22:49:19 -0500
Subject: [Mailman-Developers] substitution variables
Message-ID: <1444189759.113685.42.camel@pudina.fmp.com>

I've hacked per-post AES encryption of subscriber addresses into
Decorate.py so that I have a subscriber address identifier available for
situations where subscribers have lost redirections in force and can't
get headers to me.

Everything works, and I can use the substitution %(user_skey)s in the
footer, but I get a nasty warning on the Non-digest options page,
"Warning: The following illegal substitution variables were found in the
msg_footer string: user_skey".

Where in the code can I list this substitution variable so that it
doesn't throw this error on this page?

-- 
Lindsay Haisley       | "The only unchanging certainty
FMP Computer Services |    is the certainty of change"
512-259-1190          |
http://www.fmp.com    | - Ancient wisdom, all cultures


From mark at msapiro.net  Wed Oct  7 16:32:11 2015
From: mark at msapiro.net (Mark Sapiro)
Date: Wed, 7 Oct 2015 07:32:11 -0700
Subject: [Mailman-Developers] substitution variables
In-Reply-To: <1444189759.113685.42.camel@pudina.fmp.com>
References: <1444189759.113685.42.camel@pudina.fmp.com>
Message-ID: <56152CEB.4020909@msapiro.net>

On 10/06/2015 08:49 PM, Lindsay Haisley wrote:
> 
> Where in the code can I list this substitution variable so that it
> doesn't throw this error on this page?


Add it to the definition of PERSONALIZED_ALLOWEDS in
Mailman/Gui/NonDigest.py.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan

From fmouse at fmp.com  Wed Oct  7 17:23:52 2015
From: fmouse at fmp.com (Lindsay Haisley)
Date: Wed, 07 Oct 2015 10:23:52 -0500
Subject: [Mailman-Developers] substitution variables
In-Reply-To: <56152CEB.4020909@msapiro.net>
References: <1444189759.113685.42.camel@pudina.fmp.com>
 <56152CEB.4020909@msapiro.net>
Message-ID: <1444231432.113685.56.camel@pudina.fmp.com>

On Wed, 2015-10-07 at 07:32 -0700, Mark Sapiro wrote:
> On 10/06/2015 08:49 PM, Lindsay Haisley wrote:
> > 
> > Where in the code can I list this substitution variable so that it
> > doesn't throw this error on this page?
> 
> 
> Add it to the definition of PERSONALIZED_ALLOWEDS in
> Mailman/Gui/NonDigest.py.

And then turn on OWNERS_CAN_ENABLE_PERSONALIZATION in mm_cfg.py and
enable personalization for the list, yes?

-- 
Lindsay Haisley       | "The only unchanging certainty
FMP Computer Services |    is the certainty of change"
512-259-1190          |
http://www.fmp.com    | - Ancient wisdom, all cultures


From mark at msapiro.net  Thu Oct  8 02:13:51 2015
From: mark at msapiro.net (Mark Sapiro)
Date: Wed, 7 Oct 2015 17:13:51 -0700
Subject: [Mailman-Developers] substitution variables
In-Reply-To: <1444231432.113685.56.camel@pudina.fmp.com>
References: <1444189759.113685.42.camel@pudina.fmp.com>
 <56152CEB.4020909@msapiro.net> <1444231432.113685.56.camel@pudina.fmp.com>
Message-ID: <5615B53F.2040603@msapiro.net>

On 10/07/2015 08:23 AM, Lindsay Haisley wrote:
> On Wed, 2015-10-07 at 07:32 -0700, Mark Sapiro wrote:
>> On 10/06/2015 08:49 PM, Lindsay Haisley wrote:
>>>
>>> Where in the code can I list this substitution variable so that it
>>> doesn't throw this error on this page?
>>
>>
>> Add it to the definition of PERSONALIZED_ALLOWEDS in
>> Mailman/Gui/NonDigest.py.
> 
> And then turn on OWNERS_CAN_ENABLE_PERSONALIZATION in mm_cfg.py and
> enable personalization for the list, yes?


Well, I assumed you'd already done that since I don't see how you can
include a recipient unique token in the msg_footer if delivery isn't
personalized.

However, all you need is for the list's delivery is to be personalized.
Whether you accomplish this by turning on
OWNERS_CAN_ENABLE_PERSONALIZATION and setting personalize via the GUI or
setting personalize some other way, e.g. via withlist, that's all you need.

If for some reason you're not personalizing, but ensuring a single
recipient per SMTP transaction by setting VERP_DELIVERY_INTERVAL = 1
then yes, you would need to either set the list's personalize attribute
non-zero or you could add your variable name to ALLOWEDS in
Mailman/Gui/Digest.py, but setting personalize is the right way to do it.

If the list's delivery is not personalized and VERP_DELIVERY_INTERVAL is
or becomes unequal 1 (and SMTP_MAX_RCPTS is unequal 1) you will wind up
decorating multiple recipients messages with the same data.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan

From fmouse at fmp.com  Thu Oct  8 02:21:25 2015
From: fmouse at fmp.com (Lindsay Haisley)
Date: Wed, 07 Oct 2015 19:21:25 -0500
Subject: [Mailman-Developers] substitution variables
In-Reply-To: <5615B53F.2040603@msapiro.net>
References: <1444189759.113685.42.camel@pudina.fmp.com>
 <56152CEB.4020909@msapiro.net> <1444231432.113685.56.camel@pudina.fmp.com>
 <5615B53F.2040603@msapiro.net>
Message-ID: <1444263685.113685.192.camel@pudina.fmp.com>

On Wed, 2015-10-07 at 17:13 -0700, Mark Sapiro wrote:
> If the list's delivery is not personalized and VERP_DELIVERY_INTERVAL is
> or becomes unequal 1 (and SMTP_MAX_RCPTS is unequal 1) you will wind up
> decorating multiple recipients messages with the same data.

Thanks, Mark. I got it working just fine. The KISS principle is best, so
I just turned on list owner personalization and went from there. This
wasn't required for the solution I developed for AOL spam feedback
reports, which involved putting an AES crypt in the Resent-Message-ID
header. Only VERP personalization was needed for that.

The personalization performance hit was already on deck by enabling VERP
addresses on my lists, so I don't reckon that putting this personalized
crypt string in the footers will increase it much.

-- 
Lindsay Haisley       | "The only unchanging certainty
FMP Computer Services |    is the certainty of change"
512-259-1190          |
http://www.fmp.com    | - Ancient wisdom, all cultures


From programo.sapien at gmail.com  Fri Oct  9 08:40:29 2015
From: programo.sapien at gmail.com (Aswin kumar)
Date: Fri, 9 Oct 2015 01:40:29 -0500
Subject: [Mailman-Developers] Off topic : Regarding Python
Message-ID: <CAJ0JsiB+padWeK-xbjM+ytDgbofttJnzJK+eGCfeoivxf5DtGg@mail.gmail.com>

Hi,

I am new to python programming language and I saw that there are two
versions python2 and python3. Which one should I start learning ? Should I
learn python2 first and then switch to python 3? I am asking this question
here as maiman3 has switched to python3 , but some other projects still
stick to python2, so I am kind of confused which version to start with.

Regards,
Aswin

From mark at msapiro.net  Fri Oct  9 18:17:50 2015
From: mark at msapiro.net (Mark Sapiro)
Date: Fri, 9 Oct 2015 09:17:50 -0700
Subject: [Mailman-Developers] Off topic : Regarding Python
In-Reply-To: <CAJ0JsiB+padWeK-xbjM+ytDgbofttJnzJK+eGCfeoivxf5DtGg@mail.gmail.com>
References: <CAJ0JsiB+padWeK-xbjM+ytDgbofttJnzJK+eGCfeoivxf5DtGg@mail.gmail.com>
Message-ID: <5617E8AE.90203@msapiro.net>

On 10/08/2015 11:40 PM, Aswin kumar wrote:
> 
> I am new to python programming language and I saw that there are two
> versions python2 and python3. Which one should I start learning ? Should I
> learn python2 first and then switch to python 3? I am asking this question
> here as maiman3 has switched to python3 , but some other projects still
> stick to python2, so I am kind of confused which version to start with.


Definitely start with Python 3.

There is still much legacy Python 2 code, and many projects have a hard
time porting to Python 3 because of dependencies on other packages which
are still Python 2, but the future is Python 3, and any new code is best
as Python 3.

The only reason to learn Python 2 at this point is if you are tasked
with supporting a legacy Python 2 project.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan

From barry at list.org  Fri Oct  9 19:30:29 2015
From: barry at list.org (Barry Warsaw)
Date: Fri, 9 Oct 2015 13:30:29 -0400
Subject: [Mailman-Developers] Off topic : Regarding Python
In-Reply-To: <5617E8AE.90203@msapiro.net>
References: <CAJ0JsiB+padWeK-xbjM+ytDgbofttJnzJK+eGCfeoivxf5DtGg@mail.gmail.com>
 <5617E8AE.90203@msapiro.net>
Message-ID: <20151009133029.37650048@limelight.wooz.org>

On Oct 09, 2015, at 09:17 AM, Mark Sapiro wrote:

>Definitely start with Python 3.

Yes, definitely.  If you know Python 3, it's also not that difficult to learn
the differences that matter when you have to deal with Python 2.

Cheers,
-Barry

From andrew at hodgsonfamily.org  Fri Oct 16 23:03:10 2015
From: andrew at hodgsonfamily.org (Andrew Hodgson)
Date: Fri, 16 Oct 2015 21:03:10 +0000
Subject: [Mailman-Developers] Initial experiments with Mailman 3/Buildout
Message-ID: <DB5PR04MB147855B09233033475DFDFB6C53D0@DB5PR04MB1478.eurprd04.prod.outlook.com>

Hi,

I have looked in several list archives for different approaches for Mailman 3 installation, and am trying a few things.  I have got a new Debian 8.2 install, and installed the Python 2 and 3 developer tools etc.

I am making a virtual environment in /opt/mailman.

When running buildout, I get up until this point:

[?]
Getting distribution for 'djangorestframework>=3.0.0'.
warning: no previously-included files matching '__pycache__' found under directory '*'
warning: no previously-included files matching '*.py[co]' found under directory '*'
Got djangorestframework 3.2.4.
Getting distribution for 'django-gravatar2>=1.0.6'.
zip_safe flag not set; analyzing archive contents...
Got django-gravatar2 1.3.0.
While:
 Installing mailman-web.
Error: There is a version conflict.
We already have: pbr 1.8.1
but lockfile 0.11.0 requires 'pbr!=0.7,<1.0,>=0.6'.

Any ideas as to how to resolve?

Thanks.
Andrew.

From simon.hanna at serve-me.info  Sat Oct 17 01:26:25 2015
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Sat, 17 Oct 2015 01:26:25 +0200
Subject: [Mailman-Developers] Initial experiments with Mailman 3/Buildout
In-Reply-To: <DB5PR04MB147855B09233033475DFDFB6C53D0@DB5PR04MB1478.eurprd04.prod.outlook.com>
References: <DB5PR04MB147855B09233033475DFDFB6C53D0@DB5PR04MB1478.eurprd04.prod.outlook.com>
Message-ID: <20151016232624.GA26915@turing>

>  Installing mailman-web.
> Error: There is a version conflict.
> We already have: pbr 1.8.1
> but lockfile 0.11.0 requires 'pbr!=0.7,<1.0,>=0.6'.
It looks like lockfile does not actually need a specific version of pbr.
https://bugs.launchpad.net/pylockfile/+bug/1506679

not sure though since I manually installed mailman
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20151017/44aff4f6/attachment.sig>

From johnl at taugh.com  Tue Oct 20 23:22:22 2015
From: johnl at taugh.com (John Levine)
Date: 20 Oct 2015 21:22:22 -0000
Subject: [Mailman-Developers] New DMARC workaround hack
Message-ID: <20151020212222.1455.qmail@ary.lan>

Take a look at these:

https://tools.ietf.org/html/draft-andersen-arc-00
https://tools.ietf.org/html/draft-jones-arc-usage-00

These add new headers based on DKIM that document a chain of custody
for mail.  It would let mail providers accept list mail that DMARC
would otherwise reject.  The drafts are a mess (the XML versions are
easier to read than the txt), but I think the design is sound.  They
are intended to be implemented via upgraded DKIM libraries, and the
idea is to allow lists to work like they always have, without any
changes visible to users.  You can intuit from the authors who's
likely to implement it.

It'd be helpful for people to look at it and write to the contact
address in the first draft and tell them how likely it is that Mailman
would support it, or what problems it would present.

R's,
John

From sumanah at panix.com  Mon Oct 26 11:04:51 2015
From: sumanah at panix.com (Sumana Harihareswara)
Date: Mon, 26 Oct 2015 11:04:51 -0400
Subject: [Mailman-Developers] Mozilla funding opportunity for Mailman work
Message-ID: <562E4113.9080004@panix.com>

Hi Cabal. I'm sorry that I've neglected Mailman so much for the past
several months; I've just registered for PyCon 2016 and will be
ramping up my participation especially leading up to that.

Mozilla
https://blog.mozilla.org/blog/2015/10/23/mozilla-launches-open-source-su
pport-program/
has just announced that it wants to "identify up to 10 projects we
rely on and can fund in a thoughtful, meaningful way by December
12th". They use Mailman and if we can identify useful ways they can
spend that money to advance Mailman's development and usefulness to
them (e.g., paying for someone's time to get packaging or localization
to a better state), I'd be happy to take point on organizing that
proposal.

Again, regrets on my absence.
-- 
Sumana Harihareswara
http://brainwane.net

From barry at list.org  Tue Oct 27 18:52:21 2015
From: barry at list.org (Barry Warsaw)
Date: Tue, 27 Oct 2015 18:52:21 -0400
Subject: [Mailman-Developers] Mozilla funding opportunity for Mailman
 work
In-Reply-To: <562E4113.9080004@panix.com>
References: <562E4113.9080004@panix.com>
Message-ID: <20151027185221.508cebe8@limelight.wooz.org>

On Oct 26, 2015, at 11:04 AM, Sumana Harihareswara wrote:

>Mozilla
>https://blog.mozilla.org/blog/2015/10/23/mozilla-launches-open-source-su
>pport-program/ has just announced that it wants to "identify up to 10
>projects we rely on and can fund in a thoughtful, meaningful way by December
>12th". They use Mailman and if we can identify useful ways they can spend
>that money to advance Mailman's development and usefulness to them (e.g.,
>paying for someone's time to get packaging or localization to a better
>state), I'd be happy to take point on organizing that proposal.

I noticed that Mailman is on their list, although they haven't identified a
Mozilla contact person for us:

https://wiki.mozilla.org/MOSS/Projects_in_use_by_Mozilla

I think it would be very interesting for us to apply, although I'd like to
know what they are interested in getting out of Mailman (presumably v3).  What
would *we* like to do that we can't do because of resources?

Ideas very welcome.

Cheers,
-Barry

From stephen at xemacs.org  Wed Oct 28 00:30:57 2015
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: Wed, 28 Oct 2015 13:30:57 +0900
Subject: [Mailman-Developers] Mozilla funding opportunity for Mailman
 work
In-Reply-To: <20151027185221.508cebe8@limelight.wooz.org>
References: <562E4113.9080004@panix.com>
 <20151027185221.508cebe8@limelight.wooz.org>
Message-ID: <22064.20353.452872.487566@turnbull.sk.tsukuba.ac.jp>

Barry Warsaw writes:

 > What would *we* like to do that we can't do because of resources?

That's where I'm blocked.  I can't honestly put myself up as a
money-sink: I wouldn't mind getting paid for my effort, but I wouldn't
put in any less effort just because I'm unpaid. ;-)

Hm ... maybe generalizing Andrew's authn/authz stuff?  Mozilla would
probably like improved Persona support?  Andrew would be good for that
and ISTR he's a freelancer.

I wouldn't mind compensating a student even if she/he would do the
work for free anyway.  But except for Abhilash, I can't recommend any
recent student for a project I think is valuable enough to request
such resources without at the same time volunteering to mentor
intensively.  Not enough experience, and the project I have in mind is
not an extension of any GSoC project.  And I can't really volunteer at
this point.

As for Abhilash himself, I think he should concentrate on his graduate
studies.  There's no problem with him volunteering the occasional
half-day to Mailman, but I've seen far too many students seriously
hindered in getting the most out of their graduate study by committing
to consulting work.  I won't oppose him if he wants to do something,
but I won't nominate him.  (@Abhilash: If you decide to apply against
my advice, it's just advice.  I'll be happy to support you to the
extent I can in your Mailman work or your graduate study, whatever you
want to do -- though I suspect I'd be useless in the latter role. ;-)

If somebody else is willing to mentor *my* suggestions: there are a
couple of new proposals for DMARC mitigation out there.  I would be
willing to evaluate them (in the remaining hours) for use in Mailman.
Why "evaluate"?  As most of you probably know, message signing and
signature validation are normally considered MTA functions, and
Mailman needn't/shouldn't try to duplicate that functionality.
However, with experimental protocols, of course they won't be in
production MTAs for a while, so it *might* be worthwhile to offer
support in Mailman as a spur/temporary workaround while the MTAs
implement the new protocols.

I don't (yet) have a developer to recommend, though.

Steve


From raj.abhilash1 at gmail.com  Wed Oct 28 03:01:53 2015
From: raj.abhilash1 at gmail.com (Abhilash Raj)
Date: Wed, 28 Oct 2015 00:01:53 -0700
Subject: [Mailman-Developers] Mozilla funding opportunity for Mailman
 work
In-Reply-To: <22064.20353.452872.487566@turnbull.sk.tsukuba.ac.jp>
References: <562E4113.9080004@panix.com>
 <20151027185221.508cebe8@limelight.wooz.org>
 <22064.20353.452872.487566@turnbull.sk.tsukuba.ac.jp>
Message-ID: <563072E1.70504@gmail.com>


On 10/27/2015 09:30 PM, Stephen J. Turnbull wrote:
> Barry Warsaw writes:
> 
>  > What would *we* like to do that we can't do because of resources?
> 
> That's where I'm blocked.  I can't honestly put myself up as a
> money-sink: I wouldn't mind getting paid for my effort, but I wouldn't
> put in any less effort just because I'm unpaid. ;-)
> 
> Hm ... maybe generalizing Andrew's authn/authz stuff?  Mozilla would
> probably like improved Persona support?  Andrew would be good for that
> and ISTR he's a freelancer.
> 
> I wouldn't mind compensating a student even if she/he would do the
> work for free anyway.  But except for Abhilash, I can't recommend any
> recent student for a project I think is valuable enough to request
> such resources without at the same time volunteering to mentor
> intensively.  Not enough experience, and the project I have in mind is
> not an extension of any GSoC project.  And I can't really volunteer at
> this point.
> 
> As for Abhilash himself, I think he should concentrate on his graduate
> studies.  There's no problem with him volunteering the occasional
> half-day to Mailman, but I've seen far too many students seriously
> hindered in getting the most out of their graduate study by committing
> to consulting work.  I won't oppose him if he wants to do something,
> but I won't nominate him.

I doubt I would be able to actually do a contract work at all, even if I
wanted to, due to F-1 Visa regulations. Not for next 8 months at least.
And I have too much in my plate right now, grad school keep me busy! But
I can spend some time to mentor/help someone else as a volunteer for
whichever project we choose.

  (@Abhilash: If you decide to apply against
> my advice, it's just advice.  I'll be happy to support you to the
> extent I can in your Mailman work or your graduate study, whatever you
> want to do -- though I suspect I'd be useless in the latter role. ;-)

(off the topic: Prof, I would like to talk to you sometime for some
advice about studies :)

> 
> If somebody else is willing to mentor *my* suggestions: there are a
> couple of new proposals for DMARC mitigation out there.  I would be
> willing to evaluate them (in the remaining hours) for use in Mailman.
> Why "evaluate"?  As most of you probably know, message signing and
> signature validation are normally considered MTA functions, and
> Mailman needn't/shouldn't try to duplicate that functionality.
> However, with experimental protocols, of course they won't be in
> production MTAs for a while, so it *might* be worthwhile to offer
> support in Mailman as a spur/temporary workaround while the MTAs
> implement the new protocols.

Again off the topic: I recently came to know (from my Advisor) that my
implementation of Digital signatures in mailman can be improved by
implementing algorithms where the server does not actually decrypts the
email, just transforms it so that the recipients can decrypt that. I
don't have the details, but I am gonna explore more in that front soon.


> 
> I don't (yet) have a developer to recommend, though.
> 
> Steve
> 
> _______________________________________________
> Mailman-Developers mailing list
> Mailman-Developers at python.org
> https://mail.python.org/mailman/listinfo/mailman-developers
> Mailman FAQ: http://wiki.list.org/x/AgA3
> Searchable Archives: http://www.mail-archive.com/mailman-developers%40python.org/
> Unsubscribe: https://mail.python.org/mailman/options/mailman-developers/raj.abhilash1%40gmail.com
> 
> Security Policy: http://wiki.list.org/x/QIA9
> 

-- 
thanks,
Abhilash Raj

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20151028/534e1f79/attachment.sig>

From ned at inventati.org  Wed Oct 28 15:05:56 2015
From: ned at inventati.org (n)
Date: Wed, 28 Oct 2015 20:05:56 +0100
Subject: [Mailman-Developers] listadmin script
Message-ID: <20151028190555.GB8495@localhost>

Hello,


(I hope I'm not off topic..)


Since I maintain a couple of dozen lists, I need a way to reduce the
time taken to process them. I use listadmin[1], a command line tool to
manipulate the queues of messages held for moderator approval and
subscription requests.

It's a great tool that saved me a lot of time.


But, sadly, the upstream author does not work anymore on the listadmin
script since years.

So, new versions of mailman brings new bugs, and make it less and less
usable[2].


It's not a big script, but I haven't the technical skills to maintain
it.

I was hoping that someone on this list, with the inside knowledge of
mailman, could maintain it, or write something similar.


It would save me (and others) so much time !


Thank you for your work on mailman,


ned


[1] http://heim.ifi.uio.no/kjetilho/hacks/#listadmin
[2] https://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=listadmin
    and others unreported..

-- 

From mark at msapiro.net  Wed Oct 28 15:58:57 2015
From: mark at msapiro.net (Mark Sapiro)
Date: Wed, 28 Oct 2015 12:58:57 -0700
Subject: [Mailman-Developers] listadmin script
In-Reply-To: <20151028190555.GB8495@localhost>
References: <20151028190555.GB8495@localhost>
Message-ID: <56312901.6000208@msapiro.net>

On 10/28/2015 12:05 PM, n wrote:

> 
> It's not a big script, but I haven't the technical skills to maintain
> it.
> 
> I was hoping that someone on this list, with the inside knowledge of
> mailman, could maintain it, or write something similar.


Do you need something that runs on your workstation by screenscraping
the web UI, or can you use something which you can run in a command
shell on the host?

If the latter, look at <https://www.msapiro.net/scripts/list_requests>
(mirrored at <http://fog.ccsf.edu/~msapiro/scripts/list_requests>).

Let me know if that can work for you. If not, I can look at listadmin,
but as you can see, screen scraping interfaces are fragile and difficult
to maintain, and I'm not a strong perl coder, so no promises. Also,
realistically because of recent disasters (the death of the old
mail.python.org server) and upcoming travel plans, I probably don't have
time for this until December.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan

From ned at inventati.org  Wed Oct 28 16:27:53 2015
From: ned at inventati.org (n)
Date: Wed, 28 Oct 2015 21:27:53 +0100
Subject: [Mailman-Developers] listadmin script
In-Reply-To: <56312901.6000208@msapiro.net>
References: <20151028190555.GB8495@localhost>
 <56312901.6000208@msapiro.net>
Message-ID: <20151028202753.GC8495@localhost>

Mark Sapiro disait (le 28/10/15 ? 12:58) :
> Do you need something that runs on your workstation by screenscraping
> the web UI, or can you use something which you can run in a command
> shell on the host?
>
> If the latter, look at <https://www.msapiro.net/scripts/list_requests>
> (mirrored at <http://fog.ccsf.edu/~msapiro/scripts/list_requests>).

Thank you for your reply.

I'm moderator for lists on several mailman servers,
so it's not what I'm searching for.

Instead of screenscraping, 
I was thinking about using the email interface.

ned

> 
> Let me know if that can work for you. If not, I can look at listadmin,
> but as you can see, screen scraping interfaces are fragile and difficult
> to maintain, and I'm not a strong perl coder, so no promises. Also,
> realistically because of recent disasters (the death of the old
> mail.python.org server) and upcoming travel plans, I probably don't have
> time for this until December.
> 
> -- 
> Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
> San Francisco Bay Area, California    better use your sense - B. Dylan
> _______________________________________________
> Mailman-Developers mailing list
> Mailman-Developers at python.org
> https://mail.python.org/mailman/listinfo/mailman-developers
> Mailman FAQ: http://wiki.list.org/x/AgA3
> Searchable Archives: http://www.mail-archive.com/mailman-developers%40python.org/
> Unsubscribe: https://mail.python.org/mailman/options/mailman-developers/ned%40inventati.org
> 
> Security Policy: http://wiki.list.org/x/QIA9
> 

-- 

From mark at msapiro.net  Wed Oct 28 17:50:35 2015
From: mark at msapiro.net (Mark Sapiro)
Date: Wed, 28 Oct 2015 14:50:35 -0700
Subject: [Mailman-Developers] listadmin script
In-Reply-To: <20151028202753.GC8495@localhost>
References: <20151028190555.GB8495@localhost> <56312901.6000208@msapiro.net>
 <20151028202753.GC8495@localhost>
Message-ID: <5631432B.5010206@msapiro.net>

On 10/28/2015 01:27 PM, n wrote:
> 
> Instead of screenscraping, 
> I was thinking about using the email interface.


The email interface is designed for users, not moderators. A moderator
has no way to get info about moderator requests via email. The only
thing a moderator can do via email is approve or discard requests via
confirmation token if she already knows the tokens which she only gets
in the email notice of a held request.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan

From adam-mailman at amyl.org.uk  Thu Oct 29 07:26:40 2015
From: adam-mailman at amyl.org.uk (Adam McGreggor)
Date: Thu, 29 Oct 2015 11:26:40 +0000
Subject: [Mailman-Developers] listadmin script
In-Reply-To: <20151028190555.GB8495@localhost>
References: <20151028190555.GB8495@localhost>
Message-ID: <20151029112640.GS9573@hendricks.amyl.org.uk>

On Wed, Oct 28, 2015 at 08:05:56PM +0100, n wrote:
> Since I maintain a couple of dozen lists, I need a way to reduce the
> time taken to process them. I use listadmin[1], a command line tool to
> manipulate the queues of messages held for moderator approval and
> subscription requests.

How many of those things could be fixed by fine-tuning the list
settings, to prevent those happening, or to use filters/regexpes
better?

(I spent time looking at the actions I and moderators were taking, to
define some rules to reduce things from several a day to a couple a
month.)

[?]

> I was hoping that someone on this list, with the inside knowledge of
> mailman, could maintain it, or write something similar.

Whilst I see Mark's replied already, I wonder if it would be better
use of building tools around Mailman 3, for when 3.1 is
released/ready.

This should be easier to maintain, as it would use the various APIs,
rather than scraping.


-- 
"I am invariably puzzled by the attitude of Oxbridge to its 
 ecclesiastical past. Hearing grace recited before dinner by
 people you know don?t believe a word of it seems to be 
 considered quite normal."
   -- David Colquhoun

From ned at inventati.org  Thu Oct 29 08:44:00 2015
From: ned at inventati.org (n)
Date: Thu, 29 Oct 2015 13:44:00 +0100
Subject: [Mailman-Developers] listadmin script
In-Reply-To: <20151029112640.GS9573@hendricks.amyl.org.uk>
References: <20151028190555.GB8495@localhost>
 <20151029112640.GS9573@hendricks.amyl.org.uk>
Message-ID: <20151029124400.GA3521@localhost>

Adam McGreggor disait (le 29/10/15 ? 11:26) :
> > time taken to process them. I use listadmin[1], a command line tool to
> > manipulate the queues of messages held for moderator approval and
> > subscription requests.
> 
> How many of those things could be fixed by fine-tuning the list
> settings, to prevent those happening, or to use filters/regexpes
> better?

I will think about it, but I doubt it will do the trick ; in addition
to spam, there is a lot of irrelevant or off topic emails.

> > I was hoping that someone on this list, with the inside knowledge of
> > mailman, could maintain it, or write something similar.
> 
> Whilst I see Mark's replied already, I wonder if it would be better
> use of building tools around Mailman 3, for when 3.1 is
> released/ready.
> This should be easier to maintain, as it would use the various APIs,
> rather than scraping.

If it will be easier with the 3.1 version,
I can wait a bit longer.

ned

--