From anirudhdahiya9 at gmail.com  Fri Apr  1 20:47:30 2016
From: anirudhdahiya9 at gmail.com (Anirudh Dahiya)
Date: Sat, 2 Apr 2016 06:17:30 +0530
Subject: [Mailman-Developers] Unable to commit changes from mailman shell to
 database
Message-ID: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>

Hi
I have been playing with mailman shell but upon creating objects like
domains and addresses, and subsequestly committing them to database by
calling transaction.commit() I am unable to see changes in the database(and
thus on Postorius)
The current approach I am taking towards committing transaction is first
creating a connection object by calling config.db.engine.connect()  . After
that, I am creating a transaction object by calling connection.begin() .
Finally after creating domain and address objects, I call
transaction.commit(),which executes succesfully, but I am unable to see
changes being reflected in the database.
Thanks
Anirudh Dahiya
(IRC spark)

From barry at list.org  Sat Apr  2 15:44:26 2016
From: barry at list.org (Barry Warsaw)
Date: Sat, 2 Apr 2016 15:44:26 -0400
Subject: [Mailman-Developers] Unable to commit changes from mailman
 shell to database
In-Reply-To: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>
References: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>
Message-ID: <20160402154426.2fda1913@subdivisions.wooz.org>

On Apr 02, 2016, at 06:17 AM, Anirudh Dahiya wrote:

>I have been playing with mailman shell but upon creating objects like
>domains and addresses, and subsequestly committing them to database by
>calling transaction.commit() I am unable to see changes in the database(and
>thus on Postorius)

>The current approach I am taking towards committing transaction is first
>creating a connection object by calling config.db.engine.connect() . After
>that, I am creating a transaction object by calling connection.begin() .
>Finally after creating domain and address objects, I call
>transaction.commit(),which executes succesfully, but I am unable to see
>changes being reflected in the database.

Which database backend are you using, SQLite or PostgreSQL?

When you're checking the database, is this from a different process?   How are
you checking the database?  Can you give us the exact set of commands you're
using both to change things and to check things?  E.g. maybe paste a console
session so we can reproduce it.

`mailman shell` exposes some globals, such as commit(), abort(), the config
object, and the mailing list `m` object if you gave the -l option.  I don't
think you normally need to do the config.db.engine.connect() call or the
connection.begin() call, and I don't know whether those are create
subtransactions or otherwise interfering with your operations.

Cheers,
-Barry

From gurkirpal204 at gmail.com  Sat Apr  2 16:00:12 2016
From: gurkirpal204 at gmail.com (Gurkirpal Singh)
Date: Sun, 3 Apr 2016 01:30:12 +0530
Subject: [Mailman-Developers] Unable to commit changes from mailman shell to
 database
In-Reply-To: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>
References: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>
Message-ID: <570024CB.4080002@gmail.com>

I tried the doing it the way you did and got the same result. After 
checking with "dir()" I found a "commit" method. I tried running it and 
it worked.

 >>> dir()
['IUserManager', '__builtins__', 'a', 'abort', 'commit', 'config', 
'getUtility', 'l', 'm', 'now', 'um', 'user_manager', 'x']
 >>> commit
<bound method SABaseDatabase.commit of 
<mailman.database.postgresql.PostgreSQLDatabase object at 0x7efbf969c9e8>>
 >>> commit()
 >>>


Maybe it was there because I'm using PostgreSQL but it worked for me.

From barry at list.org  Sat Apr  2 17:38:09 2016
From: barry at list.org (Barry Warsaw)
Date: Sat, 2 Apr 2016 17:38:09 -0400
Subject: [Mailman-Developers] Unable to commit changes from mailman
 shell to database
In-Reply-To: <570024CB.4080002@gmail.com>
References: <CAJ0CPtHa8awo332jzhLrrZmEN26AZTed03U49C+oqRskXBNEFA@mail.gmail.com>
 <570024CB.4080002@gmail.com>
Message-ID: <20160402173809.62c3c2bd@subdivisions.wooz.org>

On Apr 03, 2016, at 01:30 AM, Gurkirpal Singh wrote:

>Maybe it was there because I'm using PostgreSQL but it worked for me.

Cool, so it's working now?

I believe SQLAlchemy opens a transaction automatically, so that the bound
commit() and abort() methods operate on that transaction.  By creating a new
connection and a new transaction, I think you might have created a
subtransaction which didn't get flushed to the database.  It probably doesn't
have anything to do with the db backend specifically, but just how SA works.

Cheers,
-Barry

From mark at msapiro.net  Tue Apr  5 15:13:42 2016
From: mark at msapiro.net (Mark Sapiro)
Date: Tue, 5 Apr 2016 12:13:42 -0700
Subject: [Mailman-Developers] New Mailman 3 users list
Message-ID: <57040E66.9050705@msapiro.net>

I'm happy to announce a new list for Mailman 3 users and others
interested in Mailman 3. This list is running on a new Mailman 3
installation at lists.mailman3.org.

The list's posting address is <mailman-users at mailman3.org>.

The list's info page is
<https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/>
and its archives are at
<https://lists.mailman3.org/archives/list/mailman-users at mailman3.org/>.

The top level Postorius page is <https://lists.mailman3.org/mailman3/>
and the HyperKitty archiver is at <https://lists.mailman3.org/archives/>.

You can subscribe by email by sending a blank message to
<mailman-users-join at mailman3.org> and responding to the resulting
confirmation request. You can also subscribe by going to the info page
at
<https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/>,
logging in and subscribing.

Log in offers three options. There are Google and Facebook OAuth2 login
links on the log in page to log in with a Google or Facebook account,
and there is a 'Login by email' link to login via Persona with any
address you register and confirm with Persona.

This is a new installation of Mailman 3 so there are still kinks to work
out, and any issues can be discussed on the list or reported to the
trackers at <https://gitlab.com/mailman/postorius/issues>,
<https://gitlab.com/mailman/hyperkitty/issues> or
<https://gitlab.com/mailman/mailman/issues> as appropriate.

Enjoy!

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

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20160405/fe599f33/attachment.sig>

From aurelien at bompard.org  Wed Apr  6 12:08:04 2016
From: aurelien at bompard.org (Aurelien Bompard)
Date: Wed, 6 Apr 2016 18:08:04 +0200
Subject: [Mailman-Developers] Postorius and verified email addresses
Message-ID: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>

Hey people,

Mailman has the notion of "verified" email addresses. When a user is
created / registered in Postorius, a Mailman user can be created. It's off
by default but it seems like a logical thing to do.
This user will be created when the client visits the user profile or
preferences pages anyway, because it needs it in Mailman at that point.

However, the address that the user is created with is not verified. It's a
good thing because Django, by itself, does not verify email addresses. The
social auth providers that we use do validate them, but not Django, and
when internal auth is involved then it's only Django.

In that case, how should this address be validated? Should Postorius
consider that the login system always validates addresses and set them as
verified in Mailman? Should it ask mailman to verify the email addresses
when it encounters a user's un-verified address? This does not seem
possible in REST at the moment (unless I missed it), and should be
protected against multiple checks.

Ideas?

Aur?lien

From barry at list.org  Wed Apr  6 17:23:47 2016
From: barry at list.org (Barry Warsaw)
Date: Wed, 6 Apr 2016 17:23:47 -0400
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
Message-ID: <20160406172347.6868a76d@anarchist.wooz.org>

On Apr 06, 2016, at 06:08 PM, Aurelien Bompard wrote:

>In that case, how should this address be validated? Should Postorius
>consider that the login system always validates addresses and set them as
>verified in Mailman? Should it ask mailman to verify the email addresses
>when it encounters a user's un-verified address? This does not seem
>possible in REST at the moment (unless I missed it), and should be
>protected against multiple checks.

This is why POST on members (i.e. create a subscription) has a pre_verified
flag, which defaults to False.  The core already has a subscription workflow
to send the address a confirmation email if the subscribing address is not
already verified, and pre_verified is False.

(It will send a similar confirmation email if pre_confirmed is False and the
mailing list is set to confirm or confirm_then_moderate.)

By default, confirmation can only effectively happen by email reply, but the
intent is that you could modify the confirm.txt template[1] to include the
appropriate link back into Postorius which would effect the same verification
step as a mail-back.  This link would POST to <api>/addresses/<email>/verify
to verify the user's email address.

Thinking about it the terms you describe above, I guess there's another
workflow that isn't directly covered.  When Postorius creates the user, an
address is also created and linked to the user, but it cannot be set as the
preferred address until it's verified.  I can see where you might want to send
the verification email at some point before a subscription event, so that the
linked address gets verified and thus could be set as their preferred
address.  If that's a use case you think we need, do file a bug.  I don't
think it would be too difficult to implement.

Cheers,
-Barry

[1] Or w.r.t. GL issue #112, set a 'confirm.txt' template URL.

From simon.hanna at serve-me.info  Wed Apr  6 18:26:33 2016
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Thu, 07 Apr 2016 00:26:33 +0200
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
Message-ID: <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>

>In that case, how should this address be validated? Should Postorius
>consider that the login system always validates addresses and set them
>as
>verified in Mailman? Should it ask mailman to verify the email
>addresses
>when it encounters a user's un-verified address? This does not seem
>possible in REST at the moment (unless I missed it), and should be
>protected against multiple checks.
>
>Ideas?
I started a thread about this a couple of months back. I was asking about what authentication system to switch to, since persona is being dropped. I found a couple of solutions and have my personal favorite: django-allauth

I have a merge request pending that adds basic functionality for it in postorius. I think that it's better than python-social-auth that hyperkitty is currently using. 
Short version: it supports both external (social) and internal (django) auth systems and offers options to combine/switch between them . Allauth provides Signals that I used to verify the addresses in Mailman. 

The link to my merge request:
https://gitlab.com/mailman/postorius/merge_requests/130

The link to my previous post:

https://mail.python.org/pipermail/mailman-developers/2016-January/025338.html


-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

From barry at list.org  Wed Apr  6 19:29:07 2016
From: barry at list.org (Barry Warsaw)
Date: Wed, 6 Apr 2016 19:29:07 -0400
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
 <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
Message-ID: <20160406192907.6980b814@subdivisions.wooz.org>

On Apr 07, 2016, at 12:26 AM, Simon Hanna wrote:

>Short version: it supports both external (social) and internal (django) auth
>systems and offers options to combine/switch between them . Allauth provides
>Signals that I used to verify the addresses in Mailman.

I think we have to decide how and where addresses will be verified.  Are they
going to be via confirmations emailed by core or via Postorius?

I think the core has to support emailed confirmation messages because
Postorius is technically an optional component.  So if a site were to build
their own REST front-end, they'd at least want to allow the core to handle
email verifications without having to build that into their front-end.

That doesn't necessarily prevent Postorius from doing it, and when used with
Persona, we see how nicely that can work.  It's also of course possible that
any 3rd party front-end will have its own way of verifying email addresses.

The other thing to think about is that the core already must know how to talk
to the outgoing MTA, to provide proper reputation services, signing, etc.  I
don't know that we want to make site admins have to configure that in two
places, and we almost certainly don't want Postorius to send out emails
directly.

Cheers,
-Barry

From simon.hanna at serve-me.info  Wed Apr  6 20:46:40 2016
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Thu, 7 Apr 2016 02:46:40 +0200
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <20160406192907.6980b814@subdivisions.wooz.org>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
 <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
 <20160406192907.6980b814@subdivisions.wooz.org>
Message-ID: <5705ADF0.8080503@serve-me.info>

On 04/07/2016 01:29 AM, Barry Warsaw wrote:
> On Apr 07, 2016, at 12:26 AM, Simon Hanna wrote:
> 
>> Short version: it supports both external (social) and internal (django) auth
>> systems and offers options to combine/switch between them . Allauth provides
>> Signals that I used to verify the addresses in Mailman.
> 
> I think we have to decide how and where addresses will be verified.  Are they
> going to be via confirmations emailed by core or via Postorius?
> 
> I think the core has to support emailed confirmation messages because
> Postorius is technically an optional component.  So if a site were to build
> their own REST front-end, they'd at least want to allow the core to handle
> email verifications without having to build that into their front-end.
I don't mind the core being able to handle verfications. But I'm pretty sure everyone that offers a
web interface for managing mailman will want the ability to confirm emails using http links. I don't
see a reason why we should implement that in mailman, if it can easily be added in the front-end.
Doing this has one downside in my opinion. Storing the same addresses in several places (which isn't
bad perse, as a matter of fact microservices encourage duplicating data and synchronizing it)
would need synchronization. Ideally mailman would offer signals for various events that front-ends
can hook in to. They would probably be similar to the hyperkitty archiver plugin I guess.

Another thing I believe in is blocking access until an account is confirmed, which really shouldn't
depend on mailman. While postorius might be a project that should always have an active connection
to the core, and archive doesn't necessarily need it.
> 
> That doesn't necessarily prevent Postorius from doing it, and when used with
> Persona, we see how nicely that can work.  It's also of course possible that
> any 3rd party front-end will have its own way of verifying email addresses.
> 
> The other thing to think about is that the core already must know how to talk
> to the outgoing MTA, to provide proper reputation services, signing, etc.  I
> don't know that we want to make site admins have to configure that in two
> places, and we almost certainly don't want Postorius to send out emails
> directly.
Sorry but I have to disagree with that. Postorius _has_ to be able to send out mails.
In case any server errors occur, django tries to send out emails to administrators defined in the
settings. I strongly recommend setting this up for a production system.
The mta should take care of the rest (dkim signatures, ...). If you are referring to gpg signing and
encryption, there are django apps for that. A quick search revealed
https://github.com/stephenmcd/django-email-extras
I haven't tried it, but I don't think it would be that hard to integrate if the core supports them.

There is one more issue that needs to be discussed which is relevant to all templates: Translation.
Django has builtin methods to translate and through the browser's preferred language can choose one.
The core would require associating a language with each user in the settings.

>From a usability point of view I would like Postorius to be able to set all templates and not just
link to files in mailman. There are a couple of businesses that manage thousands of lists and I
guess they would appreciate it if list owners could do this without direct access to the mailman server.

From terri at toybox.ca  Sun Apr 10 17:06:24 2016
From: terri at toybox.ca (Terri Oda)
Date: Sun, 10 Apr 2016 14:06:24 -0700
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <20160406192907.6980b814@subdivisions.wooz.org>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
 <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
 <20160406192907.6980b814@subdivisions.wooz.org>
Message-ID: <570AC050.2010009@toybox.ca>

On 2016-04-06 4:29 PM, Barry Warsaw wrote:
> On Apr 07, 2016, at 12:26 AM, Simon Hanna wrote:
>
>> Short version: it supports both external (social) and internal (django) auth
>> systems and offers options to combine/switch between them . Allauth provides
>> Signals that I used to verify the addresses in Mailman.
>
> I think we have to decide how and where addresses will be verified.  Are they
> going to be via confirmations emailed by core or via Postorius?
>
> I think the core has to support emailed confirmation messages because
> Postorius is technically an optional component.  So if a site were to build
> their own REST front-end, they'd at least want to allow the core to handle
> email verifications without having to build that into their front-end.

I'd prefer core to do it, so that it's consistent regardless of 
front-end and there isn't duplicated work in every single web component 
for Mailman, which is one of the reasons I haven't reviewed any of the 
proposed Postorius code to do this.  (The other big one being that I 
need to do some research on auth libraries, because past experience has 
told me that I rarely agree with Simon. ;)  )

  Terri


From barry at list.org  Sun Apr 10 17:43:30 2016
From: barry at list.org (Barry Warsaw)
Date: Sun, 10 Apr 2016 17:43:30 -0400
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <5705ADF0.8080503@serve-me.info>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
 <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
 <20160406192907.6980b814@subdivisions.wooz.org>
 <5705ADF0.8080503@serve-me.info>
Message-ID: <20160410174330.2907ccee@subdivisions.wooz.org>

On Apr 07, 2016, at 02:46 AM, Simon Hanna wrote:

>I don't mind the core being able to handle verfications. But I'm pretty sure
>everyone that offers a web interface for managing mailman will want the
>ability to confirm emails using http links.

Absolutely.  This is the focus of my "templates" branch.

In MM2 we had the advantage that the web ui and the core were tightly
integrated, but because they're separated in MM3, the core can't assume
anything about the web front-end.  At a high level, it will be possible to
define a custom template for all confirmation messages, and those templates
can be filled with whatever clickable confirmation link makes sense for
whatever web ui is in place.  I imagine that for a typical Postorius-fronted
installation, we'll have some standard templates and a simple config file that
will provide a minimal effort click-through-Postorius functionality.

>I don't see a reason why we should implement that in mailman, if it can
>easily be added in the front-end.

Right, the core will *not* provide such a clickable confirmation link.

>Doing this has one downside in my opinion. Storing the same addresses in
>several places (which isn't bad perse, as a matter of fact microservices
>encourage duplicating data and synchronizing it) would need
>synchronization. Ideally mailman would offer signals for various events that
>front-ends can hook in to. They would probably be similar to the hyperkitty
>archiver plugin I guess.

Yes.  While orthogonal to the templates work, the core has a fairly rich
ability to link internal events to external notifications.  We simply need to
define the API for calling back into Postorius (or whatever web front-end
needs notification), and then we can dispatch internal events to that API.
E.g. a webhook or similar.

Alternatively, Postorius can make queries of the core's REST API at whatever
decision point makes sense.

>Sorry but I have to disagree with that. Postorius _has_ to be able to send
>out mails.  In case any server errors occur, django tries to send out emails
>to administrators defined in the settings.

That's different though.  What I mean is that any list traffic, including to
list members, owners, and moderators, should only be sent by the core.  It's
different if a Django error occurs and it wants to send an email notification
to the Django administrator (although do also think about the Mailman site
owners).

>There is one more issue that needs to be discussed which is relevant to all
>templates: Translation.  Django has builtin methods to translate and through
>the browser's preferred language can choose one.  The core would require
>associating a language with each user in the settings.

We already do that.  Members, users, and addresses have preferences, and those
preferences have a preferred language.  When sending notifications, the core
will use the preferred language of the appropriate preference.  The core will
use gettext based translations, but when using a template, we have a mechanism
for looking up a template in a particular language.  If one is available,
we'll use that, if not we always fall back to the site language (which is
usually English) or the system language (i.e. shipped templates, always
English).

Whatever translation service we end up using should support translating both
gettext and templates, but ultimately it's up to the site administrator and
system administrators of whatever front-end is used, to provide the
appropriate translated templates.

>From a usability point of view I would like Postorius to be able to set all
>templates and not just link to files in mailman. There are a couple of
>businesses that manage thousands of lists and I guess they would appreciate
>it if list owners could do this without direct access to the mailman server.

The template system I'm working on does not require file system access to the
Mailman server machine.  It will work on any URL (my intention is to use
requests for the underlying fetch machinery, with a mailman: extension for
file system access for those sites that do want to use that).  The templates
don't even need to reside in Postorius.

Cheers,
-Barry

From simon.hanna at serve-me.info  Sun Apr 10 20:39:51 2016
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Mon, 11 Apr 2016 02:39:51 +0200
Subject: [Mailman-Developers] Postorius and verified email addresses
In-Reply-To: <570AC050.2010009@toybox.ca>
References: <CAHmoxtU7_E2mh5jz9mbQH2JoJ_is98=NfUiMtC81NV5RL-GH0w@mail.gmail.com>
 <ED97E1FD-43A6-47FD-87F5-D8DDAD8484CE@serve-me.info>
 <20160406192907.6980b814@subdivisions.wooz.org> <570AC050.2010009@toybox.ca>
Message-ID: <570AF257.9030302@serve-me.info>

On 04/10/2016 11:06 PM, Terri Oda wrote:
> On 2016-04-06 4:29 PM, Barry Warsaw wrote:
>> On Apr 07, 2016, at 12:26 AM, Simon Hanna wrote:
>>
>>> Short version: it supports both external (social) and internal
>>> (django) auth
>>> systems and offers options to combine/switch between them . Allauth
>>> provides
>>> Signals that I used to verify the addresses in Mailman.
>>
>> I think we have to decide how and where addresses will be verified. 
>> Are they
>> going to be via confirmations emailed by core or via Postorius?
>>
>> I think the core has to support emailed confirmation messages because
>> Postorius is technically an optional component.  So if a site were to
>> build
>> their own REST front-end, they'd at least want to allow the core to
>> handle
>> email verifications without having to build that into their front-end.
> 
> I'd prefer core to do it, so that it's consistent regardless of
> front-end and there isn't duplicated work in every single web component
> for Mailman, which is one of the reasons I haven't reviewed any of the
> proposed Postorius code to do this.  (The other big one being that I
> need to do some research on auth libraries, because past experience has
> told me that I rarely agree with Simon. ;)  )
>From time to time I search for such projects, and I currently prefer
allauth. That doesn't mean that I know it's the best solution, it's just
the best one (in my eyes) I currently know of.

Postorius is currently confirming emails on it's own.
And Hyperkitty doesn't at all (the registration process doesn't attempt
to confirm emails).

From simon.hanna at serve-me.info  Tue Apr 12 18:27:54 2016
From: simon.hanna at serve-me.info (Simon Hanna)
Date: Wed, 13 Apr 2016 00:27:54 +0200
Subject: [Mailman-Developers] Translation
Message-ID: <570D766A.9040509@serve-me.info>

Hi,

I just completed setting up the Mailman 3 projects on zanata.org
They can be found here:

https://translate.zanata.org/project/view/mailman
https://translate.zanata.org/project/view/postorius
https://translate.zanata.org/project/view/hyperkitty

I'm hesitating to ask people to translate because zanata is really slow
and I'm afraid of scaring people away from translating mailman in the
future :D

Anyhow, I'm writing to ask for your opinions.
What do you think about zanata? Maybe you can try translating a couple
of things and report how you feel about it.

I stumbled across a few other open source solutions:
https://translatewiki.net
http://pootle.locamotion.org/
https://demo.weblate.org/

This mediagoblin issue discussed some alternatives when they switched
away from transifex. https://issues.mediagoblin.org/ticket/913


There seems to be a pootle server run by gnu itself.
https://chapters.gnu.org/pootle/


What do you think?

Simon

From mark at msapiro.net  Tue Apr 12 18:37:06 2016
From: mark at msapiro.net (Mark Sapiro)
Date: Tue, 12 Apr 2016 15:37:06 -0700
Subject: [Mailman-Developers] Translation of Mailman 3
In-Reply-To: <570D766A.9040509@serve-me.info>
References: <570D766A.9040509@serve-me.info>
Message-ID: <570D7892.3000207@msapiro.net>

Forwarding to mailman-i18n at python.org. I'm interested in the views of
that community. Note this is only about Mailman 3. No change is
contemplated to the way translations are handled for Mailman 2.1

On 04/12/2016 03:27 PM, Simon Hanna wrote to Mailman-Developers:
> Hi,
> 
> I just completed setting up the Mailman 3 projects on zanata.org
> They can be found here:
> 
> https://translate.zanata.org/project/view/mailman
> https://translate.zanata.org/project/view/postorius
> https://translate.zanata.org/project/view/hyperkitty
> 
> I'm hesitating to ask people to translate because zanata is really slow
> and I'm afraid of scaring people away from translating mailman in the
> future :D
> 
> Anyhow, I'm writing to ask for your opinions.
> What do you think about zanata? Maybe you can try translating a couple
> of things and report how you feel about it.
> 
> I stumbled across a few other open source solutions:
> https://translatewiki.net
> http://pootle.locamotion.org/
> https://demo.weblate.org/
> 
> This mediagoblin issue discussed some alternatives when they switched
> away from transifex. https://issues.mediagoblin.org/ticket/913
> 
> 
> There seems to be a pootle server run by gnu itself.
> https://chapters.gnu.org/pootle/
> 
> 
> What do you think?
> 
> Simon


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

From mark at msapiro.net  Sun Apr 17 15:59:09 2016
From: mark at msapiro.net (Mark Sapiro)
Date: Sun, 17 Apr 2016 12:59:09 -0700
Subject: [Mailman-Developers] Mailman 2.1.22 release
Message-ID: <5713EB0D.10109@msapiro.net>

I am pleased to announce the release of Mailman 2.1.22.

Python 2.4 is the minimum supported, but Python 2.7 is strongly recommended.

There are no new features in this release.  There are a few i18n updates
and some bug fixes. See the attached README for details.

Mailman is free software for managing email mailing lists and
e-newsletters. Mailman is used for all the python.org and
SourceForge.net mailing lists, as well as at hundreds of other sites.

For more information, please see our web site at one of:

http://www.list.org
http://www.gnu.org/software/mailman
http://mailman.sourceforge.net/
http://mirror.list.org/

Mailman 2.1.22 can be downloaded from

https://launchpad.net/mailman/2.1/
http://ftp.gnu.org/gnu/mailman/
https://sourceforge.net/projects/mailman/

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan
-------------- next part --------------
2.1.22 (17-Apr-2016)

  i18n

    - Fixed a typo in the German options.html template.  (LP: #1562408)

    - An error in the Brazilian Portugese translation of Quarterly has been
      fixed thanks to Kleber A. Benatti.

    - The Brazilian Portugese translation has been updated by Emerson Ribeiro
      de Mello.

  Bug fixes and other patches

    - All addresses in data/virtual-mailman are now properly appended with
      VIRTUAL_MAILMAN_LOCAL_DOMAIN and duplicates are not generated if the
      site list is in a virtual domain.  (LP: #1570630)

    - DMARC mitigations will now find the From: domain to the right of the
      rightmost '@' rather than the leftmost '@'.  (LP: #1568445)

    - DMARC mitigations for a sub-domain of an organizational domain will now
      use the organizational domain's sp= policy if any.  (LP: #1568398)

    - Modified NewsRunner.py to ensure that messages gated to Usenet have a
      non-blank Subject: header and when munging the Message-ID to add the
      original to References: to help with threading.  (LP: #557955)

    - Fixed the pipermail archiver to do a better job of figuring the date of
      a post when its Date: header is missing, unparseable or has an obviously
      out of range date.  This should only affect bin/arch as ArchRunner has
      code to fix dates at least if ARCHIVER_CLOBBER_DATE_POLICY has not been
      set to 0 in mm_cfg.py.  If posts have been added in the past to a list's
      archive using bin/arch and an imported mbox, running bin/arch again could
      result is some of those posts being archived with a different date.
      (LP: #1555798)

    - Fixed an issue with CommandRunner shunting a malformed message with a
      null byte in the body.  (LP: #1553888)

    - Don't collapse multipart with a single sub-part inside multipart/signed
      parts.  (LP: #1551075)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://mail.python.org/pipermail/mailman-developers/attachments/20160417/3ede3c90/attachment.sig>

From barry at list.org  Wed Apr 20 09:58:15 2016
From: barry at list.org (Barry Warsaw)
Date: Wed, 20 Apr 2016 09:58:15 -0400
Subject: [Mailman-Developers] [Mailman-i18n]  Translation of Mailman 3
In-Reply-To: <570D7892.3000207@msapiro.net>
References: <570D766A.9040509@serve-me.info>
	<570D7892.3000207@msapiro.net>
Message-ID: <20160420095815.0f175625@subdivisions.wooz.org>

>On 04/12/2016 03:27 PM, Simon Hanna wrote to Mailman-Developers:
>> Hi,
>> 
>> I just completed setting up the Mailman 3 projects on zanata.org
>> They can be found here:
>> 
>> https://translate.zanata.org/project/view/mailman
>> https://translate.zanata.org/project/view/postorius
>> https://translate.zanata.org/project/view/hyperkitty
>> 
>> I'm hesitating to ask people to translate because zanata is really slow
>> and I'm afraid of scaring people away from translating mailman in the
>> future :D
>> 
>> Anyhow, I'm writing to ask for your opinions.
>> What do you think about zanata? Maybe you can try translating a couple
>> of things and report how you feel about it.
>> 
>> I stumbled across a few other open source solutions:
>> https://translatewiki.net
>> http://pootle.locamotion.org/
>> https://demo.weblate.org/
>> 
>> This mediagoblin issue discussed some alternatives when they switched
>> away from transifex. https://issues.mediagoblin.org/ticket/913
>> 
>> There seems to be a pootle server run by gnu itself.
>> https://chapters.gnu.org/pootle/
>> 
>> What do you think?

From a project standpoint, we have a few considerations.

We're a GNU project, so we have a hard requirement that any service we choose
must run on free software.

From my own perspective, I'd like to separate out what we as project leaders
and developers care about from what translators care about.  Meaning, we
should be able to extract the pot file and templates and upload them whenever
it makes sense from a project standpoint.  Translators can do their job on
their own schedule and don't need to be tied directly to project milestones.
To the extent that we have to include any translation artifacts in the
software we release, we should be able to get a snapshot when needed and just
include it in our release.

Then all I care about is that translators are happy enough with the system
that they'll actively translate!  It does us no good if the service is too
painful for translators to use.  As a monolinguist, I don't really have any
insight into that. ;)

We need a Mailman 3 translation champion, someone who understand the technical
and more importantly, social issues involved, and can spend time and energy on
helping bring a good story to fruition.  I'm happy to give wide latitude to
the champion to help shape a solution that works for us.  Maybe that's you
Simon?

Cheers,
-Barry

From stephen at xemacs.org  Wed Apr 20 22:47:44 2016
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: Thu, 21 Apr 2016 11:47:44 +0900
Subject: [Mailman-Developers] [Mailman-i18n]  Translation of Mailman 3
In-Reply-To: <20160420095815.0f175625@subdivisions.wooz.org>
References: <570D766A.9040509@serve-me.info> <570D7892.3000207@msapiro.net>
 <20160420095815.0f175625@subdivisions.wooz.org>
Message-ID: <22296.16208.742523.449355@turnbull.sk.tsukuba.ac.jp>

-i18n removed, I'm not subscribed.

Barry Warsaw writes:

 > We need a Mailman 3 translation champion, someone who understand
 > the technical and more importantly, social issues involved, and can
 > spend time and energy on helping bring a good story to fruition.
 > I'm happy to give wide latitude to the champion to help shape a
 > solution that works for us.  Maybe that's you Simon?

I can help somewhat, I know the technology, I have a "friend in the
business" (a lawyer buddy who's heavily invested in legal translation
software), I've been involved with the Mailman and Debian translation
communitiess in the past.  But right now I'm "busy as Barry", and for
the near future GSoC is going to sop up most of my Mailman time.

I'll subscribe to -i18n shortly.

Steve