[mlmmj] Resent-From + subonlypost

[mlmmj] Resent-From + subonlypost

[Sabahattin Gucukoglu]

Hi all,

One of my lists is a news update list for friends.  Sometimes I "Redirect" (RFC 5322 Resent-* fields) directly to my list.  The From: header thus is the original From:.  You might guess the problem: I'll redirect, then because subonlypost is set, a bounce will confuse the original sender.

Please can mlmmj support the Resent-From for the checks requiring a check against the original sender prior to the From: field?

As an aside, I'm using TMDA; many TMDA users will be using From as I do, to hold the original address.  So mlmmj works well for action required by or posts received from a TMDA user, such as this one.

BTW: would it be possible to implement nomail subscription automatically using a response to an initial post?  This would allow non-members to easily CC a list when reply-all'ing on a thread containing that list.  I saw discussion from 2004 about that, only nothing seems to have come of it.

Cheers,
Sabahattin

Re: [mlmmj] Resent-From + subonlypost

[Morten Shearman Kirkegaard]

Hi Sabahattin,

On Thu, 2010-04-08 at 00:07 +0100, Sabahattin Gucukoglu wrote:
> Sometimes I "Redirect" (RFC 5322 Resent-* fields) directly to my list. 
> The From: header thus is the original From:.  You might guess the
> problem: I'll redirect, then because subonlypost is set, a bounce will
> confuse the original sender.
> 
> Please can mlmmj support the Resent-From for the checks requiring a
> check against the original sender prior to the From: field?

No.

Quoting RFC 5322:

Resent fields are strictly informational.  They MUST NOT be used in the
normal processing of replies or other such automatic actions on
messages.

> BTW: would it be possible to implement nomail subscription
> automatically using a response to an initial post?

Could you rephrase that?


Best regards,
Morten

-- 
Morten Shearman Kirkegaard <moki@fabletech.com>
CTO, FableTech
http://fabletech.com/

Re: [mlmmj] Resent-From + subonlypost

[Sabahattin Gucukoglu]

[-- Attachment #1: Type: text/plain, Size: 1539 bytes --]

On 10 Apr 2010, at 20:34, Morten Shearman Kirkegaard wrote:
> On Thu, 2010-04-08 at 00:07 +0100, Sabahattin Gucukoglu wrote:
>> Sometimes I "Redirect" (RFC 5322 Resent-* fields) directly to my list. 
>> The From: header thus is the original From:.  You might guess the
>> problem: I'll redirect, then because subonlypost is set, a bounce will
>> confuse the original sender.
>> 
>> Please can mlmmj support the Resent-From for the checks requiring a
>> check against the original sender prior to the From: field?
> 
> No.
> 
> Quoting RFC 5322:
> 
> Resent fields are strictly informational.  They MUST NOT be used in the
> normal processing of replies or other such automatic actions on
> messages.

That's right.  So how does this prevent using Resent-From before From?  (Remember, you don't ever use those fields for any kind of non-delivery notifications, only the return address, envelope sender.)

> BTW: would it be possible to implement nomail subscription
>> automatically using a response to an initial post?
> 
> Could you rephrase that?

The idea is, sometimes someone wants to reply-all, including to a list, when they get into a thread someone quotes a list on.  Given the chance they may not want to subscribe but would, one time, reply to a message asking them if they want to post (like Dan's qsecretary).  When they reply, they get subscribed nomail, and the original post goes through.  So it's a bit like TMDA in front of a list, only much faster and more useful. :-)

Cheers,
Sabahattin


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2655 bytes --]

Re: [mlmmj] Resent-From + subonlypost

[Morten Shearman Kirkegaard]

On Mon, 2010-04-12 at 00:32 +0100, Sabahattin Gucukoglu wrote:
> On 10 Apr 2010, at 20:34, Morten Shearman Kirkegaard wrote:
> > They MUST NOT be used in
...
> > other such automatic actions on messages.
> 
> That's right.  So how does this prevent using Resent-From before From?

That's how.

> sometimes someone wants to reply-all, including to a list,
> when they get into a thread someone quotes a list on.  Given the
> chance they may not want to subscribe but would, one time, reply to a
> message asking them if they want to post (like Dan's qsecretary). 
> When they reply, they get subscribed nomail, and the original post
> goes through.

Holding the original mail and releasing it when the sender subscribes
would require some changes, but it's doable. Automatically sending a
subscription request shouldn't be too hard. Patches are welcome.


Morten

-- 
Morten Shearman Kirkegaard <moki@fabletech.com>
CTO, FableTech
http://fabletech.com/

Re: [mlmmj] Resent-From + subonlypost

[Sabahattin Gucukoglu]

[-- Attachment #1: Type: text/plain, Size: 1253 bytes --]

On 13 Apr 2010, at 06:31, Morten Shearman Kirkegaard wrote:
On Mon, 2010-04-12 at 00:32 +0100, Sabahattin Gucukoglu wrote:
>> On 10 Apr 2010, at 20:34, Morten Shearman Kirkegaard wrote:
>>> They MUST NOT be used in
> ....
>>> other such automatic actions on messages.
>> 
>> That's right.  So how does this prevent using Resent-From before From?
> 
> That's how.

Okay, in case there was any doubt or misunderstanding on either side I brought it to YAM.  Here's the thread start:
http://www.ietf.org/mail-archive/web/yam/current/msg00410.html

You can choose either behaviour, but I'll look at patching mine.

> sometimes someone wants to reply-all, including to a list,
>> when they get into a thread someone quotes a list on.  Given the
>> chance they may not want to subscribe but would, one time, reply to a
>> message asking them if they want to post (like Dan's qsecretary). 
>> When they reply, they get subscribed nomail, and the original post
>> goes through.
> 
> Holding the original mail and releasing it when the sender subscribes
> would require some changes, but it's doable. Automatically sending a
> subscription request shouldn't be too hard. Patches are welcome.

Looking into it.

Cheers,
Sabahattin


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2655 bytes --]