Feature request: shared address
Matus UHLAR - fantomas
the whole idea:
a "shared address" is an address being used by more people to receive
address and respond. They may respond person who originally sent e-mail
(via reply), back to the shared address (list-reply), or to both
addresses (group-reply), but the From address is always set to the shared
address.
The reason for this is to have some common address for helpdesks, where
users only should communicate with one address, and helpdesk operators
don't want users to see their address.
This is similar to mailing lists: user from outside the list sends a
message, people can discuss it internally, reply to user or reply to user
or reply to all.
However for mailing lists, people don't set "From" to the list address.
I know about no mailing list supporting this feature :(
There's a feature in mutt that supports alternate addresses (alternates)
and allows people optionally reply mail from alternate address. This is
nice, but does not have the nice feature of replying back to the alternate
address (you have to type it on your own). Also, people may want to set to
'off' for their own addresses, and it's impossible to use it then for
this purpose.
So:
What I would like to have, is a "[un]shared" command, that would define
regexp's of addresses that user is member of.
Mail to those addresses would be displayed normally, but replying those
mail with "reply", "list-reply" or "group-reply" command would reply in
standard way with address in From: set to address that matched the list.
Any comments?
Thank you.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Support bacteria - they're the only culture some people have.
Matus UHLAR - fantomas
> On Mon, 21 Mar 2005 10:45:29 +0000 (UTC), Matus UHLAR - fantomas wrote:
>> I was discussing this some time before, but it took some time to make up
>> the whole idea:
>>
>> a "shared address" is an address being used by more people to receive
>> address and respond. They may respond person who originally sent e-mail
>> (via reply), back to the shared address (list-reply), or to both
>> addresses (group-reply), but the From address is always set to the shared
>> address.
>>
>> The reason for this is to have some common address for helpdesks, where
>> users only should communicate with one address, and helpdesk operators
>> don't want users to see their address.
>
>
> A) 'helpdesk' does not want to show its real name
I was talking about e-mail address. The helpdesk address is forwarded to
more addresses and those should not be visible to outside users.
Example follows...
Initial request from user to helpdesk:
From: user <us...@example.com>
To: helpdesk <help...@example.com>
Subject: I need help.
This mail will be forwarded to helpdesk operators real addresses
<h...@example.com> and <h...@example.com>. Helpdesk addresses is shared
between them, so both should communicate using that address.
Reply from helpdesk operator 1 <h...@example.com>, that has address defined
as 'shared' will look like:
From: Mr. Helpdesk Operator 1 <help...@example.com>
To: user <us...@example.com>
Subject: We need more info (Re: I need help.)
So the user gets reply from <help...@example.com> will NOT see address
<h...@example.com>, and will further communicate with helpdesk address.
From: user<us...@example.com>
To: Mr. Helpdesk Operator <help...@example.com>
Subject: providing more info.
Now the helpdesk 2 operator issues group reply:
From: Mrs. Helpdesk Operator 2 <help...@example.com>
To: user <us...@example.com>
Cc: helpdesk <help...@example.com>
Subject: solved.
...
If any mail in between would be replied with list-reply between operators,
the user address won't appear in headers:
From: Mrs. Helpdesk Operator 2 <help...@example.com>
To: <help...@example.com>
Subject: I will take care of this one
This is acceptable.
Handling shared addresses should be very similar to mailing lists, but the
From: address should be replaced with shared address.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Microsoft dick is soft to do no harm
Matus UHLAR - fantomas
> On Tue, 22 Mar 2005 11:13:07 +0000 (UTC), Matus UHLAR - fantomas wrote:
>> Martin Trautmann <t-...@gmx.net> wrote:
>>> Sorry - I don't get the idea.
>>>
>>> A) 'helpdesk' does not want to show its real name
>>
>> I was talking about e-mail address. The helpdesk address is forwarded to
>> more addresses and those should not be visible to outside users.
>>
>> Example follows...
>>
>> Initial request from user to helpdesk:
>>
>> From: user <us...@example.com>
>> To: helpdesk <help...@example.com>
>> Subject: I need help.
>>
>> This mail will be forwarded to helpdesk operators real addresses
>><h...@example.com> and <h...@example.com>. Helpdesk addresses is shared
>> between them, so both should communicate using that address.
>>
>> Reply from helpdesk operator 1 <h...@example.com>, that has address defined
>> as 'shared' will look like:
>>
>> From: Mr. Helpdesk Operator 1 <help...@example.com>
>> To: user <us...@example.com>
>> Subject: We need more info (Re: I need help.)
>
> message-hook . "set realname="
> message-hook "~t helpdesk" "set from=help...@example.com"
> message-hook "~t helpdesk" "set realname='Mr. Helpdesk Operator'"
setting realname is outside the point. I don't want to change anyone's
realname, they can use their own realnamed.
I know I can set from address using message hooks, but that doesn't help
me using list-reply (and probably even group-reply - will I reply to
"helpdesk" if I'll set from to "helpdesk" and have reply_self unset?)
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Christian Science Programming: "Let God Debug It!".
Matus UHLAR - fantomas
> On Tue, 22 Mar 2005 11:13:07 +0000 (UTC), Matus UHLAR - fantomas wrote:
>> > Sorry - I don't get the idea.
>> >
>> > A) 'helpdesk' does not want to show its real name
>>
>> I was talking about e-mail address. The helpdesk address is forwarded to
>> more addresses and those should not be visible to outside users.
>
>
>>> -> 'From': help...@example.com
Pardon me, but:
1. this is not my example
2. this is NOT the realname but the address.
From: Adam <help...@example.com>
^^^^ ^^^^^^^^^^^^^^^^^^^^ address
||||
realname
From: Eva <help...@example.com>
^^^ ^^^^^^^^^^^^^^^^^^^^ address
||||
realname
The users can use their realnames, but should use common (shared) address
>> From: Mr. Helpdesk Operator 1 <help...@example.com>
>> To: user <us...@example.com>
>> Subject: We need more info (Re: I need help.)
>
> Bad idea to change this subject anyway.
it's useless in this point.
>> If any mail in between would be replied with list-reply between operators,
>> the user address won't appear in headers:
>>
>> From: Mrs. Helpdesk Operator 2 <help...@example.com>
>> To: <help...@example.com>
>> Subject: I will take care of this one
>
> I don't know why you actually would choose to get the same message once
> again to yourself - but here I'd recommend forwarding the question
> instead of replying (I use a patched mutt version that keeps References
> here).
You probably missed the point that <help...@example.com> is shared,
forwarded to multiple users, so mail sent to the address will come to all
of them. (Shared is the address, not the mailbox.)
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
(R)etry, (A)bort, (C)ancer
Sven Guckes
> Peter H. Coffin <hel...@ninehells.com>:
>> message-hook . "set from="
>> message-hook . "set realname="
>> message-hook "~t helpdesk" "set from=help...@example.com"
>> message-hook "~t helpdesk" "set realname='Mr. Helpdesk Operator'"
>
> setting realname is outside the point. I don't want to
> change anyone's realname, they can use their own realnamed.
a realname daemon? kewl!
> I know I can set from address using message hooks, but that doesn't help
> me using list-reply (and probably even group-reply - will I reply to
> "helpdesk" if I'll set from to "helpdesk" and have reply_self unset?)
err... who says that group-reply or list-reply is an answer?
just reply to the sender - with "Bcc: help...@example.com".
thus you'll answer the "customer" and your colleagues. done.
just add the BCC via header editing or by alias - it's up to you!
Sven
Matus UHLAR - fantomas
>> change anyone's realname, they can use their own realnamed.
Sven Guckes <use...@guckes.net> wrote:
> a realname daemon? kewl!
should be "realnames" :)
>> I know I can set from address using message hooks, but that doesn't help
>> me using list-reply (and probably even group-reply - will I reply to
>> "helpdesk" if I'll set from to "helpdesk" and have reply_self unset?)
>
> err... who says that group-reply or list-reply is an answer?
> just reply to the sender - with "Bcc: help...@example.com".
> thus you'll answer the "customer" and your colleagues. done.
> just add the BCC via header editing or by alias - it's up to you!
it is not the answer. and whole purpose of using group and lst replies is
to do some work for helpdesk operators. They can hit 'r', 'g' or 'l' and
the result will automatically be similar to using mailing lists.
They won't need to manipulate To/Cc/Bcc addresses. Isn't this the reason
why mutt has mailing-lists support?
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
On the other hand, you have different fingers.
Sven Guckes
> On Thu, 24 Mar 2005 13:11:36 +0000 (UTC), Matus UHLAR - fantomas wrote:
>> it is not the answer. and whole purpose of using group and lst
>> replies is to do some work for helpdesk operators. They can hit
>> 'r', 'g' or 'l' and the result will automatically be similar to
>> using mailing lists. They won't need to manipulate To/Cc/Bcc
>> addresses. Isn't this the reason why mutt has mailing-lists support?
>
> mutt has mailing list support. I am sorry that mailing list
> features don't work on regular email, but they just don't,
> because they don't have all the features of list mail.
exactly.
repeat: the helpdesk *alias* (aka distributor) is *not* a list. so
list-reply does not apply - and other list related stuff, neither.
Sven
Matus UHLAR - fantomas
>>> it is not the answer. and whole purpose of using group and lst
>>> replies is to do some work for helpdesk operators. They can hit
>>> 'r', 'g' or 'l' and the result will automatically be similar to
>>> using mailing lists. They won't need to manipulate To/Cc/Bcc
>>> addresses. Isn't this the reason why mutt has mailing-lists support?
> * Peter H. Coffin <hel...@ninehells.com> [2005-03-24]:
>> Yes. For mailing lists that follow mailing list conventions,
>> mutt has mailing list support. I am sorry that mailing list
>> features don't work on regular email, but they just don't,
>> because they don't have all the features of list mail.
Sven Guckes <use...@guckes.net> wrote:
> repeat: the helpdesk *alias* (aka distributor) is *not* a list. so
> list-reply does not apply - and other list related stuff, neither.
And this is exactly a feature I'd like to have in mutt: the possibility to
support the shared address feature, with list-like features (that are
already implemented in mutt, directives "lists" and "subscribe") and
setting from address to address that is configured as shared (this is
already implemented in mutt, directive "alternates" when option
"reverse_name" is set).
I probably could set up mutt to behave this way by defining the shared
address in "alternates", setting reverse_name and metoo. But I'd like to
have "metoo" unset, and would like not to show mail sent to the shared
addresses as sent to me.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Nothing is fool-proof to a talented fool.
Sven Guckes
> Sven Guckes <use...@guckes.net> wrote:
>> repeat: the helpdesk *alias* (aka distributor) is *not* a list. so
>> list-reply does not apply - and other list related stuff, neither.
>
> And this is exactly a feature I'd like to have in mutt: the possibility to
> support the shared address feature, with list-like features (that are
> already implemented in mutt, directives "lists" and "subscribe") and
> setting from address to address that is configured as shared (this is
> already implemented in mutt, directive "alternates" when option
> "reverse_name" is set).
>
> I probably could set up mutt to behave this way by defining the shared
> address in "alternates", setting reverse_name and metoo. But I'd like to
> have "metoo" unset, and would like not to show mail sent to the shared
> addresses as sent to me.
hm... let's see..
so you are *one* of the people responding to that "helpdesk" address -
but you do *not* want mutt to show that the message is addressed to you?
however, when you *do* reply to such a message then you *do* want
mutt to use that address as if it were one of of your own messages
so it keeps the address in the from: header as a sender address?
i think the bottom line is that the address you are using is *not*
a "list address" and changing the command already in place simply
do *not* work as you want them to be. instead, you want a *new*
kind of functionality - but i don't see why you cannot use any
of the existing features for that.
if you want you cow-orkes to use mutt without having to think,
well, i think you had better give a them a point-and-click tool.
this won't solve your problems - but it
allows us to filter them out more easily. ;-)
Sven
Matus UHLAR - fantomas
>> it is not the answer. and whole purpose of using group and lst replies is
>> to do some work for helpdesk operators. They can hit 'r', 'g' or 'l' and
>> the result will automatically be similar to using mailing lists.
>>
>> They won't need to manipulate To/Cc/Bcc addresses. Isn't this the reason
>> why mutt has mailing-lists support?
Peter H. Coffin <hel...@ninehells.com> wrote:
> Yes. For mailing lists that follow mailing list conventions, mutt has
> mailing list support. I am sorry that mailing list features don't work
> on regular email, but they just don't, because they don't have all the
> features of list mail.
Pardon, which conventions exactly? There are more of them, different list
processing software add different headers (if any) and it seems to me that
the only way how to detect that a mail was sent to mailing list is to find
the list name in To:/Cc: headers.
In such case, I can set up any address as list and mutt will behave as if
it was a mailing list (list-reply will work and "To: list" will be
displayed).
But maybe I'm wrong, mutt uses different logic, and that is what makes my
request somehow strange.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
WinError #99999: Out of error messages.
Matus UHLAR - fantomas
>> Sven Guckes <use...@guckes.net> wrote:
>>> repeat: the helpdesk *alias* (aka distributor) is *not* a list. so
>>> list-reply does not apply - and other list related stuff, neither.
>>
>> And this is exactly a feature I'd like to have in mutt: the possibility to
>> support the shared address feature, with list-like features (that are
>> already implemented in mutt, directives "lists" and "subscribe") and
>> setting from address to address that is configured as shared (this is
>> already implemented in mutt, directive "alternates" when option
>> "reverse_name" is set).
>>
>> I probably could set up mutt to behave this way by defining the shared
>> address in "alternates", setting reverse_name and metoo. But I'd like to
>> have "metoo" unset, and would like not to show mail sent to the shared
>> addresses as sent to me.
Sven Guckes <use...@guckes.net> wrote:
> so you are *one* of the people responding to that "helpdesk" address -
> but you do *not* want mutt to show that the message is addressed to you?
Exactly.
> however, when you *do* reply to such a message then you *do* want
> mutt to use that address as if it were one of of your own messages
> so it keeps the address in the from: header as a sender address?
Yes.
> i think the bottom line is that the address you are using is *not*
> a "list address" and changing the command already in place simply
> do *not* work as you want them to be.
Actually, there are more addressess I use this way. Some of them are
"real" mailing lists, some of them are just MTA's aliases with multiple
recipients, but they all do the same work - send mail from outside users
to all participants.
> instead, you want a *new* kind of functionality - but i don't see why
> you cannot use any of the existing features for that.
For the functionality I want, I need currently to:
1. set reverse_name. That's OK in most cases.
2. define the same address as mailing list and alternate address.
This will allow me to use list-reply and also set From address to
address mail was sent to
However alternates prevail, so mail is shown as sent to "me" when
it is currently sent to "shared address" a.k.a. list or alias.
3. set 'metoo' option for the mail to be sent to the address defined
in alternates.
But I don't like setting 'metoo' and receiving copies of e-mails.
So, i can get the functionality when I turn on one feature i really do not
want (metoo), and combine two other features (lists/subscribe and alternates)
where the final result is not the best what I can get.
Also it requires reverse_name to be set, which in some cases may be
unwanted too.
> if you want you cow-orkes to use mutt without having to think,
> well, i think you had better give a them a point-and-click tool.
I'm not the only one using mutt this way :) at least one co-worker (if I
may call him this way) uses the same functionality (and is probably
forced to solve the same problem). And I did not tell him to do so, he
somehow found out it's good idea too.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
WinError #98652: Operation completed successfully.
Matus UHLAR - fantomas
>> Pardon, which conventions exactly? There are more of them, different list
>> processing software add different headers (if any) and it seems to me that
>> the only way how to detect that a mail was sent to mailing list is to find
>> the list name in To:/Cc: headers.
Peter H. Coffin <hel...@ninehells.com> wrote:
> Yes. You need to make sure your faked list includes the headers
> described in RFC 2369, and then your help desk people put the
> appropriate list commands in their muttrc files.
I use mutt from debian distribution, versions 1.3.28 and 1.5.6
(with some debian patches, see http://packages.debian.org/mutt)
and both versions are able to treat mail as if it was sent to mailing
list, if the mailing list address appears in To: or Cc:.
Maybe this is caused by a patch, however I'm still convinced I do not need
real mailing lists for this. Is this wrong?
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
"Where do you want to go to die?" [Microsoft]
Matus UHLAR - fantomas
>>> On Fri, 1 Apr 2005 23:17:21 +0000 (UTC), Matus UHLAR - fantomas wrote:
>>>> Pardon, which conventions exactly? There are more of them, different list
>>>> processing software add different headers (if any) and it seems to me that
>>>> the only way how to detect that a mail was sent to mailing list is to find
>>>> the list name in To:/Cc: headers.
>>
>> Peter H. Coffin <hel...@ninehells.com> wrote:
>>> Yes. You need to make sure your faked list includes the headers
>>> described in RFC 2369, and then your help desk people put the
>>> appropriate list commands in their muttrc files.
>>
>> I use mutt from debian distribution, versions 1.3.28 and 1.5.6
>> (with some debian patches, see http://packages.debian.org/mutt)
>> and both versions are able to treat mail as if it was sent to mailing
>> list, if the mailing list address appears in To: or Cc:.
>
> mailing lists.
No. When using debian both stable/testing versions of mutt, mail sent to any
address that I define as 'list' address works as list address (list-reply
works and sends mail to that address).
I just checked this on one address that is surely not a list, but an
alias to me and my colleague.
My settings:
subscribe hostm...@nextra.sk
(or lists hostm...@nextra.sk, does not matter)
alternates ((postmaster|abuse|hostmaster)@(in\.)?nextra\.sk)
and the mail:
From: some.a...@some.domain
To: hostm...@nextra.sk
Now the 'r'eply causes:
From: Matus UHLAR - fantomas <hostm...@nextra.sk>
To: some.a...@some.domain
(good)
the 'g'roup-reply causes:
From: Matus UHLAR - fantomas <hostm...@nextra.sk>
To: some.a...@some.domain
(I'd like to add <hostm...@nextra.sk> to Cc: list, it's not done because
I have 'metoo' unset).
the 'l'list-reply causes:
From: Matus UHLAR - fantomas <hostm...@nextra.sk>
To: hostm...@nextra.sk
(good)
> where list is an MTA-based multi-address alias, then using Group Reply
> will send the reply to both the user, and the alias list. Where this
> differs from the mailing list functions is that there's no way to reply
> to just list@, only to user@ or everything, and if someone forgets to
> add Cc: even once, they have to put it back by hand.
This is one of things I'd like to avoid with this requested feature, people
should know that 'reply', 'group-reply' and 'list-reply' work as expected,
and it does not affect other settings.
Currently, mutt shows mail sent to hostm...@nextra.sk as sent to me, not
to the list, probably because alternates has precedence over
lists/subscribe.
>> Maybe this is caused by a patch, however I'm still convinced I do not need
>> real mailing lists for this. Is this wrong?
>
> It's not wrong if you have accessed to a mutt patched to do that. But
> the majority of mutt users don't have mutts patched that way, and won't
> be able to help you do that.
The stronger need for "shared address" support I have then ;-)
I can't require users have patched mutt, for being able to handle addresses
as mailing lists.
But what I really want here, is to ask: do you know that mutt DOES require
special headers for treating address as list address, even if the address is
defined as list address in config file?
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Windows found: (R)emove, (E)rase, (D)elete
Alain Bench
On Wednesday, April 20, 2005 at 7:06:10 PM +0000, Matus Uhlar wrote:
>| alternates ((postmaster|abuse|hostmaster)@(in\.)?nextra\.sk)
Or stricter:
| alternates ^(postmaster|abuse|hostmaster)@(in\\.)?nextra\\.sk$
> do you know that mutt DOES require special headers for treating
> address as list address, even if the address is defined as list
> address in config file?
Sorry Peter: No. The configured address has to appear in "To:" or
"Cc:" fields for Mutt to fully consider a mail as a list mail. Old
versions until 1.5.6 don't even look at special "List-*:" fields.
Right, there exist some (dozens of) patches around this. And recent
Mutts since July 20 2004 or 1.5.7 make use of "List-Post:" header to
enable <list-reply> even to non-configured lists. But if non-configured,
then no L flag, no ~l matching, no MFT, no %L format specifier, no
nothing else. Only <list-reply> works. For other list features you still
need to declare address in lists/subscribe, and address still must
appear in To/Cc.
AFAICS that's true for both stock and Debian Mutts.
Bye! Alain.
--
Mutt muttrc tip for mailing lists: set followup_to=yes and declare the list as
- subscribe ^list@ddress$ if you are subscribed and don't want courtesy copy.
- lists ^list@ddress$ if you are not subscribed or want a courtesy copy.
Matus UHLAR - fantomas
>>> address as list address, even if the address is defined as list
>>> address in config file?
> On Fri, 22 Apr 2005 15:55:20 +0200 (CEST), Alain Bench wrote:
>> Sorry Peter: No. The configured address has to appear in "To:" or
>> "Cc:" fields for Mutt to fully consider a mail as a list mail. Old
>> versions until 1.5.6 don't even look at special "List-*:" fields.
Peter H. Coffin <hel...@ninehells.com> wrote:
> Yes, I know this. But that's the version Matus was working with, so I
> didn't consider that older version behavior was particularly important.
Well, the old behaviour is important for me, because:
1. it allows recognition of mailing list that do not send required headers,
without playing with procmail
2. we still need to specify 'subscribe' for recognition of lists we're
subscribed to
and 3. it allows to do what I want, even with brute workarounds and
side-effecte :)
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Matus UHLAR - fantomas
>>> On Fri, 22 Apr 2005 15:55:20 +0200 (CEST), Alain Bench wrote:
>>>> Sorry Peter: No. The configured address has to appear in "To:" or
>>>> "Cc:" fields for Mutt to fully consider a mail as a list mail. Old
>>>> versions until 1.5.6 don't even look at special "List-*:" fields.
>>
>> Peter H. Coffin <hel...@ninehells.com> wrote:
>>> Yes, I know this. But that's the version Matus was working with, so I
>>> didn't consider that older version behavior was particularly important.
>>
>> Well, the old behaviour is important for me, because:
>>
>> 1. it allows recognition of mailing list that do not send required headers,
>> without playing with procmail
>>
>> 2. we still need to specify 'subscribe' for recognition of lists we're
>> subscribed to
>>
>> and 3. it allows to do what I want, even with brute workarounds and
>> side-effecte :)
>
> or make a real list. Here's a third: find the patch in question and
> apply it. I have no clue what that patch is, or whether it works on
> newer versions of mutt. Here's a fourth: wait until Debian releases a
> new port.
Sorry, but you probably missed original point. Let me repeat it once more:
I would like to have a feature that is currently NOT in mutt nor any other
mailing lists.
The feature is similar to the way how mutt handles mailing lists and
alternative addresses: comparing addresses in To: and Cc: with configured
reegexp's.
Debian already released new port, I have currently installed 1.5.9i, and it
behaves exactly like older versions.
I do not care about List-* addressed. I doubt the old way of list handling
(extracting addresses from To: and Cc: headers) disappears, because the new
way (extracting addresses from List- headers) is not enough, because:
1. some mailing lists do not use proper (or any) list headers
2. if someone posts to the list and to me too, I get two mails.
I would be able to list-reply only to one of them if mutt only accepted
mail with List-* headers as mail from mailing list.
That would cause 2 mails in an inbox behave differently... bad.
3. people will still define "subscribe" with list address for proper
functionality.
A fact that List-* address appears in mail to me it not enough, because
someone can bounce such mail to me as non list-subscriber.
4. mutt will stil have to extract addressess from To: and Cc: to process
'alternates'.
So, that is not about List-* headers and new version of mutt that
recognizes them. I do not need those headers and don't care about them.
I'd like to have another list ("shared"?) for which the behaviour will be
similar to lists and alternates, with specific behaviour.
Should I fill up a wishlist bugreport somewhere?
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Windows 2000: 640 MB ought to be enough for anybody
Rocco Rutte
>I do not care about List-* addressed. I doubt the old way of list handling
>(extracting addresses from To: and Cc: headers) disappears, because the new
>way (extracting addresses from List- headers) is not enough, because:
>1. some mailing lists do not use proper (or any) list headers
Yes. And the problem is? Before, there was no recognition of mailing
list specific headers like List-Post: and people complained they have to
use subscribe/lists command. Now there is recognition and people
complain because it's not the very final solution to this problem. Huh?
>2. if someone posts to the list and to me too, I get two mails.
> I would be able to list-reply only to one of them if mutt only accepted
> mail with List-* headers as mail from mailing list.
> That would cause 2 mails in an inbox behave differently... bad.
It's correct that you get 2 mails, but there's nothing you can do about.
But mutt won't behave differently. If you have a mail sent to a mailing
list and do _list_ reply, the mail goes out to the list. What else would
one expect? If you do a simple reply on both, the mail goes out to the
author.
Mutt only considers the List-Post: header if you try to do a list reply
and will only complain that no known mailing list was found if both
fails, checking List-Post: and subsribe/lists. And there is
Mail-Followup-To:...
>3. people will still define "subscribe" with list address for proper
> functionality.
...for those mailing lists not adding the proper headers.
> A fact that List-* address appears in mail to me it not enough, because
> someone can bounce such mail to me as non list-subscriber.
Yes. But in mutt there're three (semantically) different types of
replies: simple, group and list. I think you should look their meaning
up in the manual (again).
bye, Rocco
--
:wq!
Matus UHLAR - fantomas
> * Matus UHLAR - fantomas wrote:
>
>>I do not care about List-* addressed. I doubt the old way of list handling
>>(extracting addresses from To: and Cc: headers) disappears, because the new
>>way (extracting addresses from List- headers) is not enough, because:
>
>>1. some mailing lists do not use proper (or any) list headers
>
> Yes. And the problem is? Before, there was no recognition of mailing
> list specific headers like List-Post: and people complained they have to
> use subscribe/lists command. Now there is recognition and people
> complain because it's not the very final solution to this problem. Huh?
Exactly. And I think tere won't be any final solution soon.
However, the fact that some mailing lists do not use proper headers means,
that people need to specify lists/subscribe, and mutt has to check for
addresses in To:/Cc: heaers.
>>2. if someone posts to the list and to me too, I get two mails.
>> I would be able to list-reply only to one of them if mutt only accepted
>> mail with List-* headers as mail from mailing list.
>> That would cause 2 mails in an inbox behave differently... bad.
>
> It's correct that you get 2 mails, but there's nothing you can do about.
> But mutt won't behave differently.
Yes, because it detects mailing lists the way above - there won't be any
List-* addresses in one of those mails (the one that came directly to me).
Good we have lists/subscribe config optiona and mutt looks at To:/Cc: headers
>>3. people will still define "subscribe" with list address for proper
>> functionality.
>
> ...for those mailing lists not adding the proper headers.
for ANY lists. If a list message with List-* headers sits in your mailbox,
how do you know If you are subscribed to the list or not?
Anyone can bounce such message (with those list-* headers - you know what
bouce is, right?) to you. In such case, you will have list message sitting
in your mailbox, and you are NOT subscribed to the list.
And without knowing if you are subscribed, mutt will not know, what
Mail-Followup-To: header to fill.
>> A fact that List-* address appears in mail to me it not enough, because
>> someone can bounce such mail to me as non list-subscriber.
>
> Yes. But in mutt there're three (semantically) different types of
> replies: simple, group and list. I think you should look their meaning
> up in the manual (again).
Yes, but without knowing that user is subscribed, Mutt will have problems
filling up proper Mail-Followup-To:.
Blindly expecting, that you only read mails from lists you are subscribed
to, will result into sending replies to other lists, that will not have your
address in Mail-Followup-To: and you won't see how people react to your
mail.
...
I hope I was clear enough. Working with List-* headers is a step forward,
but it's not enough to replace the old functionality of handling mailing
lists that is already in mutt.
And I have to repeat, this discussion does not belong to this thread.
It's request for feature that is similar to list handling, not the same.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Depression is merely anger without enthusiasm.
Rocco Rutte
>>>3. people will still define "subscribe" with list address for proper
>>> functionality.
>> ...for those mailing lists not adding the proper headers.
>for ANY lists. If a list message with List-* headers sits in your mailbox,
>how do you know If you are subscribed to the list or not?
%Z of index_format does exactly that. The manual doesn't say 'any
%mailbox' but 'known mailbox' IIRC
The %Z flag btw, is not the only reason why mutt knows about
subsribe/list command. That's more or less a nice side effect...
bye, Rocco
--
Matus UHLAR - fantomas
>>Rocco Rutte <pd...@gmx.net> wrote:
>
>>>>3. people will still define "subscribe" with list address for proper
>>>> functionality.
>
>>> ...for those mailing lists not adding the proper headers.
>
>>for ANY lists. If a list message with List-* headers sits in your mailbox,
>>how do you know If you are subscribed to the list or not?
Rocco Rutte <pd...@gmx.net> wrote:
> %Z of index_format does exactly that. The manual doesn't say 'any
> %mailbox' but 'known mailbox' IIRC
I didn't ask, who knows that, but HOW it knows that.
And I explained that List-* headers do NOT say this, so the user has to
specify 'subscribe', or mutt will not KNOW it.
--
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Despite the cost of living, have you noticed how popular it remains?