warn about a BCC field

warn about a BCC field

[Uwe Brauer]

Hi

I suppose I receive a message which contains a bcc field. Now I don't
want by mistake hit, gnus-summary-followup-with-original because then
the original recipient will get into the CC field. Is there a way to be
warned about that?

Like: the original message contained a BCC are you sure you want to use
followup instead reply to?

Regards

Uwe Brauer 

Re: warn about a BCC field

[Lars Ingebrigtsen]

Uwe Brauer <oub@mat.ucm.es> writes:

> I suppose I receive a message which contains a bcc field. Now I don't
> want by mistake hit, gnus-summary-followup-with-original because then
> the original recipient will get into the CC field. Is there a way to be
> warned about that?

Hm.  That's not supposed to happen at all -- that function is not
supposed to put the Bcc recipients into the Cc header.

Does Gnus really do that?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: warn about a BCC field

[Dan Christensen]

On Apr 11, 2018, Lars Ingebrigtsen <larsi@gnus.org> wrote:

> Uwe Brauer <oub@mat.ucm.es> writes:
>
>> I suppose I receive a message which contains a bcc field. Now I don't
>> want by mistake hit, gnus-summary-followup-with-original because then
>> the original recipient will get into the CC field. Is there a way to be
>> warned about that?
>
> Hm.  That's not supposed to happen at all -- that function is not
> supposed to put the Bcc recipients into the Cc header.
>
> Does Gnus really do that?

First, you shouldn't receive messages with Bcc headers, since they are
stripped before sending.

Second, even if you have a message with a Bcc header (say, in your Sent
folder), Gnus does not seem to copy those recipients into the wide
reply / followup.  (I just tested.)

Dan

Re: warn about a BCC field

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> I suppose I receive a message which contains a bcc field. Now I don't
   >> want by mistake hit, gnus-summary-followup-with-original because then
   >> the original recipient will get into the CC field. Is there a way to be
   >> warned about that?

   > Hm.  That's not supposed to happen at all -- that function is not
   > supposed to put the Bcc recipients into the Cc header.

   > Does Gnus really do that?

Hm I start to suspect that this has to do with gmail. My university
switched form its internal server to gmail. I will do some testing again.

Re: warn about a BCC field

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> I suppose I receive a message which contains a bcc field. Now I don't
   >> want by mistake hit, gnus-summary-followup-with-original because then
   >> the original recipient will get into the CC field. Is there a way to be
   >> warned about that?

   > Hm.  That's not supposed to happen at all -- that function is not
   > supposed to put the Bcc recipients into the Cc header.

   > Does Gnus really do that?

I investigated further and indeed there is a problem.
As I said my university switched to gmail. Now someone sent me, using
his Apple Mail app, a mail with the following headers

From: User1 <user@ucm.es>
Subject: Blabla
To: User2 <user2@ucm.es>
Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
Bcc: oub@mat.ucm.es


So you see I am only on the BCC field which is displayed. Now I hit F
(as I do automatically and which in this case was a big mistake because
it results in

To: User1 <user@ucm.es>
Subject: Blabla
CC: User2 <user2@ucm.es>
Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
Bcc: oub@mat.ucm.es

And resulted in a somehow uncomfortable situation.

Meanwhile I did the following

(defun my-check-bcc ()
  "Try to avoid painful situations when using followup, and there is a Bcc field."
  (interactive)
  (cond ((and gnus-article-reply
              (message-with-reply-buffer
                (re-search-forward
                 "^Bcc:\\|BCC:\\|bcc:"
                 nil t)))
         (condition-case nil
             (progn
               (message-remove-header "cc")
               (message-goto-body)
               (message "There is a BCC field, we have deleted it!!!"))
           (error
            (progn
                (message "Followup??")))))))

(add-hook 'gnus-message-setup-hook 'my-check-bcc)


Any thoughts or comments?

Uwe Brauer 

Re: warn about a BCC field

[Lars Ingebrigtsen]

Uwe Brauer <oub@mat.ucm.es> writes:

> I investigated further and indeed there is a problem.
> As I said my university switched to gmail. Now someone sent me, using
> his Apple Mail app, a mail with the following headers
>
> From: User1 <user@ucm.es>
> Subject: Blabla
> To: User2 <user2@ucm.es>
> Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
> Bcc: oub@mat.ucm.es
>
> So you see I am only on the BCC field which is displayed. Now I hit F

Hm...  `F'?  What's that bound to?  Is this in a mail group?

> (as I do automatically and which in this case was a big mistake because
> it results in

> To: User1 <user@ucm.es>
> Subject: Blabla
> CC: User2 <user2@ucm.es>
> Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
> Bcc: oub@mat.ucm.es

Hm?  That's exactly the same as the first message?  Including Date: and
all? 

If I `F' on a mail with a Bcc in it, Message seems to ignore the Bcc as
it should...

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: warn about a BCC field

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> I investigated further and indeed there is a problem.
   >> As I said my university switched to gmail. Now someone sent me, using
   >> his Apple Mail app, a mail with the following headers
   >> 
   >> From: User1 <user@ucm.es>
   >> Subject: Blabla
   >> To: User2 <user2@ucm.es>
   >> Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
   >> Bcc: oub@mat.ucm.es
   >> 
   >> So you see I am only on the BCC field which is displayed. Now I hit F

   > Hm...  `F'?  What's that bound to?  Is this in a mail group?

gnus-summary-followup-with-original

I thought this is the standard binding


   >> (as I do automatically and which in this case was a big mistake because
   >> it results in

   >> To: User1 <user@ucm.es>
   >> Subject: Blabla
   >> CC: User2 <user2@ucm.es>
   >> Date: Wed, 22 Feb 2017 08:39:23 +0100 (1 year, 7 weeks, 1 day ago)
   >> Bcc: oub@mat.ucm.es

   > Hm?  That's exactly the same as the first message?  Including Date: and
   > all? 


   > If I `F' on a mail with a Bcc in it, Message seems to ignore the Bcc as
   > it should...

Ok this is a misunderstanding. It is not about the BCC is about the CC
field when the original message contained a BCC field.

I try to say it again.

I am the recipient in the BCC field.

The message was not sent to me, (I was only on the BCC)
I did not pay attention and wanted to answer the sender, so  when hitting
gnus-summary-followup-with-original
gnus put up the correct
TO field, which was ok
but also the CC field
extracting that information from the CC field.

Gnus behavior is correct!

If I had used gnus-summary-reply-with-original (as I should) everything
would have been fine, but I am so used to followup, (because people
often complained in the that they don't get a message when they are only
on the CC field and I use gnus-summary-reply-with-original, that is why
I am so used to gnus-summary-followup-with-original, but in the
situation I described that can for the careless user lead to
embarrassing solution, and that is why I wrote that small hack, which
might come in handy for people in similar situation.

Did I explain my problem now better?

Re: warn about a BCC field

[Lars Ingebrigtsen]

Uwe Brauer <oub@mat.ucm.es> writes:

> Did I explain my problem now better?

Oh, I see.  You were Bcc'd on a message, but then did a wide reply that
revealed to the other people in the conversation that you were party to
the conversation, and you didn't want to do that.

Well, first of all: Mail agents are supposed to strip the Bcc header
before sending the mail, so you shouldn't have gotten that Bcc header.
So that's a bug in the software that sent you the mail.

But the question is: If you're not listed in To/Cc in a mail, and you do
a wide reply, should Gnus warn you about what you're doing?  I don't
think so: It's a very common situation, especially with mailing lists.

So I don't really think there's anything here to be done on the Gnus
side, but your my-check-bcc sounds nice for your specific use case,
though.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: warn about a BCC field

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> Did I explain my problem now better?

   > Oh, I see.  You were Bcc'd on a message, but then did a wide reply that
   > revealed to the other people in the conversation that you were party to
   > the conversation, and you didn't want to do that.

   > Well, first of all: Mail agents are supposed to strip the Bcc header
   > before sending the mail, so you shouldn't have gotten that Bcc header.
   > So that's a bug in the software that sent you the mail.

Well well. I just sent myself with gnus the following message

From: Uwe Brauer <oub@mat.ucm.es>
Subject: this
To: Uwe Brauer <oub@mat.ucm.es> 
Cc: Uwe Brauer <burrurr@gmail.com> 
Date: Sat, 14 Apr 2018 00:14:28 +0200 (48 seconds ago)
Bcc: oub.oub.oub@gmail.com


As you can see that is a BCC field.

Now when I open in my oub.oub.oubGmail.com account the message and
indeed the BCC field is gone. I did the same with thunderbird and the
same behavior. Odd no?

   > But the question is: If you're not listed in To/Cc in a mail, and
   > you do a wide reply, should Gnus warn you about what you're doing?
   > I don't think so: It's a very common situation, especially with
   > mailing lists.

I see your point concerning mailing list, but what's about the other use
case I just mentioned? What would you do in such a situation?

However I just realized that my code is not failsafe. 

As it seems I can't trust that such a mail (where I am not in the TO/CC)
would include a bcc field, and then my hack would fail.

Couldn't we set a variable (gnus-for-the-careless-repliers) 
but then it is not clear to me what to search for? Any ideas how to deal
with that?

It seems that most people are either not very often in such a situation
or are just more careful then I am.

   > So I don't really think there's anything here to be done on the Gnus
   > side, but your my-check-bcc sounds nice for your specific use case,
   > though.

Re: warn about a BCC field

[Lars Ingebrigtsen]

Uwe Brauer <oub@mat.ucm.es> writes:

> Well well. I just sent myself with gnus the following message
>
> From: Uwe Brauer <oub@mat.ucm.es>
> Subject: this
> To: Uwe Brauer <oub@mat.ucm.es> 
> Cc: Uwe Brauer <burrurr@gmail.com> 
> Date: Sat, 14 Apr 2018 00:14:28 +0200 (48 seconds ago)
> Bcc: oub.oub.oub@gmail.com
>
> As you can see that is a BCC field.
>
> Now when I open in my oub.oub.oubGmail.com account the message and
> indeed the BCC field is gone. I did the same with thunderbird and the
> same behavior. Odd no?

No, mail user agents are supposed to strip the Bcc header before
sending, and that's what Gnus is doing, so everything works as it
should.

The whole point of the Bcc header is that isn't not supposed to be sent
out.

> I see your point concerning mailing list, but what's about the other use
> case I just mentioned? What would you do in such a situation?
>
> However I just realized that my code is not failsafe. 
>
> As it seems I can't trust that such a mail (where I am not in the TO/CC)
> would include a bcc field, and then my hack would fail.
>
> Couldn't we set a variable (gnus-for-the-careless-repliers) 
> but then it is not clear to me what to search for? Any ideas how to deal
> with that?

Just look for your email address in the headers.  If it's not there,
it's dangerous for you to do a wide reply on it.

> It seems that most people are either not very often in such a situation
> or are just more careful then I am.

I think so.  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: warn about a BCC field

[Uwe Brauer]

   > Uwe Brauer <oub@mat.ucm.es> writes:

   > No, mail user agents are supposed to strip the Bcc header before
   > sending, and that's what Gnus is doing, so everything works as it
   > should.

   > The whole point of the Bcc header is that isn't not supposed to be sent
   > out.

Again a misunderstanding the following 

,----
| From: Uwe Brauer <oub@mat.ucm.es>
| Subject: test bdd
| To: Uwe Brauer <oub@mat.ucm.es> 
| Date: Sat, 14 Apr 2018 00:10:57 +0200 (8 hours, 59 minutes, 22 seconds ago)
| Bcc: oub.oub.oub@gmail.com
`----

Is a header in my oub@mat.ucm.es

Mailbox, after receiving it. It is *not* a message I want to sent.


This is the header of a mail in my oub.oub.oub@gmail.com mailbox


,----
| From: Uwe Brauer <oub@mat.ucm.es>
| Subject: test bdd
| To: Uwe Brauer <oub@mat.ucm.es> 
| Date: Sat, 14 Apr 2018 00:10:57 +0200 (8 hours, 59 minutes, 43 seconds ago)
`----

When I do the opposite sent from my oub.oub.oub@gmail.com account to
oub@mat.ucm.es the bcc is no displayed. And this behavior is independent
from the MUA (gnus/thunderbird)

Re: warn about a BCC field

[Adam Sjøgren]

Uwe writes:

>> No, mail user agents are supposed to strip the Bcc header before
>> sending, and that's what Gnus is doing, so everything works as it
>> should.
>
>> The whole point of the Bcc header is that isn't not supposed to be sent
>> out.
>
> Again a misunderstanding the following 
>
> ,----
> | From: Uwe Brauer <oub@mat.ucm.es>
> | Subject: test bdd
> | To: Uwe Brauer <oub@mat.ucm.es> 
> | Date: Sat, 14 Apr 2018 00:10:57 +0200 (8 hours, 59 minutes, 22 seconds ago)
> | Bcc: oub.oub.oub@gmail.com
> `----
>
> Is a header in my oub@mat.ucm.es
>
> Mailbox, after receiving it. It is *not* a message I want to sent.

You keep saying this, and it keeps being baffling.

Are you SURE that the email you are looking at is one you _received_ and
not a copy that was _stored_ in your local archive (equivalent of "Sent"
folder)?

When I send an email with Gnus, the local copy stored in
nnml+archive:mail-2018 has the Bcc header¹, while the emails sent out to
the recipients in To, Cc and Bcc do not.

That's the way it is supposed to work. If you see something different,
something is wrong.

You should probably do your experiment with different From and To. That
is less likely to confuse, as it will clearly separate received emails
from locally stored copies.


  Best regards,

    Adam


¹ So I can see that I used Bcc - while nobody else csn (the recipients
  in the Bcc might infer that they were Bcc'ed, but really can't be
  sure).

-- 
 "I'm a dessertarian."                                        Adam Sjøgren
                                                         asjo@koldfront.dk

Re: warn about a BCC field

[Uwe Brauer]

   > Uwe writes:

   > You keep saying this, and it keeps being baffling.



   > Are you SURE that the email you are looking at is one you _received_ and
   > not a copy that was _stored_ in your local archive (equivalent of "Sent"
   > folder)?


Well now I tried out the webmail interface of this particular gmail
account and the fields

from:	Uwe Brauer <oub@mat.ucm.es>
to:	Uwe Brauer <oub.oub.oub@gmail.com>
cc:	Uwe Brauer <oub@mat.ucm.es>
bcc:	Uwe Brauer <burrurr@gmail.com>
date:	Sat, Apr 14, 2018 at 12:07 PM
subject:	test bcc via webmail
mailed-by:	mat.ucm.es


Then after sending in my INBOX there is a message
which contains the BCC field.
And I received mails from colleagues which contained BCC fields

Ok what about the following? I sent you offline a mail with a BCC field
and you tell me whether you can see it?

That would settle it?

I did that sending to my other gmail account and in that INOBOX I could
see the BCC field.

Would you do that experiment?

Uwe

Re: warn about a BCC field

[Adam Sjøgren]

Uwe writes:

> Well now I tried out the webmail interface of this particular gmail
> account and the fields
>
> from:	Uwe Brauer <oub@mat.ucm.es>
> to:	Uwe Brauer <oub.oub.oub@gmail.com>
> cc:	Uwe Brauer <oub@mat.ucm.es>
> bcc:	Uwe Brauer <burrurr@gmail.com>
> date:	Sat, Apr 14, 2018 at 12:07 PM
> subject:	test bcc via webmail
> mailed-by:	mat.ucm.es
>
> Then after sending in my INBOX there is a message
> which contains the BCC field.

I just tried sending an email from a Gmail account, like this:

  from: asjoegren@gmail.com
  to:   asjo@koldfront.dk
  bcc:  asjo@asjo.org

In "Sent Mail" I see the bcc-header on Gmail. The email received on
koldfront.dk does NOT have the Bcc header. The email received on
asjo.org does NOT have the Bcc header.

This is how I expect things to behave.

The only difference I see is that I have not Cc'ed myself on the email
(you have).

Maybe Gmail realizes that From and Cc is the same, and doesn't strip the
Bcc field?

Let me try again, this time I do this:

  from: asjoegren@gmail.com
  to:   asjo@koldfront.dk
  cc:   asjoegren@gmail.com
  bcc:  asjo@asjo.org

(Like you). And lo and behold, in Inbox in Gmail I _do_ see the Bcc
field.

I can see two explanations:

 a) Either Gmail realizes that From and Cc are identical, and therefore
    it doesn't strip the Bcc (as to not confuse you), OR

 b) The search based approach to email means that you have ONE local
    email showing up both in Sent Mail and in Inbox, which would of
    course retain the Bcc.

And the answer is: b).

If I go to Inbox and choose "Show original" I get the exact same URL as
if I go to "Sent Mail" and choose "Show original".

So what you are seeing is an artefact of how Gmail stores your email.

> And I received mails from colleagues which contained BCC fields

This is very weird, I don't think that should happen, even when using
Gmail.

> Ok what about the following? I sent you offline a mail with a BCC field
> and you tell me whether you can see it?

Shoot!


  Best regards,

    Adam

-- 
 "I'm a dessertarian."                                        Adam Sjøgren
                                                         asjo@koldfront.dk

Re: warn about a BCC field

[Uwe Brauer]

   > Uwe Brauer <oub@mat.ucm.es> writes:

   > Just look for your email address in the headers.  If it's not there,
   > it's dangerous for you to do a wide reply on it.

I thought about this point.

As it turns out (thanks an experiment I did with Adam), you were right,
BCC is not, usually ,included in the message, only in some wired gmail
setting.

Now, that makes my approach useless in the general case. Moreover my
approach was based on a function which I put in the setup-hook so it
checked for a BCC field once the reply buffer was generated.

But if I want to deal with the general case I should check before
creating the reply buffer.

How do I do that? I mean which function can the article buffer and
extract the information for the TO and CC fields?

Thanks

Uwe Brauer  

Re: warn about a BCC field

[Lars Ingebrigtsen]

Uwe Brauer <oub@mat.ucm.es> writes:

> How do I do that? I mean which function can the article buffer and
> extract the information for the TO and CC fields?

(message-fetch-field "From"), for instance.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

[Partial solution: how to warn about a possible BCC field] (was: warn about a BCC field)

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> How do I do that? I mean which function can the article buffer and
   >> extract the information for the TO and CC fields?

   > (message-fetch-field "From"), for instance.

Thanks. I thought how to implement this.

    -  first the question is warn or change. Change would mean to kill
       all CC fields, which I consider as to much, in case the message
       comes from a mailing list.

    -  so a warning seems appropriate, (although Adam is right in
       pointing out that maybe you would automatically say yes.) 

    -  The issue is then that the function you propose works well in an
       article buffer, but I can't get it to work in a summary buffer.

    -  I could either modify the gnus-article-followup-with-original,
       but I have the feeling Lars does not like that idea.

    -  write a wrapper and bind it.

    -  use advice (which I usually try to avoid) but this time it was
       convenient.



(defun check-my-to-field ()
  "This function will be run before you execute followup. And it warns you if your email address is not 
in the To field. That warning is a safty message if you had been only on the BCC field!!"
  (interactive)
  (unless (string-match "oub@mat.ucm.es" (message-fetch-field "From"))
    (if (y-or-n-p "Warning: Followup? but you are not on the TO field, proceed?")
      (message "I hope you are sure.")
      (error
       (progn
         (message "Aborted, you are not in the To: field!!!"))))))


(defadvice gnus-article-followup-with-original (before mychecktofield activate) ;Version:1.131
  "This function will be run before you execute followup. It warns you if your email address is not in the To field. That
warning is a safty message if you had been only on the BCC field!!"
  (check-my-to-field))


So I came up with this. However what I am supposed to do in the summary
buffer?

Any thoughts??

Uwe Brauer 

Re: warn about a BCC field

[Uwe Brauer]

>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

   > Uwe Brauer <oub@mat.ucm.es> writes:
   >> Did I explain my problem now better?

   > But the question is: If you're not listed in To/Cc in a mail, and
   > you do a wide reply, should Gnus warn you about what you're doing?
   > I don't think so: It's a very common situation, especially with
   > mailing lists.

On a second thought if  in such a situation I will use

gnus-summary-reply-with-original

I will be asked

Really reply by mail to article author? (y or n) y

Couldn't we have (with a setting of a variable say) the same for followup?

Really followup by mail to article authors (including  the CC) ? (y or n) y

Re: warn about a BCC field

[Adam Sjøgren]

Uwe writes:

>>>> "Lars" == Lars Ingebrigtsen <larsi@gnus.org> writes:

>> But the question is: If you're not listed in To/Cc in a mail, and
>> you do a wide reply, should Gnus warn you about what you're doing?
>> I don't think so: It's a very common situation, especially with
>> mailing lists.
>
> On a second thought if  in such a situation I will use
>
> gnus-summary-reply-with-original
>
> I will be asked
>
> Really reply by mail to article author? (y or n) y

That's something you've configured yourself, right?

I know I have an advice adding exactly that.

> Couldn't we have (with a setting of a variable say) the same for followup?
>
> Really followup by mail to article authors (including  the CC) ? (y or n) y

So regardless of whether you reply or follow up, you want to be asked if
that's really what you want to do?

I think your fingers will quickly learn to press 'y' automatically.


  Best regards,

    Adam

-- 
 "Ja, det lyder jo som radioen vender på hovedet!"            Adam Sjøgren
                                                         asjo@koldfront.dk

Re: warn about a BCC field

[Uwe Brauer]

   > Uwe writes:


   > That's something you've configured yourself, right?

   > I know I have an advice adding exactly that.


   > So regardless of whether you reply or follow up, you want to be asked if
   > that's really what you want to do?

   > I think your fingers will quickly learn to press 'y' automatically.

Sigh, might be true (for example I have an advice asking me to encrypt or
decrypt. Since most recipients don't use encryption I say n, even
sometimes to people I do want to encrypt).......

   >   Best regards,

   >     Adam