548885 - bcc list not shown in Gmail IMAP sent mail (whenever gmail is used as smtp server or a mail from another smtp is copied into gmail -any folder- the BCCs disappear. Only BCCs in notes sent from gmail web interface are preserved and visible in TB)

Status: RESOLVED DUPLICATE of bug 55073

(MailNews Core :: Composition, defect)

Description

[pclough]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2) Gecko/20100115 Firefox/3.6 (.NET CLR 3.5.30729)
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100216 Thunderbird/3.0.2

When a mail is sent to a bcc list, there is no record to whom it was sent.  This is true whether looking at normal header, expanded header or source.

Reproducible: Always

Steps to Reproduce:
1.send any mail to just a bcc address or list
2.look in sent mail to see who recipients were
3.
Actual Results:  
no "To: or CC: or BCC: recipients listed

Expected Results:  
see the bcc list

Comment 1

[Nathan Tuggy (:tuggyne)]

Unable to reproduce on Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.1.8) Gecko/20100216 Lightning/1.0b1 Thunderbird/3.0.2 ID:20100216161403. All email addresses show up at most one click and some scrolling away.

Does this happen in Safe Mode? (See https://support.mozillamessaging.com/en-US/kb/Safe+Mode for details on how to enter Thunderbird Safe Mode.) Do any error messages show up in the Error Console?

Also, what account types (IMAP, POP, etc) are you using?

Comment 2

[pclough]

Attachment: result of test 1

Comment 3

[pclough]

Attachment: source note for test 1

Comment 4

[pclough]

Attachment: result of test 2

Comment 5

[pclough]

Attachment: source note for test 2

Comment 6

[pclough]

Attachment: result of test 3 (sent from gmail)

Comment 7

[pclough]

Attachment: source note for test 3

Comment 8

[pclough]

I am using IMAP with gmail.

There were no errors in error console as a result of the tests I just ran in safe mode (or from previous tests, for that matter).

Comment 9

[Nathan Tuggy (:tuggyne)]

This is probably specific to Thunderbird's integration with GMail; as I don't have a GMail account, will need additional repro.

[=> MailNews Core | Composition like bug 61621; this isn't necessarily quite the right place, though.]

Comment 10

[pclough]

Nathan, is there anything I can do to help?  Peter

Comment 11

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #8)
> I am using IMAP with gmail.

(Q1) Did you send the mail via Gmail's SMTP server?
(Q2) What is your setting for "folder to save copy of sent mail"?
     (Account Settings, Copies & Folders)

(Q3) What headers do you see at each step of next?

(0) Account Settings, Copies & Folders, enable copy of sent mail, and select
   "Sent" of "Local Folders". (If no Sent, create manually.)
    Select "Drafts" of "Local Folders" as draft folder for ease of test.
(1) Compose a text mail, select Gmail's mail address at From:,
    set your other mail address in To: (never use same address as From:),
    and set another your mail address in BCC:. 
(2) Save as Draft, check mail source of mail in "Drafts" of "Local Folders"
(3) "Send Later", check mail source of mail in "Outbox" of "Local Folders"
(4) "Send Unsent Messages" => mail data is passed to Gmail's SMTP server
(5) Check mail source of mail in "[Gmail]/Sent Mail"
    This is mail data saved by Gmail's SMTP/Gmail, is presented by Gmail IMAP.
    See also "[Gmail]/All Mail". Same mail data is seen. (Gmail's design)
(6) Check mail in "Sent" of "Local Folders".
    This is mail data saved by Tb after passing data to Gmail's SMTP.
(7) At Inbox of your other account(s), check mail source of arrived mail.

(8) Create an Gmail IMAP folder via UI of Tb (say "Test").
(9) Copy mail of step (6) (mail in "Sent" of "Local Folders") to Test.
    Check mail source of mail in "Test".
(Q4) Which data? Data at step (5)? Data at step (6)? 

(10) Check mail source of mail(s) in "[Gmail]/All Mail".
(Q5) Does Gmail/Gmail IMAP generate two versions of mails?
     (one: no BCC: header, another: with BCC: header)

Comment 12

[pclough]

Note - just tested from a machine with TB2 installed and had same problem.
Q1 - notes are sent via gmail smtp server 
Q2 - "sent" folder on account "[e-mail]"
Q3.2 - bcc visible
Q3.3 - bcc visible
Q3.5 - bcc not visible in Gmail "sent" or "all mail"
Q3.6 - bcc visible
Q3.7 - To visible in both, bcc in neither
Q3.9 - bcc not visible in copy, still visible in Local folders "sent"
Q4 - covered above
Q5 - no
Note 2 - when creating and sending from gmail on web, all works properly, whether in draft or sent file; the bcc in such a case also shows up in TB in sent file. When I do a global search in gmail and find the bcc email, it shows the bccs.  Same in All Mail in TB.

Comment 13

[WADA:World Anti-bad-Duping Agency]

Confirming per test results by bug opener.

Simplest/easiest workaround:
  Use SMTP server other than Gmail's SMTP always, or if BCC: is required.
By other SMTP choice for identity of [e-mail], mail data without BCC: is not written in [Gmail]/Sent Mail (nor [Gmail]/All Mail) by Gmail's SMTP & Gmail. So, "sent mail copy with BCC: by Tb" is written in Sent(and [Gmail]/All Mail) by next your setting.
> Q2 - "sent" folder on account "[e-mail]"

Note:
If you send mail which has BCC: to yourself([e-mail] in To:/CC:/BCC:) using other SMTP server, arrived mail(in Inbox and in [Gmail]/All Mail) may have BCC: header(same data as sent mail copy in "Sent"). I don't know Gmail considers "mail with Received: headers" as dup'ed mail or not.
Please never open bug for "BCC: is exposed by Tb!" in such case.

Comment 14

[pclough]

>"Use SMTP server other than Gmail's SMTP always, or if BCC: is required." 
- this is not practical especially when I am travelling.  Actually the easier workaround is to use the web gmail to send bcc notes, as awkward as that may be.

>By other SMTP choice for identity of [e-mail], mail data without BCC:
>is not written in [Gmail]/Sent Mail (nor [Gmail]/All Mail) by Gmail's SMTP &
>Gmail. So, "sent mail copy with BCC: by Tb" is written in Sent(and [Gmail]/All
>Mail) by next your setting.
If I understand what this is saying, sending using some other SMTP is fine, except that if I try to place a copy of the sent mail into the gmail sent folder the bcc's disappear.  Another reason why the workaround above works best.

>Note:
>If you send mail which has BCC: to yourself([e-mail] in To:/CC:/BCC:)
>using other SMTP server, arrived mail(in Inbox and in [Gmail]/All Mail) may
>have BCC: header(same data as sent mail copy in "Sent"). I don't know Gmail
>considers "mail with Received: headers" as dup'ed mail or not.
>Please never open bug for "BCC: is exposed by Tb!" in such case.

Is that meant to be funny?  I assure you I will open bugs that are bugs ... like the one I opened here.  We still need a fix which works with gmail smtp as a server.

Comment 15

[nathan]

I agree that this is a bug that should be taken seriously. I've found it because I just set up a thunderbird/gmail combination for someone who isn't that computer savvy, and they have just contacted me because they need to know who they sent some bcc'd emails to. 

Pretty embarrassing to have to explain that the stable version of Thunderbird has lost that information for them!

Comment 16

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #14)
> If I understand what this is saying, sending using some other SMTP is fine,
> except that if I try to place a copy of the sent mail into the gmail sent
> folder the bcc's disappear.

[e-mail], did you really checked it without sending mail to your Gmail's mail address?
  - via SMTP server other than Gmail's one
  - Copy sent mail to "Sent" of Gmail IMAP account 
  - No your Gmail's mail address in To:, CC:, BCC:

Following are copies of sent mail in "Sent" folder of Gmail IMAP.
(mail-2, SMTP=smtp.opd.dti.ne.jp)
> BCC: [e-mail]
> Message-ID: <[e-mail]>
> Date: Tue, 02 Mar 2010 13:17:47 +0900
> From: "M. Wada" <[e-mail]>
> User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100216 Thunderbird/3.0.2
> MIME-Version: 1.0
> To: [e-mail]
> Subject: BCC test 02
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> Content-Transfer-Encoding: 7bit
> 
> BCC test 02
(mail-3, SMTP=smtp.opd.dti.ne.jp)
> BCC: [e-mail]
> Message-ID: <[e-mail]>
> Date: Tue, 02 Mar 2010 13:22:33 +0900
> From: Yatter One <[e-mail]>
> User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100216 Thunderbird/3.0.2
> MIME-Version: 1.0
> To: [e-mail]
> Subject: BCC test 03
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> Content-Transfer-Encoding: 7bit
>
> BCC test 03

If Gmail's address is set in To:/CC:/BCC:, Gmail replaced the copied sent mail data with BCC: by newly arrived version(Received: headers exist, no BCC:).
  - Same Message-ID, Subject, but longer one, then replases by new one?
    Similar to check of newer version of draft mail?

(mail-1, SMTP=smtp.opd.dti.ne.jp
> Delivered-To: [e-mail]
> Received: by 10.142.180.9 with SMTP id c9cs583919wff; Mon, 1 Mar 2010 20:15:37 -0800 (PST)
> Received: by 10.114.23.2 with SMTP id 2mr3141570waw.132.1267503336826; Mon, 01 Mar 2010 20:15:36 -0800 (PST)
> Return-Path: <[e-mail]>
> Received: from smtp09.dti.ne.jp (smtp09.dti.ne.jp [202.216.231.184]) (snip)
> Received-SPF: pass (google.com: domain of [e-mail] (snip)
> Received: from [192.168.0.10] (snip)
> Message-ID: <[e-mail]>
> Date: Tue, 02 Mar 2010 13:15:33 +0900
> From: "M. Wada" <[e-mail]>
> User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100216 Thunderbird/3.0.2
> MIME-Version: 1.0
> To: Yatter One <[e-mail]>
> CC: [e-mail]
> Subject: BCC test 01
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> Content-Transfer-Encoding: 7bit
>
> BCC test 01

Comment 17

[pclough]

>[e-mail], did you really checked it without sending mail to your
>Gmail's mail address?
>  - via SMTP server other than Gmail's one
>  - Copy sent mail to "Sent" of Gmail IMAP account 
>  - No your Gmail's mail address in To:, CC:, BCC:

Yes, I REALLY did.  See attachments for a further test.

Shouldn't TB be passing the entire email including bcc's to gmail and letting it sort out who sees which addresses?  Gmail is certainly capable of not showing bcc's when the email is sent from its web interface.

And, in any case, a user should be able to bcc himself at the from address without losing the bcc's of the sent message.  I have done this for years with other mail programs usually to see what attachments will look like inline when viewed by the recipients.

Peter

Comment 18

[pclough]

Attachment: test of bcc with no copy to gmail - orig

Comment 19

[pclough]

Attachment: test of bcc with no copy to gmail - sent folder

Comment 20

[pclough]

WADA,

I see you also specified sending via another non-gmail smtp.  That is not an acceptable solution or workaround.  If you need it for a test, I'll do it, but I'm not sure what it would prove.

My whole purpose in using gmail and IMAP is to use the same SMTP everywhere (home and travel) and to synchronize mail on desktop and laptop.  We've already determined (and posted the evidence within this thread) that when I send via gmail smtp with a copy to Local Folders "sent" folder, the bcc's are there.  If I send a note from gmail web interface with bcc's, those bcc's appear in gmail "sent" folder in TB.  We have also established that if I copy the note from the Local Folders "sent" folder to the gmail "sent" folder within (and using) TB, the bcc's disappear. I don't see that that behavior fits your theory of what is the cause of the problem.

Comment 21

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #19)
> test of bcc with no copy to gmail - sent folder

"Sent Mail"(==[Gmail]/Sent Mail) instead of "sent"(or "Sent"), isn't it?
Which SMTP server did you really use to send the mail?
Did you change SMTP server choice of identity for your Gmail mail address?
Note:
If main identity(first identity, listed at top of "manage identity"), SMTP choice change at "manage identity"/Edit is not effective. If main identity, SMTP choice change at main Account Setting panel is required. It's known bug.

(In reply to comment #17)
> >[e-mail], did you really checked it without sending mail to your
> >Gmail's mail address?
> >  - via SMTP server other than Gmail's one
> >  - Copy sent mail to "Sent" of Gmail IMAP account 
> >  - No your Gmail's mail address in To:, CC:, BCC:
> 
> Yes, I REALLY did.  See attachments for a further test.

"REALLY did" which? All?
Please misunderstand my test result. My test result is to present you that BCC: header is REALLY saved in Gmail's mail folder, unless same mail is sent to Gmail.

To avoid confusion, to understand what Gmail does, to see "BCC: header by Tb" in Gmail's folder by yourself, try next.
 (1) Add SMTP server definition(other than Gmail's one), label it "SMTP-X".
     Same server/username as already defined one can be defined.
 (2) Add identity for your Gmail's mail address.
     Same mail address as already defined one can be defined.
     Set name="My Gmail SMTP-X", and choose "SMTP-X" for SMTP server.
     Folder to save copy of sent mail = Gmail IMAP account's root level "Sent"
 (3) Compose a mail, choose "My Gmail SMTP-X" at From:.
     Set To:, CC:, BCC: (no Gmail's mail address in it), and send.
 (4) Change folder view to "All Folders" to avoid confusion.
 (5) Check root level "Sent" and "[Gmail]/Sent Mail".
     No BCC: in mail in "Sent"?
     Does the sent mail data exist in "[Gmail]/Sent Mail"? 

If "always BCC: to myself only", loss of BCC: header in sent mail copy is not a problem, because no my mail address in To:/CC:==sent as BCC:.
If "always no my mail-addr in BCC:", send via other SMTP server is a workaround.
However, if "other people and myself in BCC:", BCC: header in sent mail copy for housekeeping is lost even when other SMTP is used, if sent mail copy is saved in Gmail's folder. In this case, as you already know, "save copy of sent mail in local mail folder" is a workaround, if you are mobile user(use same PC always at different locations). I prefer this to "always send mail via Web mail".

Comment 22

[WADA:World Anti-bad-Duping Agency]

Quick summary of Gmail's behaviour.
(1) There are three versions of mail data.
  (#1) Mail data passed to Gmail's SMTP. No BCC:, no Receved: header etc.
       Gmail Label of "[Gmail]/Sent Mail", "[Gmail]/All Mail" are added.
       Mail data passed by Gmail IMAP to Tb == #1.
  (#2) Copy of sent mail uploaded(append) by Tb with BCC: (root level "Sent").
       Gmail Label of "Sent", "[Gmail]/All Mail" are added.
       Mail data passed by Gmail IMAP to Tb == #1.
       Gmail looks to consider #2 as DUP of #1.
  (#3) Arrived mail by To:/CC:/BCC: of the mail, with Received: etc., no BCC: 
       Gmail Label of "Inbox", "[Gmail]/All Mail" are added.
       Mail data passed by Gmail IMAP to Tb == #3.
       Gmail looks to replace #1 with #3.
  To keep #2 with BCC: by Tb in Gmail's folder, #1 and #3 shouldn't happen.
(2) If mail is sent via Gmail's Web interface, Gmail looks to ignore
    own mail-addr in To:/CC:/BCC:. So it looks that "replace of #2 with BCC:
    generated by Gmail with arrived #3" wont' happen.
    This #2 with BCC: may be replaced if sent mail data is uploaded to Gmail
    via Gmail IMAP.

Comment 23

[pclough]

WADA,

We seem to be talking at cross purposes.  You have a theory that two copies of the same mail are being sent to gmail, one with and one without BCCs and that the one without BCCs is replacing the one with BCCs because gmail is treating the one with BCCs as a duplicate.

You propose a lot of tests using another non-gmail smtp server.

If those tests will help in providing a TB fix to the problem, I'll do them.  BUT, using an alternate smtp server as a "solution" to this problem is NOT acceptable.  

I just ran a very simple test:
1. created a note in an Account with a different smtp server (not gmail)
2. specified in account settings to place a copy in Local Folders "sent"
3. sent
4. copy appeared in proper account inbox, with no indication that it was bcc (ok)
5. copy in Local Folders "sent" had bcc shown clearly
6. copied the note from Local Folders "sent" to "_Test BCC" a new subfolder in gmail.  The BCC disappeared during that copy.

Note: no other copies had been sent or copied to gmail.

My observation:  something bizarre is happening when a note with BCC is being copied to gmail, regardless of the smtp origination.

Peter

Comment 24

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #23)
> You propose a lot of tests using another non-gmail smtp server.
> If those tests will help in providing a TB fix to the problem, I'll do them. 
> BUT, using an alternate smtp server as a "solution" to this problem is NOT
> acceptable.  

???
I never propose workaround of "different SMTP server from Gmail's one" to you.
I merely requested you to corectly understand what Gmail does do with checking several cases, (a-1) mail sent via Gmail's SMTP, (a-2) mail sent via other than Gmail's SMTP, (b-1) mail data is written to Gmail's folder by only Tb with BCC: header as copy of sent mail, (b-2) mail data is also written by Gmail as mail data passed to Gmail's SMTP, (b-3) mail data is also written by Gmail as mail data arrived to Gmail.

Do you correctly understand test results of your comment #12?

Comment 25

[pclough]

Wada,

We are either struggling with language or one of us is not understanding the other.  You say in comment 24:

>I never propose workaround of "different SMTP server from Gmail's one" to you.

But in comment 13 you said:

> Simplest/easiest workaround:
>  Use SMTP server other than Gmail's SMTP always, or if BCC: is required.

I'm not sure how to rationalize those two statements.

Here is how I see the situation:

1. Notes sent from non-gmail smtp and filed in non-gmail folders properly show the BCCs in sent folder.

2. Notes sent from non-gmail-smtp and filed in gmail folders do not show the BCCs in sent mail folder.

3. Notes sent from gmail's web interface and filed in gmail folders properly show the BCCs in sent mail folder.

4. Notes sent from non-gmail smtp, filed in non-gmail folders lose the BCC data when copied to any gmail folder.

5. From this I conclude that multiple notes going to gmail has nothing to do with the problem unless TB under the covers is generating multiple notes which are overlaying one another, which is still a TB bug requiring a fix, it seems to me.

I have posted test results to support the first four of those assertions and logic supports assertion 5.  Now, what do you propose we do from here to get this fixed?

Peter

Comment 26

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #25)
> We are either struggling with language or one of us is not understanding the
> other.  You say in comment 24:
> >I never propose workaround of "different SMTP server from Gmail's one" to you.
> But in comment 13 you said:
> > Simplest/easiest workaround:
> >  Use SMTP server other than Gmail's SMTP always, or if BCC: is required.
> I'm not sure how to rationalize those two statements.

Sorry for unclear explanation.
 Comment #13: as general workaround of "loss of copy with BCC:" with Gmail
 Comment #24: as workaround in your case.
              - Use of Gmail's SMTP server sounds mandatory.
              - BCC: sounds often used to send mail copy to Gmail 
 
> which is still a TB bug requiring a fix, it seems to me.

Your observations is same as my observations.
"loss of sent mail copy with BCC: uploaded by Tb" is never Tb's bug. Tb simply does do next 1/2.
  1. Tb passes mail data stream(no BCC: header) to SMTP server.
     If Gmail's SMTP is used, passed mail data is saved with Gmail Label
     of [Gmail]/Sent Mail.
  2. Tb uploads "copy of sent mail with BCC:" to Gmail via Gmail IMAP.
     Gmail IMAP returns OK to append.
     If used SMTP server at step 1 is not Gmail's SMTP,
     "copy of sent mail with BCC:" is really held in Gmail's IMAP folder.
  3. Mail arrives at Gmail, as a result of step 1.
     Gmail replaces mail data by data of arrived mail(no BCC:, Received: etc.)

If Gmail's SMTP is used at step 1, Gmail doesn't keep mail data at step 2(with BCC: by Tb) even though Gmail IMAP returns OK to append. And, Gmail IMAP returns mail data of step 1(no BCC:) as mail data at step 2.
Trick in "mail send with BCC: via Web interface" is probably one of next:
 - step 2 happens first and step 1 happens, then data at step 2 wins,
   and step 3 doesn't happen.
 - Both step 1 and step 3 doesn't happen.
 - Gmail keeps newer data(with BCC:) at step 2, and step 3 doesn't happen.

> Now, what do you propose we do from here to get this fixed?

All of above is current implementation of Gmail, Gmail's SMTP, and Gmail IMAP. Tb(and any ordinal mail client) can do nothing for next phenomenon. Next phenomenon is never produced by Tb's fault.
  loss of sent mail copy with BCC: uploaded by mailer, with Gmail's SMTP/IMAP
Your should request improvement to Google instead of Mozilla foundation/Mozilla Messaging company/Mozilla community.

Possible quirks by Tb to bypass this bug's problem with Gmail:
  Change Message-ID of sent mail copy from original, in addition to adding BCC:
    Original       : Message-ID: <abc-def>
    Sent mail copy : Message-ID: <abc-def-sent-mail-copy-by-tb-with-bcc>
                     BCC: ...
  If Gmail doesn't consider as DUP if different Message-ID:,
  the sent mail copy is kept as different mail in "[Gmail]/All Mail".

Comment 27

[pclough]

Wada,

You are hung up on your theory.  But there are two important observations of mine which you do not seem to have taken into account and which cannot be explained by your theory:

1. use NON-gmail SMTP to send and ask to save a copy in imap gmail sent-mail folder;  the result is that the BCCs are lost.  Note that gmail has seen just one copy.

2. use NON-gmail SMTP to send and save a copy in Local Folders.  BCCs are visible; BUT, if you then, as a separate operation, copy that email to IMAP gmail folder of any name, the BCCs disappear.

Your theory does not explain either of those phenomena.  Nor does it explain why a note sent using gmail web interface handles the BCCs properly and is copyable and moveable from folder to folder within TB without losing anything.

Net: 

- your theory does not explain all behaviors we are seeing
- you have therefore NOT demonstrated that this is a gmail bug; in fact, given the rest of the observations you have not considered, it seems to me that the evidence points at TB and its interface to gmail
- in the next few days I will try to install another IMAP mail program to see if it can handle BCCs through gmail any better.

I see that others have experienced something similar in Outlook - see

http://help.wugnet.com/office/bcc-folder-IMAP-account-Outlook-2007-ftopict1071569.html

However, in that case, copying the note to the sent folder solved the problem.
What I suspect is that gmail has some non-standard rules about the BCCs and deletes them if the mail program doesn't follow them.

Peter

Comment 28

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #27)
> 1. use NON-gmail SMTP to send and ask to save a copy in imap gmail sent-mail folder;

mail-2/mail-3(bcc: really exists) in my Comment #16 is mail data in root-level Sent of Gmail IMAP, generated by steps you are talking.

> the result is that the BCCs are lost.

Which mail was seen?
(a) sent mail copy(no Received: headers), lack of BCC: only 
(b) arrived mail(some Received: headers, no BCC: header)
If (a), did you really send via non-Gmail SMTP? Check Received: headers which arrived at other account by To:/CC:/BCC:.
If (b), did you specified Gmail's mail-addr in To:/CC:/BCC:?

> Note that gmail has seen just one copy.

It's right. Gmail keeps single copy of mail data only. Duplicated mail data is removed by Gmail. Gmails considered next three versions "same mail".
  (1) mail saved by Gmail's SMTP, (2) mail appended by Tb(with BCC:),
  (3) mail arrived at Gmail.
  (mail data in "[Gmail]/All Mail" can be said the "single mail copy".) 

> 2. use NON-gmail SMTP to send and save a copy in Local Folders.  BCCs are
> visible; BUT, if you then, as a separate operation, copy that email to IMAP
> gmail folder of any name, the BCCs disappear.

Unable to reproduce it.
You already had arrived mail or other (single) copy for same mail in Gmail's folder(check [Gmail]/All Mail, [Gmail]/Sent Mail, Inbox, ...), didn't you?

Gmail keeps single copy of mail data only. If Gmail thinks same mail, Gmail won't keep different copy of mail data, even though Gmail returns OK to append. It's similar to next.
  1. Tb copies data with BCC: from local to IMAP(append) => OK, UID=(N) 
  2. Other cient deletes mail of UID=(N), and issues expunge
  3. Tb accesses folder, and knows that the mail of UID=(N) is deleted/expunged 
It's Gmail's design/implementation. 

> Nor does it explain why a note sent using gmail web interface handles the BCCs properly

First, ask to Google for why, instead of me.
Second, check always wheter Gmail already has a copy of same mail or not in any testing, please.

Check next, please. (Show "Order Received" column. UID is shown)
  1. Send a mail to your POP3 account via Gmail's Web Interface, with BCC:
  2. Sent mail copy is held in "Sent Mail"(==[Gmail]/Sent Mail via Gmail IMAP).
     The mail is also held in [Gmail]/All Mail. Say UID=(X).
  3. Copy arrived mail in Inbox of POP3 acount(no BCC:, many Received:) to
     a Gmail IMAP folder. Mail is also held in [Gmail]/All Mail. Say UID=(Y).
  4. Two different mails of different UID appear at [Gmail]/All Mail?
     (one : with BCC:, no Received. another : without BCC:, many Received:)
     If single mail only, which mail? UID of the mail is changed?

> and is copyable and moveable from folder to folder within TB without losing anything.

It's not surprizing. The mail is same as mail-2/mail-3 in my Comment #16.

Comment 29

[pclough]

>(In reply to comment #27)
>>1. use NON-gmail SMTP to send and ask to save a copy in imap gmail sent-mail >>folder;

>mail-2/mail-3(bcc: really exists) in my Comment #16 is mail data in root-level
>Sent of Gmail IMAP, generated by steps you are talking.

I fail to understand why a note generated by TB and sent with another (non-gmail) smtp cannot be copied by TB into the sent mail folder of gmail without losing BCCs.


>>the result is that the BCCs are lost.

>Which mail was seen?
>(a) sent mail copy(no Received: headers), lack of BCC: only 
>(b) arrived mail(some Received: headers, no BCC: header)
>If (a), did you really send via non-Gmail SMTP? Check Received: headers which
>arrived at other account by To:/CC:/BCC:.
>If (b), did you specified Gmail's mail-addr in To:/CC:/BCC:?

(a) only - as I said;  YES, it was sent via another smtp! - as I said.


>>Note that gmail has seen just one copy.

>It's right. Gmail keeps single copy of mail data only. Duplicated mail data is
>removed by Gmail. Gmails considered next three versions "same mail".
>(1) mail saved by Gmail's SMTP, (2) mail appended by Tb(with BCC:),
>(3) mail arrived at Gmail.
>(mail data in "[Gmail]/All Mail" can be said the "single mail copy".) 

What I meant above is that gmail has not been exposed to more than one copy of this email, the "sent" version saved into its "sent mail" folder.  There was no note sent or received by gmail.  Please do not question this again.


>Unable to reproduce it.
>You already had arrived mail or other (single) copy for same mail in Gmail's
>folder(check [Gmail]/All Mail, [Gmail]/Sent Mail, Inbox, ...), didn't you?

NO NO NO.  It was sent with another smtp server.  Please do not ask me this again.


>Gmail keeps single copy of mail data only. If Gmail thinks same mail, Gmail
>won't keep different copy of mail data, even though Gmail returns OK to append.
>It's similar to next.
>1. Tb copies data with BCC: from local to IMAP(append) => OK, UID=(N) 
>2. Other cient deletes mail of UID=(N), and issues expunge
>3. Tb accesses folder, and knows that the mail of UID=(N) is deleted/expunged 
>It's Gmail's design/implementation. 

This is irrelevant to this point since gmail HAS ONLY BEEN GIVEN ONE COPY OF THIS MAIL.

>>Nor does it explain why a note sent using gmail web interface handles the >>BCCs properly

>First, ask to Google for why, instead of me.

I'm asking Mozilla since gmail manages to create a note where BCCs work with IMAP and are visible and I don't understand why TB cannot / does not do the same, since it's clearly possible. So we have an existence theorem that it CAN be done and the question becomes WHY Tb isn't doing it.  (A valid answer might be an undocumented gmail requirement for a certain format.  Regardless, I'd like to see TB format its BCCs and headers in such a way that they can be processed by gmail IMAP just as gmail-generated notes with BCCs are.)

>Second, check always wheter Gmail already has a copy of same mail or not in any
>testing, please.

I have done this and specified it in several notes describing my testing.  You seem unwilling to accept my word for it.  Doesn't make it any less true.

>Check next, please. (Show "Order Received" column. UID is shown)
>1. Send a mail to your POP3 account via Gmail's Web Interface, with BCC:
>2. Sent mail copy is held in "Sent Mail"(==[Gmail]/Sent Mail via Gmail IMAP).
>The mail is also held in [Gmail]/All Mail. Say UID=(X).
>3. Copy arrived mail in Inbox of POP3 acount(no BCC:, many Received:) to
>a Gmail IMAP folder. Mail is also held in [Gmail]/All Mail. Say UID=(Y).
>4. Two different mails of different UID appear at [Gmail]/All Mail?
>(one : with BCC:, no Received. another : without BCC:, many Received:)
>If single mail only, which mail? UID of the mail is changed?

Here is the source of the note sent as you requested as it appears in Pop client as received, in gmail sent mail and as it appears in gmail ALL Mail:

-------------------------------------------

as it appears in BCC addresse's inbox:


From - Wed Mar 03 07:49:19 2010
X-Account-Key: account7
X-UIDL: 131-1226436196
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
X-Mozilla-Keys:                                                                                 
Return-path: <[e-mail]>
Received: from mta15.srv.hcvlny.cv.net
 (mta15.srv.hcvlny.cv.net [167.206.5.109]) by mstr3.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail]; Wed, 03 Mar 2010 07:45:34 -0500 (EST)
Received: from mail-fx0-f216.google.com
 (mail-fx0-f216.google.com [209.85.220.216]) by mta15.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail] (ORCPT [e-mail]); Wed,
 03 Mar 2010 07:45:33 -0500 (EST)
Received: by mail-fx0-f216.google.com with SMTP id 8so1449348fxm.31 for
 <[e-mail]>; Wed, 03 Mar 2010 04:45:33 -0800 (PST)
Received: by 10.223.76.132 with SMTP id c4mr8195808fak.106.1267620332789; Wed,
 03 Mar 2010 04:45:32 -0800 (PST)
Date: Wed, 03 Mar 2010 07:45:32 -0500
From: Peter Clough <[e-mail]>
Subject: send from gmail web
Message-id: <[e-mail]>
MIME-version: 1.0
Content-type: text/plain; charset=ISO-8859-1
Content-transfer-encoding: 7BIT
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;        d=gmail.com;
 s=gamma; h=domainkey-signature:mime-version:received:date:message-id:subject
 :from:content-type;        bh=AskMxZ5NqbTgJeUAWpFlTKruLtXXHajZ6hAOogvxMQg=;
  b=UCYgTmATAAqJs1reZnwGPnubIn4xDuW4SLDSDYR+cyBce5WrcoGae5uKeStgX3Pa3T
 wSaWL/fXaORjUcp7k/1a2b55Wd7hy9EAylGuMTiO2G3F5q4x+ajDNbHnfMZKkEbG73Ro
 ZO4eNvYH9MqhDt5jNIrzEb3PV5OzP04CtARD4=
DomainKey-Signature: a=rsa-sha1; c=nofws;        d=gmail.com; s=gamma;
 h=mime-version:date:message-id:subject:from:content-type;
 b=jtKx3vxBmDSTm9+jTHQp7kO+IbqkUf+US87Iim2M75Kz8WyXZwZwcP1ZCsHyKRw1Gf
 Y6fCIieMvWI0queT2dIFpMEtdOcsxBsxhXqc9rLOb3rA5qxObpI0wuWk5Ya3kGGRjrXV
 GDY9zYbhBpg76oznHDQLkGCFkNGA9JwXVNMws=
Original-recipient: rfc822;[e-mail]

send from gmail web


---------------------------------------

In sent mail folder:
 
From - Wed Mar 03 07:50:48 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
MIME-Version: 1.0
Received: by 10.223.122.199 with HTTP; Wed, 3 Mar 2010 04:45:32 -0800 (PST)
Bcc: [e-mail]
Date: Wed, 3 Mar 2010 07:45:32 -0500
Delivered-To: [e-mail]
Message-ID: <[e-mail]>
Subject: send from gmail web
From: Peter Clough <[e-mail]>
Content-Type: text/plain; charset=ISO-8859-1

send from gmail web

-----------------

In ALL Mail folder:

From - Wed Mar 03 07:55:44 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
MIME-Version: 1.0
Received: by 10.223.122.199 with HTTP; Wed, 3 Mar 2010 04:45:32 -0800 (PST)
Bcc: [e-mail]
Date: Wed, 3 Mar 2010 07:45:32 -0500
Delivered-To: [e-mail]
Message-ID: <[e-mail]>
Subject: send from gmail web
From: Peter Clough <[e-mail]>
Content-Type: text/plain; charset=ISO-8859-1

send from gmail web

-----------------------------------------------


>>and is copyable and moveable from folder to folder within TB without losing >>anything.

>It's not surprizing. The mail is same as mail-2/mail-3 in my Comment #16.

It surprises me.  It is not the same as the examples in comment 16 since it was NOT SENT BY GMAIL.SMTP.  It was sent from a non-gmail POP account.

Wada, please don't assume that I am giving you false information when the testing results don't match your theory.  I want this fixed and I am very careful with my tests so that we have an accurate idea of what and where the problem lies.  As you will see from some of my shouting above, I am frustrated when you assume - since the results don't match your theory of what is happening - that I have made a mistake.

Peter

Comment 30

[WADA:World Anti-bad-Duping Agency]

> >Check next, please. (Show "Order Received" column. UID is shown)
> >1. Send a mail to your POP3 account via Gmail's Web Interface, with BCC:
> >2. Sent mail copy is held in "Sent Mail"(==[Gmail]/Sent Mail via Gmail IMAP).
> >The mail is also held in [Gmail]/All Mail. Say UID=(X).
> >3. Copy arrived mail in Inbox of POP3 acount(no BCC:, many Received:) to
> >a Gmail IMAP folder. Mail is also held in [Gmail]/All Mail. Say UID=(Y).
> >4. Two different mails of different UID appear at [Gmail]/All Mail?
> >(one : with BCC:, no Received. another : without BCC:, many Received:)
> >If single mail only, which mail? UID of the mail is changed?
> 
> Here is the source of the note sent as you requested as it appears in Pop
> client as received, in gmail sent mail and as it appears in gmail ALL Mail:
>(snip)

Thanks for first attaching of real data(==evidence) instead of your wording only.
(1) Gmail(Web) put mail with BCC:, with next Received:
    and added Gmail Label of [Gmail]/Sent Mail and [Gmail]/Sent Mail.
> Received: by 10.223.122.199 with HTTP; Wed, 3 Mar 2010 04:45:32 -0800 (PST)
(2) After upload of mail arrived at POP3 server(many Received:, no BCC:) by Tb,
    Gmail kept mail data at step (1) instead of mail data uploaded at step (2).
    And, Gmail presents data at step (1) as data of uploaded mail at step (2).  

Can you check next, and compare with above case?
1. Send a mail to your POP3 account via SMTP of non Gmail, with BCC:,
   with no Gmail's mail address in To:/CC:/BCC:,
   with saving sent mail copy(with BCC:) to folder of "Local Folders".
   Automatic BCC-ing to Gmail's mail-addr should be disabled.
   Or, mail is sent to Gmail and confusion is produced.
2. Difference from "via Gmail Web" case:
     Gmail's SMTP, Gmail IMAP, Gmail is irrelevant.
   At this step, Gmail should be absolutely irrelevant for Tb.
3. Upload sent mail copy in "Local Folders"(with BCC:, usually no Received:)
   to [Gmail]/Sent Mail via Gmail IMAP using Tb.
   Check mail in "[Gmail]/Sent Mail" and "[Gmail]/All Mail".
   Is BCC: header seen in mail source?
4. Upload arrived mail to pop3 acount(no BCC:, many Received:)
   to "Inbox" via Gmail IMAP using Tb.
   Check mail in "Ibox", "[Gmail]/All Mail", and "[Gmail]/Sent Mail". 
   Is BCC: header and and Received: headers seen in mail source?

By the way, following is mail source in [Gmail]/Sent Mail(Inbox and [Gmail]/All Mail too) when next To:/CC:/BCC: is specified at mail composition form of Gmail's Web Interface.
    From: Yatter One <[e-mail]>
    To:   To myself  <[e-mail]> 
    CC:   CC myself  <[e-mail]> 
    BCC:  BCC myself <[e-mail]>, [e-mail]
Gmail looks to ignore excess CC:/BCC: or myself in CC:/BCC:.
As BCC: header exists, the mail data is "copy of sent mail" by Gmail instead of "mail data stream transfered to outside of Gmail's SMTP".
As you saw, Gmail presented me this mail data(with BCC:, one Received:) after uploaded of "arrived mail to pop3 account"(no BCC:, many Received:).
 
> MIME-Version: 1.0
> Received: by 10.142.180.9 with HTTP; Wed, 3 Mar 2010 17:52:50 -0800 (PST)
> Bcc: [e-mail]
> Date: Thu, 4 Mar 2010 10:52:50 +0900
> Delivered-To: [e-mail]
> Message-ID: <[e-mail]>
> Subject: To/CC/BCC myself, BCC pop3 account, via Gmail Web
> From: Yatter One <[e-mail]>
> To: To myself <[e-mail]>
> Content-Type: text/plain; charset=ISO-8859-1
>
> To/CC/BCC myself, BCC pop3 account, via Gmail Web

Comment 31

[pclough]

>Thanks for first attaching of real data(==evidence) instead of your wording
>only.

I hope you had a nice holiday yesterday.  perhaps you will be more receptive today.  This was far from the first "evidence " I've posted.  Take a look at the attachments.

>(1) Gmail(Web) put mail with BCC:, with next Received:
>    and added Gmail Label of [Gmail]/Sent Mail and [Gmail]/Sent Mail.
>> Received: by 10.223.122.199 with HTTP; Wed, 3 Mar 2010 04:45:32 -0800 (PST)
>(2) After upload of mail arrived at POP3 server(many Received:, no BCC:) by Tb,
>    Gmail kept mail data at step (1) instead of mail data uploaded at step (2).
>    And, Gmail presents data at step (1) as data of uploaded mail at step (2).  

>Can you check next, and compare with above case?
>1. Send a mail to your POP3 account via SMTP of non Gmail, with BCC:,
>   with no Gmail's mail address in To:/CC:/BCC:,
>   with saving sent mail copy(with BCC:) to folder of "Local Folders".
>   Automatic BCC-ing to Gmail's mail-addr should be disabled.
>   Or, mail is sent to Gmail and confusion is produced.
>2. Difference from "via Gmail Web" case:
>     Gmail's SMTP, Gmail IMAP, Gmail is irrelevant.
>   At this step, Gmail should be absolutely irrelevant for Tb.
>3. Upload sent mail copy in "Local Folders"(with BCC:, usually no Received:)
>   to [Gmail]/Sent Mail via Gmail IMAP using Tb.
>   Check mail in "[Gmail]/Sent Mail" and "[Gmail]/All Mail".
>   Is BCC: header seen in mail source?
>4. Upload arrived mail to pop3 acount(no BCC:, many Received:)
>   to "Inbox" via Gmail IMAP using Tb.
>   Check mail in "Ibox", "[Gmail]/All Mail", and "[Gmail]/Sent Mail". 
>   Is BCC: header and and Received: headers seen in mail source?

----------------------------------

As shown in pop (non-gmail) Sent folder

From - Thu Mar 04 08:19:13 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00800000
X-Mozilla-Keys:                                                                                 
BCC: Peter N Clough <[e-mail]>
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 08:19:08 -0500
From: "Peter N. Clough" <[e-mail]>
Reply-To: [e-mail]
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: bcc from pop to pop
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

bcc from pop to pop

----------------------------------

as shown in gmail sent mail folder after I copied it there:


From - Thu Mar 04 08:24:03 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
Return-Path: <[e-mail]>
Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
        by mx.google.com with ESMTPS id 3sm2784898fge.20.2010.03.04.05.19.09
        (version=TLSv1/SSLv3 cipher=RC4-MD5);
        Thu, 04 Mar 2010 05:19:10 -0800 (PST)
Sender: Peter Clough <[e-mail]>
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 08:19:08 -0500
From: "Peter N. Clough" <[e-mail]>
Reply-To: [e-mail]
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: bcc from pop to pop
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

bcc from pop to pop

----------------------------------

as shown in all mail:

From - Thu Mar 04 08:27:27 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
Return-Path: <[e-mail]>
Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
        by mx.google.com with ESMTPS id 3sm2784898fge.20.2010.03.04.05.19.09
        (version=TLSv1/SSLv3 cipher=RC4-MD5);
        Thu, 04 Mar 2010 05:19:10 -0800 (PST)
Sender: Peter Clough <[e-mail]>
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 08:19:08 -0500
From: "Peter N. Clough" <[e-mail]>
Reply-To: [e-mail]
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: bcc from pop to pop
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

bcc from pop to pop

----------------------------




The bcc has disappeared, yet gmail has never seen more than one copied mail.

I did not do your step 4.  There was no point, since the bcc's were already gone.

Comment 32

[WADA:World Anti-bad-Duping Agency]

(In reply to comment #31)

Thanks for second attaching of real data(==evidence) to comment #31(29th & 31th comment!) instead of your wording only.

> I did not do your step 4.  There was no point, since the bcc's were already gone.

Why I requested you to check step 4 is I wanted to say;
  Please check whether Gmail replaces mail data or not, if you upload mail
  data which arrived at POP3 server(you say sent by non Gmail's SMTP. So,
  BCC: shouldn't exist, and many Received: headers usually exist).

Following is Received: header(only one) in mail data which you say "Gmail has".
> Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
>        by mx.google.com with ESMTPS id 3sm2784898fge.20.2010.03.04.05.19.09
>        (version=TLSv1/SSLv3 cipher=RC4-MD5);
>        Thu, 04 Mar 2010 05:19:10 -0800 (PST)

"by mx.google.com" in the only one Received: header indicates that fisrt server received the mail data is mx.google.com(from ool-43572b14.dyn.optonline.net [67.87.43.20], which is probably you used to send the mail).
Why can such header appear in mail data which Gmail has, even though you sent via non-Gmail SMTP without requesting to send mail to Gmail's mail address, by test procedure on comment #30/comment #31?

Can you attach mail data arrived at POP3 server(not Gmail) which probablyy has many Received: headers and no BCC: header produced by your test of procedure of comment #30/comment #31?

Comment 33

[pclough]

In reply to comment 32:

>(In reply to comment #31)

>Thanks for second attaching of real data(==evidence) to comment #31(29th & 31th
>comment!) instead of your wording only.

For some reason you persist in saying things in an offensive way.  I have submitted many attachments as "evidence," and it makes me angry when you say this is only the second time I have included evidence.  

>> I did not do your step 4.  There was no point, since the bcc's were already gone.

>Why I requested you to check step 4 is I wanted to say;
>  Please check whether Gmail replaces mail data or not, if you upload mail
>  data which arrived at POP3 server(you say sent by non Gmail's SMTP. So,
>  BCC: shouldn't exist, and many Received: headers usually exist).


>Following is Received: header(only one) in mail data which you say "Gmail has".
>> Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
>>        by mx.google.com with ESMTPS id 3sm2784898fge.20.2010.03.04.05.19.09
>>        (version=TLSv1/SSLv3 cipher=RC4-MD5);
>>        Thu, 04 Mar 2010 05:19:10 -0800 (PST)

>"by mx.google.com" in the only one Received: header indicates that fisrt server
>received the mail data is mx.google.com(from ool-43572b14.dyn.optonline.net
>[67.87.43.20], which is probably you used to send the mail).
>Why can such header appear in mail data which Gmail has, even though you sent
>via non-Gmail SMTP without requesting to send mail to Gmail's mail address, by
>test procedure on comment #30/comment #31?

>Can you attach mail data arrived at POP3 server(not Gmail) which probablyy has
>many Received: headers and no BCC: header produced by your test of procedure of
>comment #30/comment #31?

I'll do a retest:

from POP sent folder:

From - Thu Mar 04 14:29:12 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00800000
X-Mozilla-Keys:                                                                                 
BCC: here <[e-mail]>
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 14:29:09 -0500
From: Peter n Clough <[e-mail]>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: retest pop to pop 14:29
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
retest pop to pop 14:29
</body>
</html>

---------------------

as received in POP account:
From - Thu Mar 04 14:32:01 2010
X-Account-Key: account1
X-UIDL: 835-1251170511
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
X-Mozilla-Keys:                                                                                 
Return-path: <[e-mail]>
Return-path: <[e-mail]>
Received: from mta25.srv.hcvlny.cv.net
 (mta25.srv.hcvlny.cv.net [167.206.5.186]) by mstr27.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail]; Thu, 04 Mar 2010 14:29:13 -0500 (EST)
Received: from mail-bw0-f225.google.com
 (mail-bw0-f225.google.com [209.85.218.225]) by mta25.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail] (ORCPT [e-mail]); Thu,
 04 Mar 2010 14:29:13 -0500 (EST)
Received: by bwz25 with SMTP id 25so840692bwz.37 for <[e-mail]>;
 Thu, 04 Mar 2010 11:29:12 -0800 (PST)
Received: by 10.103.21.14 with SMTP id y14mr259553mui.87.1267730952122; Thu,
 04 Mar 2010 11:29:12 -0800 (PST)
Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
 by mx.google.com with ESMTPS id 14sm612918bwz.10.2010.03.04.11.29.10
 (version=TLSv1/SSLv3 cipher=RC4-MD5); Thu, 04 Mar 2010 11:29:11 -0800 (PST)
Date: Thu, 04 Mar 2010 14:29:09 -0500
From: Peter n Clough <[e-mail]>
Subject: retest pop to pop 14:29
Message-id: <[e-mail]>
MIME-version: 1.0
Content-type: text/html; charset=ISO-8859-1
Content-transfer-encoding: 7BIT
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;        d=gmail.com;
 s=gamma;        h=domainkey-signature:received:received:message-id:date:from
     :user-agent:mime-version:subject:content-type :content-transfer-encoding;
 bh=O8DYXhYMkJfS6RI9tHDEdf9O4JE+Ne8kTl/CHldnDXM=;
 b=K7TThS74+AxGU24SN7UthtC2uMPlfkUpZrvx1tlnNXI5Wk72EWE3hbWY8Trw+c9vCY
 Ez0YXeef5e407zTffBVaWsgKqfWoGGER9eVpwwNSb1e4j0CRqFOakBDN3Gb1FIPzCflI
 sKrTV0+8yLWFfV3HMY7FylqVfCmB7vg907L6I=
DomainKey-Signature: a=rsa-sha1; c=nofws;        d=gmail.com; s=gamma;
 h=message-id:date:from:user-agent:mime-version:subject:content-type
 :content-transfer-encoding;
 b=K7BXPIvgheRhWRCH5kDaBy+jzAeRL4RaUkEJDsj6n7oiKBs8GccC5qYX+d2DvJRfUG
 Xw1rdBJPJdE8NWaYNahjpLJcOb/pRVCF4I/UoEOn0qt+ZXFxT86yJ+KQEhsInS+8Gu3t
 HQXeLaxg0F1hKmTgD1pOqaf5bzyGiMvKSSjGM=
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
Original-recipient: rfc822;[e-mail]

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
retest pop to pop 14:29
</body>
</html>


----------------------------------



as shown in gmail sent mail folder after copying it there from POP sent folder:

From - Thu Mar 04 14:37:20 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
Return-Path: <[e-mail]>
Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
        by mx.google.com with ESMTPS id 14sm612918bwz.10.2010.03.04.11.29.10
        (version=TLSv1/SSLv3 cipher=RC4-MD5);
        Thu, 04 Mar 2010 11:29:11 -0800 (PST)
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 14:29:09 -0500
From: Peter n Clough <[e-mail]>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: retest pop to pop 14:29
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
retest pop to pop 14:29
</body>
</html>

-------------------------------------

as shown in gmail inbox after copy from POP inbox:

From - Thu Mar 04 14:41:41 2010
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
Return-Path: <[e-mail]>
Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
        by mx.google.com with ESMTPS id 14sm612918bwz.10.2010.03.04.11.29.10
        (version=TLSv1/SSLv3 cipher=RC4-MD5);
        Thu, 04 Mar 2010 11:29:11 -0800 (PST)
Message-ID: <[e-mail]>
Date: Thu, 04 Mar 2010 14:29:09 -0500
From: Peter n Clough <[e-mail]>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
Subject: retest pop to pop 14:29
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
retest pop to pop 14:29
</body>
</html>


---------------------

I don't know what all that proves that we didn't know before, except that not just BCCs get stripped out when notes are copied into gmail.  Who is doing the stripping?

Comment 34

[WADA:World Anti-bad-Duping Agency]

Following is first Received:(bottom most in mail source) which is sent from my PC to a my POP3 account.

(A) via SMTP server of DTI(www.dti.ne.jp) 
> Received: from [192.168.0.10] (KHP222009106185.ppp-bb.dion.ne.jp [222.9.106.185])
> by smtp09.dti.ne.jp (3.11s) with ESMTP AUTH id o224MXZr003188;Tue, 2 Mar 2010 13:22:33 +0900 (JST)

(B) via SMTP server of Gmail(smtp.gmail.com)
> Received: from [192.168.0.10] (KHP222009106185.ppp-bb.dion.ne.jp [222.9.106.185])
> by mx.google.com with ESMTPS id 23sm693258qyk.7.2010.03.04.16.35.16
> (version=TLSv1/SSLv3 cipher=RC4-MD5);

Following is first Received:(bottom most in mail source) of mail you pasted.

> Received: from [10.0.0.2] (ool-43572b14.dyn.optonline.net [67.87.43.20])
> by mx.google.com with ESMTPS id 14sm612918bwz.10.2010.03.04.11.29.10
> (version=TLSv1/SSLv3 cipher=RC4-MD5); Thu, 04 Mar 2010 11:29:11 -0800 (PST)

I assumed that your SMTP server(dyn.optonline.net, non Gmail) didn't add first Received: (log of "from your PC by your-SMTP with ESMTP id ...") when I wrote previous comment, because you repeatedly said "really sent via non Gmail SMTP server". But it seems wrong...

(In reply to comment #29)
> NO NO NO. It was sent with another smtp server. Please do not ask me this again.

I dislike this kind of response without data(evidence), without correct test result check, without check of "test is proper or not".

Comment 35

[WADA:World Anti-bad-Duping Agency]

FYI.

Next mail is sent via non Gmail SMTP.
  Folder to save sent mail copy = root-level "Sent" of [e-mail].
  From: "M. Wada" <[e-mail]>
  (checked with this from addr, but any valid addr can be used, if valid setup)
  To:   [e-mail]
  BCC:  [e-mail]
(A) "Sent mail copy by Tb(with BCC:)" was kept by Gmail. As I wrote before, this is same as "mail send via Gmail Web" case.
(B) Arrived mail didn't appear in Inbox of the Gmail IMAP account. Gmail doesn't seem to add Gmail Label of "Inbox" in this case. (difference from "mail send via Gmail Web" case). I was not aware of it in prvious test.
So, some parts of your bug summary are wrong.

(Sent mail copy by Tb to root-level "Sent" of Gmail IMAP)
> BCC: [e-mail]
> Message-ID: <[e-mail]>
> Date: Fri, 05 Mar 2010 11:23:57 +0900
> From: "M. Wada" <[e-mail]>
> User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8)
> Gecko/20100216 Thunderbird/3.0.2
> MIME-Version: 1.0
> To: [e-mail]
> Subject: mail send test 04
> Content-Type: text/plain; charset=UTF-8; format=flowed
> Content-Transfer-Encoding: 7bit
> 
> mail send test 04

(mail arrived at a POP3 account, [e-mail]. Sent because put in BCC:)
> From - Fri Mar 05 11:24:28 2010
> X-Account-Key: account410
> X-UIDL: 36704
> X-Mozilla-Status: 0001
> X-Mozilla-Status2: 00000000
> X-Mozilla-Keys:
> Return-Path: <[e-mail]>
> Received: from mail-gw.auone-net.jp ([210.141.108.180])
> 	by nm02mta.auone-net.jp (snip)
>(snip, many Received: headers)
> Received: from smtp09.dti.ne.jp (smtp09.dti.ne.jp [202.216.231.184])
> 	by mtain-mc05.r1000.mx.aol.com (Internet Inbound) with ESMTP id 0E20638000095
> 	for <[e-mail]>; Thu,  4 Mar 2010 21:24:00 -0500 (EST)
> Received: from [192.168.0.10] (KHP222009106185.ppp-bb.dion.ne.jp [222.9.106.185]) by smtp09.dti.ne.jp (3.11s) with ESMTP AUTH id o252NvEj005602;Fri, 5 Mar 2010 11:23:58 +0900 (JST)
> Message-ID: <[e-mail]>
> Date: Fri, 05 Mar 2010 11:23:57 +0900
> From: "M. Wada" <[e-mail]>
> User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1.8) Gecko/20100216 Thunderbird/3.0.2
> MIME-Version: 1.0
> To: [e-mail]
> Subject: mail send test 04
> Content-Type: text/plain; charset=UTF-8; format=flowed
> Content-Transfer-Encoding: 7bit
> x-aol-global-disposition: G
> x-aol-sid: 3039ac1d604d4b906b403a04
> X-AOL-IP: 202.216.231.184
> X-Mailer: Unknown (No Version)
> X-Spam-Flag:NO
> X-AOL-SENDER: [e-mail]
> X-SPF-AUTH: SoftFail (mail-gw.auone-net.jp: transitioning domain of ops.dti.ne.jp does not designate 64.12.100.31 as permitted sender) client-ip=64.12.100.31; envelope-from=<[e-mail]>; helo=imr-ma05.mx.aol.com; domain=ops.dti.ne.jp; txt=v=spf1 ; auth=v1;
> 
> mail send test 04

Comment 36

[pclough]

Wada,

I have no idea what your last post means.  You don't explain, and as far as I can tell we are not one millimeter closer to an explanation or a fix to the bug I submitted. than when we started.

But you have succeeded in discouraging me from continuing.  Dealing with you is frustrating to me and I am insulted by many of your comments.  Apparently I also frustrate you - not sure why.  In any case, I GIVE UP.

If someone else is willing to work this, I'm prepared to help, but no longer with WADA.

Peter Clough

Comment 37

[WADA:World Anti-bad-Duping Agency]

I was very frustrated because you repeatedly said "mail is really sent by non Gmail SMTP server" even though first Received: header of mail source was written by Gmail's SMTP sever, even though I requested you to check about "data for the mail is not held by Gmail", even though I requested you to check Received: header.
Comment #37 is example of mail data of "mail was really sent via non Gmail SMTP server".

I wanted you to check "mail is really sent via non Gmail SMTP server" case(case-3 in below) by yourself, in order to correctly understand what happens, and what is real problem.
To understand what happens on "sent mail copy with BCC: by Tb", I thought you need to know what happens in next cases by yourself.
  (case-1) mail is sent via Gmail's SMTP (all of your attached data)
  (case-2) mail is sent via Gmail's Web Interface
           (you provided test result, and I could observe too) 
  (case-3) mail is really sent via non Gmail SMTP
           (comment #37, mail-2/3 in comment #16)
           This case is almost same as case-2.
  (case-4) Upload of some versions of same mail by Tb, using crafted mail.
           For example:
           v-1 : No Date: header, no User-Agent:, no Mime-Version:, ...
           v-2 : with Mime-Version:, v-3: with User-Agent:, with Mime-Version:
           v-4 : with Date:, with User-Agent:, with Mime-Version:
           Gmail probably doesn't replace old version by newly updated
           newer version. (v-4 may be treated as different mail)
In any case, problem is never "BCC: header is lost or disappears". As I repeatedly said, phenomenon in all cases is that Gmail doesn't replace already held mail data by newly uploaded/arrived mail data of same mail but with different headers. So, I wanted you to know what happens in case-3/case-4. 

I can't help people any more who says repeatedly "mail was really sent by non Gmail SMTP server" even though first Received: header of mail source was written by Gmail's SMTP sever. So, as you want, I'll never add comment to this bug.

Comment 38

[pclough]

Wada,

This is a perfect example of how you frustrate:

>In any case, problem is never "BCC: header is lost or disappears". As I
>repeatedly said, phenomenon in all cases is that Gmail doesn't replace already
>held mail data by newly uploaded/arrived mail data of same mail but with
>different headers.

For the benefit of others reading this bug, here is my reaction:

That statement is just plain wrong.  In an example like comment 33, the mail is sent from a non gmail account not using gmail smtp.  The BCC shows up fine in Local Folders.  But when it is copied from Local Folders to gmail Sent Mail, the BCC disappears.  Gmail has seen one and only one copy of the mail:  the one with the BCC.  So it has nothing to do with replacement in that instance, even if the replacement theory is correct in general.

The replacement theory also fails a logic test for the mail created on web with Gmail which shows the BCC.  That mail shows the BCC in Sent Mail folder and anywhere it is copied to.  In fact it shows only one copy, the one with the BCC;  it never - as far as I can see - shows a copy without the BCC.  So if gmail can do it for its own notes with bcc, how come TB can't do it?

-------------------------------------

Note sent gmail web interface as shown in IMAP sent mail:

MIME-Version: 1.0
Received: by 10.223.122.199 with HTTP; Thu, 4 Mar 2010 20:50:35 -0800 (PST)
Bcc: here <[e-mail]>
Date: Thu, 4 Mar 2010 23:50:35 -0500
Delivered-To: [e-mail]
Message-ID: <[e-mail]>
Subject: gmail web to non-gmail
From: Peter Clough <[e-mail]>
Content-Type: text/plain; charset=ISO-8859-1

gmail web to non-gmail

-----------------------------------

Note as shown in recipient inbox:

From - Thu Mar 04 23:51:31 2010
X-Account-Key: account2
X-UIDL: 837-1251170511
X-Mozilla-Status: 0001
X-Mozilla-Status2: 00000000
X-Mozilla-Keys:                                                                                 
Return-path: <[e-mail]>
Received: from mta28.srv.hcvlny.cv.net (mta28.srv.hcvlny.cv.net [167.206.5.99])
 by mstr27.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail]; Thu, 04 Mar 2010 23:50:36 -0500 (EST)
Received: from mail-fx0-f221.google.com
 (mail-fx0-f221.google.com [209.85.220.221]) by mta28.srv.hcvlny.cv.net
 (Sun Java System Messaging Server 6.2-8.04 (built Feb 28 2007))
 with ESMTP id <[e-mail]> for
 [e-mail] (ORCPT [e-mail]); Thu,
 04 Mar 2010 23:50:36 -0500 (EST)
Received: by fxm21 with SMTP id 21so668929fxm.5 for <[e-mail]>;
 Thu, 04 Mar 2010 20:50:35 -0800 (PST)
Received: by 10.223.4.211 with SMTP id 19mr339156fas.72.1267764635159; Thu,
 04 Mar 2010 20:50:35 -0800 (PST)
Date: Thu, 04 Mar 2010 23:50:35 -0500
From: Peter Clough <[e-mail]>
Subject: gmail web to non-gmail
Message-id: <[e-mail]>
MIME-version: 1.0
Content-type: text/plain; charset=ISO-8859-1
Content-transfer-encoding: 7BIT
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;        d=gmail.com;
 s=gamma; h=domainkey-signature:mime-version:received:date:message-id:subject
 :from:content-type;        bh=sUh/mmg7LMMyk/kQIk/jsi045fWd6BEb5g6d5mkFeqM=;
  b=KYnWShHx3vNbSmP8PR5BBQKW2qZDC118hn7Nm8KSzqGlGpvoBd8eXGv+TM1FXTjo0M
 opKtE6ll0dhlIy4rc9+MnqVAuxPqIvINHCAVdiXapPJW1nx2ASYFkpq5jvlsaE332F/6
 n7WnwAElYh28gP9RGMVancKqLnCmr0/9ed3B8=
DomainKey-Signature: a=rsa-sha1; c=nofws;        d=gmail.com; s=gamma;
 h=mime-version:date:message-id:subject:from:content-type;
 b=gKNKrX/iwdIr4uMVGCdQCt498OY5/JaFAIC9thzk8uski3esjkw/SzSnol5A0WMiZf
 kvwdSRG9vifuxQMufnCP3eNy9SAJzCPE5QJ0Mb4JLKVD2fAougB5fgEGSjNFoWMlgyS7
 COTmAOD1yd4/Wd+bWPNP+Z9uVSpvBmFpuhv0A=
Original-recipient: rfc822;[e-mail]

gmail web to non-gmail

------------------------------------------------

Note copied from gmail sent mail to _Test BCC folder (note BCC still present):

MIME-Version: 1.0
Received: by 10.223.122.199 with HTTP; Thu, 4 Mar 2010 20:50:35 -0800 (PST)
Bcc: here <[e-mail]>
Date: Thu, 4 Mar 2010 23:50:35 -0500
Delivered-To: [e-mail]
Message-ID: <[e-mail]>
Subject: gmail web to non-gmail
From: Peter Clough <[e-mail]>
Content-Type: text/plain; charset=ISO-8859-1

gmail web to non-gmail

Comment 39

[pclough]

Wada,

I owe you an apology.  Last night I did more testing and found that what you were saying is correct.  I normally have all accounts set to use smtp.gmail.com.  I was changing the outgoing mail server to that of my service provider optonline, but because I did not close and restart TB it continued to use gmail.  

The result of my testing last night shows that you were right and I was wrong.  If the mail is sent with a non-gmail smtp a copy with BCC can be saved in sent mail (or another folder) but if gmail smtp has seen the mail it cannot.

One can imagine TB providing a solution by writing the copy before passing the note to smtp.  There are several problems with that:  having to delete the copy if the send fails, specific code for a specific smtp server, and probably internal things I don't know about.  It also would be a feature not a bug fix.

Our communication could have been better, but the fault is mine and once again I humbly apologize.

The bug should be closed with a succinct explanation.

Peter Clough

Comment 40

[pclough]

See bug 55073 for a better synopsis of the problem.