220246 - fails to authenticate when copying message to Sent folder

Status: RESOLVED FIXED

(MailNews Core :: Networking: IMAP, defect)

Description

[Brian Ryner (not reading)]

current trunk build, Linux.

If I compose a message prior to logging in to my IMAP server, when I try to send
the message, it is not able to copy it to the Sent folder.  The progress dialog
shows that it sends the message ok.  Then, while the progress dialog says
"Copying message to Sent folder", I get an alert saying "Login to server
mail.brianryner.com failed."  At no point am I prompted for my IMAP password. 
Sometimes this alert appears multiple times.  At any rate, I then get "The
message was sent successfully, but could not be copied to your Sent folder."

The problem does not occur if I log in to read mail before I send the new
message.  Also, if it matters, I'm using authenticated SMTP (no SSL).

I believe this is a recent regression as I do this regularly and have never
noticed a problem.

Comment 1

[Brian Ryner (not reading)]

David, any chance this is a regression from bug 219976?

Comment 2

[David :Bienvenu]

Brian, it's possible, I guess - do you have remember password turned on or not
for  that account?

Comment 3

[Brian Ryner (not reading)]

For SMTP, yes.  For IMAP, no.

Comment 4

[Brian Ryner (not reading)]

Attachment: fix

the bug was caused by the fact that rv was still set to a failure code if no
saved password was found... so if aMsgWindow is null, rv was never reset and it
would skip showing the dialog.

Comment 5

[David :Bienvenu]

How about just:

     // Get password entry corresponding to the host URI we are passing in.
      if (NS_SUCCEEDED(passwordMgrInt->FindPasswordEntry(currServerUri,
NS_LITERAL_STRING(""), NS_LITERAL_STRING(""),
                                             hostFound, userNameFound,
passwordFound))
        CopyUCS2toASCII(passwordFound, m_password);
 
i.e., skip the setting of rv completely? I can make the patch and check it in if
you want, I'm sorry about the regression.

Comment 6

[David :Bienvenu]

Attachment: proposed fix

Comment 7

[Brian Ryner (not reading)]

Comment on attachment 132364 [details] [diff] [review]
proposed fix

That's fine too... I was just eliminating a check on rv where it wasn't clear
what we were even checking for failure of (since any real failure would have
caused an early return before that check).

Comment 8

[David :Bienvenu]

both fixes checked in.